1. Business users gain insights from activity-based costing (ABC) information on which products, services, channels and customers are relatively more or less profitable. However, ABC alone does not provide sufficient insight into what differentiates highly profitable from less profitable customers.
2. Data mining and advanced analytics techniques like decision trees and recursive partitioning can identify the key drivers that best explain differences in profitability between high-profit and low-profit customers. Knowing these drivers can guide actions to increase profit lift from customers.
3. The paper describes how these analytical techniques were applied to determine differentiating characteristics, like customer location, that correlated with profitability levels and provided guidance on targeted marketing and sales strategies.
Acquire Grow & Retain customers - The business imperative for Big DataIBM Software India
The emergence of Big Data and Analytics has changed the way marketing decisions are made. Marketing has moved away from traditional ‘generalisation’ practices such as customer segmentation, geographical targeting etc. and is focussing more on the individual – the ‘Chief Executive Customer’.
This document discusses analytics and information architecture. It begins by describing how analytics workloads are moving away from data warehouses to more specialized platforms. It then discusses what distinguishes analytics from reporting, including that analytics involve complex summaries of information and linking analyses to business actions. The document examines various data platforms used for analytics and contends that ParAccel Analytic Database is well-suited for analytics workloads due to its columnar structure, compression, SQL support, and ability to utilize Hadoop data without replication. It concludes by proposing an information architecture with Hadoop for big data, ParAccel for analytics, and data warehouses for operational support.
This whitepaper is geared to help
bank marketing professionals
understand the scope of marketing
analytics and also on how it can
contribute value to the various
factions of a bank’s marketing
activities.
The document discusses the Silvon Stratum suite of operational planning, analysis and reporting applications. It describes how Stratum provides manufacturers greater visibility into operations, decreased variability across the supply chain, and more accurate planning capabilities to help address challenges from globalization and increased competition. It highlights key features of Stratum, including pre-defined analyses and reports, planning applications, alerts, dashboards, and integration with ERP systems. The document invites readers to learn more about how Stratum can help solve their specific business issues.
1. Sales and marketing analytics uses descriptive, diagnostic, predictive, and mechanistic analytics types to generate insights from business data in areas like consumer behavior, customer segmentation, pricing, recommendations, and sales force performance.
2. Common applications of analytics include understanding consumer behavior, customer segmentation, marketing mix optimization, and sales force efficiency.
3. Popular tools for sales and marketing analytics include Zoho Analytics, Yellowfin, Looker, Microsoft Excel, and various report generation and data visualization software.
The document describes the key features of a customer analytics platform called Quiterian Analytics. It allows users to integrate customer data from multiple sources, explore and visualize the data, enrich and cleanse the data, perform advanced analytics and data mining, create dashboards, and automate marketing campaigns. The platform aims to provide a complete view of customers and help companies gain insights, improve strategic decision making, and anticipate customer behavior.
Analytics For Retail Banking - MarketelligentMarketelligent
MarketIntelligent provides analytic services to help clients make better business decisions. They offer expertise in credit risk and marketing analytics across various banking products. Their services include developing scorecards to predict customer behavior, maximize profits from assets and fees, reduce losses, acquire profitable customers, increase activation and cross-sell revenues.
Acquire Grow & Retain customers - The business imperative for Big DataIBM Software India
The emergence of Big Data and Analytics has changed the way marketing decisions are made. Marketing has moved away from traditional ‘generalisation’ practices such as customer segmentation, geographical targeting etc. and is focussing more on the individual – the ‘Chief Executive Customer’.
This document discusses analytics and information architecture. It begins by describing how analytics workloads are moving away from data warehouses to more specialized platforms. It then discusses what distinguishes analytics from reporting, including that analytics involve complex summaries of information and linking analyses to business actions. The document examines various data platforms used for analytics and contends that ParAccel Analytic Database is well-suited for analytics workloads due to its columnar structure, compression, SQL support, and ability to utilize Hadoop data without replication. It concludes by proposing an information architecture with Hadoop for big data, ParAccel for analytics, and data warehouses for operational support.
This whitepaper is geared to help
bank marketing professionals
understand the scope of marketing
analytics and also on how it can
contribute value to the various
factions of a bank’s marketing
activities.
The document discusses the Silvon Stratum suite of operational planning, analysis and reporting applications. It describes how Stratum provides manufacturers greater visibility into operations, decreased variability across the supply chain, and more accurate planning capabilities to help address challenges from globalization and increased competition. It highlights key features of Stratum, including pre-defined analyses and reports, planning applications, alerts, dashboards, and integration with ERP systems. The document invites readers to learn more about how Stratum can help solve their specific business issues.
1. Sales and marketing analytics uses descriptive, diagnostic, predictive, and mechanistic analytics types to generate insights from business data in areas like consumer behavior, customer segmentation, pricing, recommendations, and sales force performance.
2. Common applications of analytics include understanding consumer behavior, customer segmentation, marketing mix optimization, and sales force efficiency.
3. Popular tools for sales and marketing analytics include Zoho Analytics, Yellowfin, Looker, Microsoft Excel, and various report generation and data visualization software.
The document describes the key features of a customer analytics platform called Quiterian Analytics. It allows users to integrate customer data from multiple sources, explore and visualize the data, enrich and cleanse the data, perform advanced analytics and data mining, create dashboards, and automate marketing campaigns. The platform aims to provide a complete view of customers and help companies gain insights, improve strategic decision making, and anticipate customer behavior.
Analytics For Retail Banking - MarketelligentMarketelligent
MarketIntelligent provides analytic services to help clients make better business decisions. They offer expertise in credit risk and marketing analytics across various banking products. Their services include developing scorecards to predict customer behavior, maximize profits from assets and fees, reduce losses, acquire profitable customers, increase activation and cross-sell revenues.
This document discusses data mining techniques for customer relationship management (CRM). It defines data mining as the extraction of implicit and novel knowledge from large datasets. The document outlines common data mining applications in retail, banking, telecommunications and other industries. It also discusses how data mining can be used across different stages of the customer lifecycle in CRM, such as up-selling, cross-selling and customer retention. Finally, it provides an overview of common predictive and descriptive data mining techniques like decision trees, rule induction, clustering and association rule mining.
This document discusses how The Data People uses data analytics to help businesses better understand their customers. They identify a company's most valuable customers through analyzing data on profitability, demographics, lifestyle, and usage. The Data People then builds customer profiles and segments customers to tailor marketing strategies. Case studies show how they helped Alliance & Leicester increase website visits by over 100% and Holmes Place reduce customer acquisition costs and increase retention rates through predictive modeling.
Consumer analytics is the process businesses adopt to capture and analyze customer data to make better business decisions via predictive analytics. It is a method of turning data into deep insights to predict customer behavior. It may also be regarded as the process by which data can be turned into predictive insights to develop new products, new ways to package existing products, acquire new customers, retain old customers, and enhance customer loyalty. It helps businesses break big problems into manageable answers. This paper is a primer on consumer analytics. Matthew N. O. Sadiku | Sunday S. Adekunte | Sarhan M. Musa "Consumer Analytics: A Primer" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-4 | Issue-6 , October 2020, URL: https://www.ijtsrd.com/papers/ijtsrd33511.pdf Paper Url: https://www.ijtsrd.com/other-scientific-research-area/other/33511/consumer-analytics-a-primer/matthew-n-o-sadiku
The document discusses how analytics can be used to solve business problems in the retail banking industry. It describes how analytics can be applied to various areas of a bank's profit and loss statement, including acquiring new customers, reducing customer attrition, improving account activation rates, and maximizing revenue from interest, fees, and cross-selling. It also discusses how strategic reporting, marketing analytics, and data-driven insights can be used for segmentation, customer lifetime value analysis, profitability and loyalty analysis, cross-selling strategies, and customer retention programs. The overall aim is to provide a top-down analytical approach to optimize all areas of a bank's operations and financial performance.
This document contains confidential information belonging to AAUM. It discusses various analytical techniques such as customer segmentation, market basket analysis, forecasting, and supply chain optimization that can be applied across industries. Case studies of companies like Tesco, Nieman Marcus, and Food Lion demonstrate how these techniques have been successfully used to increase sales, optimize operations, and improve customer experience.
Addressing the challenges and issues with businesses struggling to deliver successful Data Science in environments - Measure Camp, Bucharest (2 Nov. 2019)
Arun Gupta, Customer Care Associate and Group Chief Technology Officer, Shoppers Stop presented at the Premier Business Leadership Series 2010, http://www.sas.com/theserieshk.
With many retailers worldwide struggling to maintain revenues, how do you grow in such a tough competitive landscape? As a leading Indian retailer and pioneer in using technology, especially business analytics, Shoppers Stop is not only thriving but has helped revolutionise the retail sector. Gupta will share insights on using analytics to drive business value, reduce operational costs and provide better products and customer experience.
This document discusses analytics and retail analytics. It defines analytics as discovering patterns in data through statistics, programming, and research. Retail analytics specifically aims to improve customer loyalty and sales. It does this by identifying valuable customers, understanding their preferences, and creating personalized shopping experiences through offers targeted to individual needs. Retailers can gather customer data through in-store and online analytics to gain insights that optimize performance.
Today we are beyond the point where big data is in the prototype stage. We are entering an era where automation, integration and end-to-end solutions need to be built rapidly to facilitate disruption. Companies need to architect a
platform for Big Data (and traditional data) analytics
Case Studies - Customer & Marketing Analytics for Retail Gurmit Combo
The document discusses three case studies involving customer intelligence and marketing effectiveness services:
1. A luxury retailer case study where customer segmentation and profiling identified their most valuable customers to focus relationship management efforts.
2. A technology company case study where product association analysis and scoring identified accounts likely to purchase docking stations for targeted cross-selling.
3. A CPG company case study where regression modeling decomposed the impact of price, promotion, competition and cross-category effects on sales volumes, revealing promotion strategy optimizations.
Today there is a lot of buzz around customer experience. Many companies have realized that investments in customer experience improvement is important not just because it helps to boost the bottom lines of their businesses but because it takes at least 4 to 6 times more cost to acquire a new customer than to retain an existing customer.
Creating Business Value - Use Cases in CPG/RetailBig Data Pulse
This document discusses how big data analytics can help consumer packaged goods, fast moving consumer goods, retail, and e-commerce companies. It provides examples of use cases like predictive demand forecasting, pricing optimization, and markdown optimization. One case study describes how a department store used a forecasting and optimization model to improve markdown strategies and increase margins by $90 million annually. In conclusion, analyzing large, diverse customer data in real-time can provide actionable insights to increase market share, revenue and profits.
Project for System Analysis and Design (IS-6410).
By performing customer segmentation following are the three objectives which can be achieved
with the implementation of this new analytics system:
1. We can track the difference between loyal customers vs visitors, perform heat map
analysis of their browsing patterns.
2. Understanding customer demographics and to focus on high profitable segments.
3. Finally empowering our Marketing department to make better strategic decisions in
terms of online Ads/campaigns.
The document discusses the benefits of centralizing customer data into a single database. It notes that while customer data provides important insights, it is often scattered across different departments and systems, making it difficult to develop a comprehensive view. Centralizing data allows companies to better understand customers, target them across channels, and improve marketing efficiency. However, the database must be regularly maintained and enhanced with things like address updates to ensure the data remains accurate and useful over time.
Business analytics is used by industries to maximize operations and is applied in many fields including marketing, sales, finance, and human resources. Companies study consumer behavior patterns through social media, spending habits, and lifestyles to segment markets and identify target audiences. Business analytics tools help marketing and sales teams optimize their strategies, perform competitor analysis, and assess sales performance.
business analytics and its importance, marketing analytics definition and its importance, how marketing analytics helps to run the organization in effective and efficient manner.
Moving Forward with Big Data: The Future of Retail AnalyticsBill Bishop
Out new report Moving Forward with Big Data: The Future of Retail Analytics goes deeper into new territory that's relevant to changes taking place across retailing.
It calls out significant progress in the past 9 months.
• The definition of big data has grown beyond technical, i.e. “what it is,” to include “what it does.”
• A lot more companies are executing big data projects (an increase from < 20% to now 65% of sample respondents).
• Most of the focus is on driving top line growth.
This document discusses how The Data People helps companies identify their best customers and maximize profits through data-driven strategies and analytics. They analyze customer data to build detailed profiles, identify valuable customer segments, predict churn, and develop targeted marketing strategies. Case studies show how they helped companies like Alliance & Leicester increase website visits by 100% through improved targeting, and helped Nescafe launch a successful direct marketing campaign by creating an accurate customer profile model.
The document summarizes key findings from a survey of 560 marketers in the UK. Some of the main findings include:
- 91% of marketers believe the digital revolution is still ongoing and will be disrupted by younger digital native consumers.
- Social media is the channel being used most heavily, while spending on live events is also increasing.
- Marketers feel least confident about using mobile and data analytics is seen as the most important skill.
- The rise of empowered consumers and content marketing are seen as the biggest disruptors and influences on future marketing.
This document profiles different types of customers based on research analyzing how 500 volunteers interacted with brands across different channels in several sectors. Six main customer segments were identified: Lifestyle Junky, Astute Alpha, Internet Investigator, Dedicated Fan, Social Shopper, and Detached Introvert. Each segment is described in terms of demographics, communication preferences, and implications for how marketers can best engage with them.
A clash is emerging in British marketing between creative and data-focused professionals. The author argues both can learn from each other to become more well-rounded marketers. Creatives can learn that numbers can be beautiful, assumptions are often wrong, and outcomes are unpredictable. Data professionals can learn that marketing is about people, imagery is important for communication, and engaging both the head and heart is needed. The author advocates for finding "data artists" who combine imagination, creativity and empathy with technical data skills.
This document discusses data mining techniques for customer relationship management (CRM). It defines data mining as the extraction of implicit and novel knowledge from large datasets. The document outlines common data mining applications in retail, banking, telecommunications and other industries. It also discusses how data mining can be used across different stages of the customer lifecycle in CRM, such as up-selling, cross-selling and customer retention. Finally, it provides an overview of common predictive and descriptive data mining techniques like decision trees, rule induction, clustering and association rule mining.
This document discusses how The Data People uses data analytics to help businesses better understand their customers. They identify a company's most valuable customers through analyzing data on profitability, demographics, lifestyle, and usage. The Data People then builds customer profiles and segments customers to tailor marketing strategies. Case studies show how they helped Alliance & Leicester increase website visits by over 100% and Holmes Place reduce customer acquisition costs and increase retention rates through predictive modeling.
Consumer analytics is the process businesses adopt to capture and analyze customer data to make better business decisions via predictive analytics. It is a method of turning data into deep insights to predict customer behavior. It may also be regarded as the process by which data can be turned into predictive insights to develop new products, new ways to package existing products, acquire new customers, retain old customers, and enhance customer loyalty. It helps businesses break big problems into manageable answers. This paper is a primer on consumer analytics. Matthew N. O. Sadiku | Sunday S. Adekunte | Sarhan M. Musa "Consumer Analytics: A Primer" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-4 | Issue-6 , October 2020, URL: https://www.ijtsrd.com/papers/ijtsrd33511.pdf Paper Url: https://www.ijtsrd.com/other-scientific-research-area/other/33511/consumer-analytics-a-primer/matthew-n-o-sadiku
The document discusses how analytics can be used to solve business problems in the retail banking industry. It describes how analytics can be applied to various areas of a bank's profit and loss statement, including acquiring new customers, reducing customer attrition, improving account activation rates, and maximizing revenue from interest, fees, and cross-selling. It also discusses how strategic reporting, marketing analytics, and data-driven insights can be used for segmentation, customer lifetime value analysis, profitability and loyalty analysis, cross-selling strategies, and customer retention programs. The overall aim is to provide a top-down analytical approach to optimize all areas of a bank's operations and financial performance.
This document contains confidential information belonging to AAUM. It discusses various analytical techniques such as customer segmentation, market basket analysis, forecasting, and supply chain optimization that can be applied across industries. Case studies of companies like Tesco, Nieman Marcus, and Food Lion demonstrate how these techniques have been successfully used to increase sales, optimize operations, and improve customer experience.
Addressing the challenges and issues with businesses struggling to deliver successful Data Science in environments - Measure Camp, Bucharest (2 Nov. 2019)
Arun Gupta, Customer Care Associate and Group Chief Technology Officer, Shoppers Stop presented at the Premier Business Leadership Series 2010, http://www.sas.com/theserieshk.
With many retailers worldwide struggling to maintain revenues, how do you grow in such a tough competitive landscape? As a leading Indian retailer and pioneer in using technology, especially business analytics, Shoppers Stop is not only thriving but has helped revolutionise the retail sector. Gupta will share insights on using analytics to drive business value, reduce operational costs and provide better products and customer experience.
This document discusses analytics and retail analytics. It defines analytics as discovering patterns in data through statistics, programming, and research. Retail analytics specifically aims to improve customer loyalty and sales. It does this by identifying valuable customers, understanding their preferences, and creating personalized shopping experiences through offers targeted to individual needs. Retailers can gather customer data through in-store and online analytics to gain insights that optimize performance.
Today we are beyond the point where big data is in the prototype stage. We are entering an era where automation, integration and end-to-end solutions need to be built rapidly to facilitate disruption. Companies need to architect a
platform for Big Data (and traditional data) analytics
Case Studies - Customer & Marketing Analytics for Retail Gurmit Combo
The document discusses three case studies involving customer intelligence and marketing effectiveness services:
1. A luxury retailer case study where customer segmentation and profiling identified their most valuable customers to focus relationship management efforts.
2. A technology company case study where product association analysis and scoring identified accounts likely to purchase docking stations for targeted cross-selling.
3. A CPG company case study where regression modeling decomposed the impact of price, promotion, competition and cross-category effects on sales volumes, revealing promotion strategy optimizations.
Today there is a lot of buzz around customer experience. Many companies have realized that investments in customer experience improvement is important not just because it helps to boost the bottom lines of their businesses but because it takes at least 4 to 6 times more cost to acquire a new customer than to retain an existing customer.
Creating Business Value - Use Cases in CPG/RetailBig Data Pulse
This document discusses how big data analytics can help consumer packaged goods, fast moving consumer goods, retail, and e-commerce companies. It provides examples of use cases like predictive demand forecasting, pricing optimization, and markdown optimization. One case study describes how a department store used a forecasting and optimization model to improve markdown strategies and increase margins by $90 million annually. In conclusion, analyzing large, diverse customer data in real-time can provide actionable insights to increase market share, revenue and profits.
Project for System Analysis and Design (IS-6410).
By performing customer segmentation following are the three objectives which can be achieved
with the implementation of this new analytics system:
1. We can track the difference between loyal customers vs visitors, perform heat map
analysis of their browsing patterns.
2. Understanding customer demographics and to focus on high profitable segments.
3. Finally empowering our Marketing department to make better strategic decisions in
terms of online Ads/campaigns.
The document discusses the benefits of centralizing customer data into a single database. It notes that while customer data provides important insights, it is often scattered across different departments and systems, making it difficult to develop a comprehensive view. Centralizing data allows companies to better understand customers, target them across channels, and improve marketing efficiency. However, the database must be regularly maintained and enhanced with things like address updates to ensure the data remains accurate and useful over time.
Business analytics is used by industries to maximize operations and is applied in many fields including marketing, sales, finance, and human resources. Companies study consumer behavior patterns through social media, spending habits, and lifestyles to segment markets and identify target audiences. Business analytics tools help marketing and sales teams optimize their strategies, perform competitor analysis, and assess sales performance.
business analytics and its importance, marketing analytics definition and its importance, how marketing analytics helps to run the organization in effective and efficient manner.
Moving Forward with Big Data: The Future of Retail AnalyticsBill Bishop
Out new report Moving Forward with Big Data: The Future of Retail Analytics goes deeper into new territory that's relevant to changes taking place across retailing.
It calls out significant progress in the past 9 months.
• The definition of big data has grown beyond technical, i.e. “what it is,” to include “what it does.”
• A lot more companies are executing big data projects (an increase from < 20% to now 65% of sample respondents).
• Most of the focus is on driving top line growth.
This document discusses how The Data People helps companies identify their best customers and maximize profits through data-driven strategies and analytics. They analyze customer data to build detailed profiles, identify valuable customer segments, predict churn, and develop targeted marketing strategies. Case studies show how they helped companies like Alliance & Leicester increase website visits by 100% through improved targeting, and helped Nescafe launch a successful direct marketing campaign by creating an accurate customer profile model.
The document summarizes key findings from a survey of 560 marketers in the UK. Some of the main findings include:
- 91% of marketers believe the digital revolution is still ongoing and will be disrupted by younger digital native consumers.
- Social media is the channel being used most heavily, while spending on live events is also increasing.
- Marketers feel least confident about using mobile and data analytics is seen as the most important skill.
- The rise of empowered consumers and content marketing are seen as the biggest disruptors and influences on future marketing.
This document profiles different types of customers based on research analyzing how 500 volunteers interacted with brands across different channels in several sectors. Six main customer segments were identified: Lifestyle Junky, Astute Alpha, Internet Investigator, Dedicated Fan, Social Shopper, and Detached Introvert. Each segment is described in terms of demographics, communication preferences, and implications for how marketers can best engage with them.
A clash is emerging in British marketing between creative and data-focused professionals. The author argues both can learn from each other to become more well-rounded marketers. Creatives can learn that numbers can be beautiful, assumptions are often wrong, and outcomes are unpredictable. Data professionals can learn that marketing is about people, imagery is important for communication, and engaging both the head and heart is needed. The author advocates for finding "data artists" who combine imagination, creativity and empathy with technical data skills.
The document discusses how big data is composed of many small pieces of individual data, or "small data". Small data provides specific insights about individuals, like keywords from web pages they visit. When small data signals align with a customer's preferences and a product opportunity, marketers should act quickly before the situation changes. To make the most of small data, marketers will need to change their thinking from mass messaging to a model that is determined by and responds to individual customers in real time based on their changing contexts and situations.
Moderator & speaker bios posting travel times on dynamic message signs webinarraymurphy9533
This document provides information about a webinar on posting travel times on dynamic message signs and third party data, including moderator and speaker bios. The moderator, Bob Koeberlein from the Idaho Transportation Department, has engineering degrees and experience managing transportation projects. Three guest speakers will discuss their experiences: Jennifer Portanova from North Carolina DOT, Jeff Galas from Illinois DOT, and Richard Dye from Maryland SHA.
10 actions for facilities managers to improve job satisfactionMartin Leitch
Qualifications cost money and experience takes time, but these 10 low cost and timely actions will help facilities managers improve their job satisfaction
This document provides information about Ram Naresh Financial Consultancy Private Limited, including its corporate office address and contact details. It lists the directors and their details such as name, address, designation, and date of appointment. It also provides contact information for Ram Financial Network's customer support.
Financial Expert Witness Issues: How to Handle the Dangerous Financial Expert...DecosimoCPAs
This document discusses how to identify and handle dangerous financial expert witnesses. It identifies four categories of dangerous experts: 1) those who are intellectually dishonest, 2) those whose work is substandard, 3) those who do not follow accepted standards and methodologies, and 4) inexperienced and unqualified experts. It provides tips for spotting potential issues, such as an expert being too eager to provide a certain opinion or having never testified for the opposing side. The document advises vetting experts to ensure they follow proven methods, consider alternative scenarios, and use reliable underlying data. It stresses having another expert review the work for reasonableness.
FASB Proposals Affecting Government ContractorsDecosimoCPAs
The document summarizes key proposals from the FASB and IASB exposure drafts on revenue recognition. It discusses the core principle of recognizing revenue as control of goods or services is transferred to customers. It also outlines the five steps to apply the new standard: 1) identify contracts, 2) identify separate performance obligations, 3) determine transaction price, 4) allocate price to obligations, and 5) recognize revenue when obligations are satisfied. Government contractors will need to evaluate how these changes may affect their accounting and revenue recognition.
The document discusses various ideas for improving common experiences like shopping, eating ice cream, watching movies, and using stairs. It suggests motivating stair use by making it an unintended exercise, allowing custom ice cream creations, enabling in-theater food ordering on screens, adding recycling bins, and enhancing movie chairs with effects, massages, or mobility for all people. The key theme is using technology and design thinking to update experiences in a way that makes routines more enjoyable and inclusive.
This document discusses Children's Hospital's use of social media and the related HIPAA privacy issues. It provides an overview of Children's presence and strategies on various social media platforms like blogs, YouTube, Twitter, and Facebook. It outlines the goals of engaging on social media to improve communication and branding. The document also addresses concerns around HIPAA violations, negative comments, and losing control of the message. It proposes having external disclaimers and internal social media policies to help navigate these challenges and properly manage privacy and legal risks in the healthcare social media space.
Tax Principal Kim Lawrence of Decosimo presented this PowerPoint at the InfoSystems Technology Convergence Showcase on Aug. 28 at The Sheraton Read House in Chattanooga, Tenn.
Cost of the Future Newly Insured Under the Affordable Care Act (ACA) DecosimoCPAs
The document discusses the impact of the Affordable Care Act (ACA) on the newly insured population and healthcare costs. It finds that the percentage of uninsured individuals is projected to decrease significantly, with many transitioning to Medicaid, individual exchanges, or employer-provided insurance. The newly insured population is expected to have higher relative morbidity and healthcare costs compared to the currently insured, especially among older age groups, which will increase premiums. Hospitals and providers will play a key role in enrolling the uninsured by educating patients on coverage options and acting as a bridge to insurance exchanges.
The document discusses business intelligence and data warehousing in the banking sector. It defines data warehousing as a collection of integrated and non-volatile data used to support management decision making. It describes the benefits of data warehousing and business intelligence for banks, such as improved risk management, operational efficiencies, customer segmentation, and decision making. Business intelligence helps banks retain profitable customers, improve operations, and gain actionable insights into portfolio performance.
The company provides advanced analytics and data-driven decision making services. It has deep analytical capabilities across various industries, developed custom products, and has an expert team of data scientists, analysts, architects and programmers. The vision is to be a world leader in advanced analytics and enabling technology. Services include marketing, operations, supply chain and risk analytics. The company uses big data technologies like Hadoop and advanced tools to deliver solutions focused on customers across industries.
Customer analytics. Turn big data into big valueJosep Arroyo
BIRT Analytics is a customer analytics solution that allows companies to gain valuable insights from big data. It integrates data from multiple sources, analyzes large volumes of data, and provides clear and granular customer information. Tools allow users to explore data, identify patterns, profile customers, and forecast trends. Advanced analytics help optimize marketing, identify cross-sell opportunities, and understand customer behavior. The solution aims to help companies understand customer needs and adapt strategies based on real customer data.
The retailer wanted to create a unified customer data platform to provide complete visibility across their customer's omnichannel touchpoints and move from siloed data to a 360-degree view. Tredence helped build a CDP that integrated over 70 data sources, processed 250TB of data weekly, and increased addressable customer data visibility by 14%. This allowed the retailer to put the customer at the center of decisions, optimize their $3B marketing budget, and win a larger share of partners' advertising dollars in a cookie-less world.
This document discusses how analytics can be used to drive customer lifecycle management. It makes three key points:
1) Current analytical approaches used by most firms focus too much on driving new customer acquisition through the traditional marketing funnel, rather than managing the entire customer lifecycle. This leads firms to prioritize volume growth over long-term profitability.
2) To effectively use analytics across the customer lifecycle, firms must align their lifecycle perspectives and programs with the customer's decision-making process, determine the appropriate breadth and depth of analytical techniques, and use customer value and profitability as a common goal.
3) The document outlines how different analytical techniques such as segmentation, propensity modeling, and cross-
191 Castro Street, 2nd Floor, Mountain View, CA 94041 P 6.docxfelicidaddinwoodie
191 Castro Street, 2nd Floor, Mountain View, CA 94041 | P: 650-532-8155 | E: [email protected] | 1
CUSTOMER INTELLIGENCE:
THE KEY TO KEEPING
SAAS/CLOUD CUSTOMERS
There is good news and bad news for software
companies in the shift to the software subscription
model of the Cloud. The good news is that
revenues have become much more predictable and
stable. The bad news is that you have to keep
reselling the sale in order to retain those customer
income streams over time. The good news is that
there is more available data than ever before about
your customers. The bad news is that the data is
scattered all over the company and is therefore not
easily accessible.
The good news is that adding and supporting
application features and functionality is easier to do
in the Cloud. The bad news is that your
competitors will soon be adding those same
features to their applications too. The meaning is
clear. In the SaaS/Cloud business model, what is
really being sold is a relationship rather than
technological features & functions, and keeping that
relationship profitably going for as long as possible is
the core issue for long-term success as SaaS
company.
1
W
hi
te
pa
pe
r
Why should a SaaS company, especially if they think that they’re in their “land-grab” phase and therefore don’t have time
or resources to worry about churn at this point, invest time and money now in building dedicated customer retention
resources?
SaaS-Capital, a provider of debt-based growth capital for SaaS companies, answers the question. Churn is a cumulative
beast. The income that you lost last quarter continues to be lost next year and the year after. Consider their model of
two SaaS companies. Both sell only software subscriptions; no other income conduit is included. Both sign 10 new
customers per month @ $1,000.00 each. Both spend $120K per month on sales & marketing to acquire those
relationships (CAC). The only difference between them is that one has a customer retention rate of 95%; the other’s
only 80%. At the end of 5 years, the difference in bottom-line company valuation between the two was $15 million
dollars. Along the way, the company with the 95% retention rate also had increased revenues to work with, up to $24K
per month. That’s a lot of money – your money – both now and later.
THE HUGE COST OF CHURN
THE NEED TO KNOW
To make the initial sale, you needed to know quite a
bit about your prospective customer. What were
their business needs and requirements? Who were
the decision makers and influencers? What were
the timetable and the budget factors? All of that
knowledge and more made the signature on the
first contract possible. To get the renewal
signatures, however, you’ll have to keep that data up
to date and to add to it. Customer Intelligence is a
process that can’t have an end. It’s what you don’t
know about your customer relationships that can
cause you to lose them.
If the key ...
Operationalizing Customer Analytics with Azure and Power BICCG
Many organizations fail to realize the value of data science teams because they are not effectively translating the analytic findings produced by these teams into quantifiable business results. This webinar demonstrates how to visualize analytic models like churn and turn their output into action. Senior Business Solution Architect, Mike Druta, presents methods for operationalizing analytic models produced by data science teams into a repeatable process that can be automated and applied continuously using Azure.
1. The document discusses Business Intelligence and analytics using Oracle BI Foundation Suite. It provides an overview of the different components, capabilities, and features of Oracle BI including the BI Server, presentation layer, data warehousing, ETL processes, and end users.
2. It describes the different modules of Oracle BI including dashboards, KPIs, reports, predictive analysis, and graphical OLAP. It also discusses the hardware and software components needed for a complete Oracle BI solution.
3. Screenshots are provided showing how to create a database connection in Oracle BI, indicating how users can access and work with data through the presentation layer.
Smarter analytics for retailers Delivering insight to enable business successKun Le
GS Retail uses smarter analytics to gain insights into customer preferences and behaviors. This allows them to personalize promotions and optimize inventory. Intersport also uses analytics to future-proof their business and gain a competitive advantage through deeper customer understanding. Migros similarly uses analytics to deepen their insights and better respond to customers, achieving business success. Smarter analytics provides retailers with performance and organizational benefits by improving decision making and enabling more targeted strategies.
Big data analytics for telecom operators final use cases 0712-2014_prof_m erdasProf Dr Mehmed ERDAS
Big Data Analytics for TELCOs Customer Experience Management Permission Based Marketing for Location and Movement Data Data Modelling Business Use Cases Data Mining BSS OSS COTS OTT Churm Modeling Markov Processes HANA HADOOP INtegration Video Streaming Test cases
Marketing analytics is the study of consumer data to evaluate marketing performance and optimize campaigns. It involves collecting, cleaning, and analyzing consumer data using statistical techniques to understand consumer behavior, refine marketing strategies, and predict future trends. Marketing analytics helps target consumers based on their interests and serve them the right messages at the right time through the right channels. It evaluates past marketing performance, reports on previous campaigns, and predicts future trends to improve marketing plans.
The document discusses predictive analytics and its applications. It begins by defining predictive analytics as using data patterns to predict future outcomes. It then discusses how various industries like marketing, risk management, and operations are using predictive analytics for applications such as targeting customers, assessing risk, and optimizing processes. The document provides examples of how predictive models are used for response modeling, customer segmentation, loyalty/retention, and assessing customer profitability in marketing. It also discusses using predictive models for predicting defaults in risk applications.
Future of Tracking: Transforming how we do it not what we doKantar
The slides from ‘Digital Transformation of Tracking’ webinar presented on BrightTalk on 28th February 2017. In this webinar Mark Chamberlain and Alex Taylor discuss how changes in consumer behaviour, increased business pressures and new technologies have created both opportunity and disruption across all industries. Like every other industry, research is in the midst of its own transformation affecting not what we do but how we do things.
Customer insight presentation s houston - boston march 2014Stuart Houston
The document discusses unlocking customer insights to drive business growth. It notes that only 28% of companies provide a good or excellent customer experience. It then provides examples of using customer data and analytics to increase revenue and profitability, expand product offerings, and enhance the customer experience. Specific use cases discussed include measuring risk-adjusted performance, aligning with customer life stages, implementing propensity modeling, and ensuring accurate corporate client profiles. The document concludes by discussing considerations for delivering effective, timely customer insights across the enterprise.
The document discusses customer relationship management (CRM) strategies and the use of data in CRM. It describes the C-MAT model for customer management, which involves understanding customer value, behavior and attitudes. It also discusses integrating customer data into CRM strategies using tools like data warehousing and data mining to collect and analyze large amounts of customer data. The document provides examples of how companies can use data mining techniques like correlation, segmentation and propensity analysis to gain insights into customers.
Database marketing uses customer databases to send personalized communications to promote products. It emphasizes using statistical models of customer behavior to select targets. Sources of customer data include internal sales records, purchased lists, and information collected from websites and forms. The main techniques are appending additional customer data, tracking website visits, and email communications. Key drivers of database marketing are the need for relationship-building, lower processing costs, and measuring marketing impact. Challenges include managing large data volumes and understanding individual customer interests from their online behavior.
2. Increase Customer Profitability Using Data Mining and Advanced Analytics
Table of Contents
Executive Summary..........................................................................................1
Three Key Trends Affecting ABM.....................................................................2
The Suppliers’ Shift from Product-Centric to Customer-Centric...................2
The Availability of Detailed Data...................................................................3
Declining Emphasis on Process and Productivity Improvement as a Way
to Improve the Bottom Line........................................................................3
Whither ABM?...............................................................................................4
What? So What? Then What?...........................................................................4
From Seeing Costs to Understanding Them.................................................5
What Differentiates More-Profitable From Less-Profitable Customers?......5
How Can a Supplier Determine Differentiating Drivers
of Its Profits from Customers? ..................................................................7
The Explanatory Investigation Continues … and Continues ........................7
Let’s Try a Different Approach ......................................................................9
Applying the Computing Power of Data Mining
and Advanced Analytics ...........................................................................10
Where Does the Analyst Go from Here? .....................................................11
And There’s More …...................................................................................12
Conclusion......................................................................................................14
APPENDIX 1.....................................................................................................15
The Importance of Ratios............................................................................15
APPENDIX 2.....................................................................................................17
Risk Incidents: Accounting for Occasional Events
in Profitability Analyses............................................................................17
APPENDIX 3.....................................................................................................18
Risk Incidents: Better Understanding the Impact of Process Failures........18
APPENDIX 4.....................................................................................................19
Profitable Customer Acquisition.................................................................19
APPENDIX 5.....................................................................................................20
Customer Equity Analysis...........................................................................20
APPENDIX 6.....................................................................................................21
Customer Strategy......................................................................................21
About SAS.......................................................................................................23
3. Increase Customer Profitability Using Data Mining and Advanced Analytics
ii
This white paper was written by Gary Cokins and Charles Randall.
Gary Cokins is an internationally recognized expert, speaker and author on
the subject of advanced cost management and performance management
systems. He is a Principal in Global Business Advisory Services with SAS, a
leading provider of business intelligence and analytic software headquartered
in Cary, NC. Cokins received a BS in industrial engineering/operations research
from Cornell University and an MBA from Northwestern University’s Kellogg
School of Management. Cokins began his career at FMC Corporation, and
he also served as a management consultant with Deloitte, KPMG Peat
Marwick and Electronic Data Systems (EDS). His latest book is Performance
Management: Integrating Strategy Execution, Methodologies, Risk, and
Analytics. He can be reached at gary.cokins@sas.com.
Charles Randall has built his career in strategy and marketing analytics.
Randall’s career encompassed both telecommunications and management
consultancy before joining SAS as a Principal Business Consultant. In his current
role as Solutions Marketing Manager, he draws from 15 years of experience
in these fields to share deep expertise and insight in numerous articles and
papers. Randall received a BSc in economics and a PhD in econometrics from
the University of Wales. His latest research study is titled Pleased to Meet You:
How Different Customers Prefer Very Different Channels. The study is a joint
project with Professor Hugh Wilson of Cranfield School of Management. Randall
can be reached at charles.randall@suk.sas.com.
4. Increase Customer Profitability Using Data Mining and Advanced Analytics
Executive Summary
There is a trend for customers to increasingly view suppliers’ products and
standard service lines as commodities. As a result, what customers now seek
from suppliers are special services, ideas, innovation and thought leadership.
Many suppliers have actively shifted their sales and marketing functions from
product-centric to customer-centric, through the use of data mining and business
intelligence1
tools to understand their customers’ behavior – their preferences,
purchasing habits and customer affinity groups. In some companies the
accounting function has supported this shift by reporting customer profitability
information (including product gross profit margins) using activity-based costing
(ABC) principles. However, is this enough?
It is progressive for the accounting function to provide marketing and sales with
reliable and accurate visibility of which customers are more and less profitable.
Often, sales and marketing people are surprised to discover that due to special
services, their largest customers in sales are not their most profitable ones, and
that a larger subset of customers than believed are only marginally profitable –
or worse yet, unprofitable. But a ranking of profit from each customer does not
provide all the information as to why. That is where data mining and analytical
techniques can help.
The use of ABC data leads to activity-based management (ABM). There are
some low-hanging fruit insights from ABC data. For example, one can see
relative magnitudes of activity costs consumed among customers. There is also
visibility into the quantity of activity drivers – such as the number of deliveries –
that cause activity costs to be high or low. But this does not provide sufficient
insight to differentiate relatively highly profitable customers from lower-profit or
unprofitable customers.
One can speculate what the differentiating characteristics or traits might be, such
as sales magnitude or location; but hypothesizing (although an important analytics
practice) can be time-consuming. It is like finding a diamond in a coal mine. One
cannot flog the data until it confesses. In attempting to identify the differentiating
traits between more and less profitable customers, the major traits may not be
intuitively obvious to an analyst. A more progressive technique is to use data
mining and advanced statistical analytics techniques.
This paper describes, in particular, the use of segmentation analysis based on
decision trees and recursive partitioning. These techniques can give the sales and
marketing functions insights into what actions, deals, services, unbundled pricing
and other decisions can elicit profit lift from customers.
1 Data mining is the process of extracting patterns from large amounts of stored data by
combining methods from statistics and database management systems. It is seen as
an increasingly important tool for transforming unprecedented quantities of digital data
into meaningful information (nicknamed “business intelligence”), to give organizations an
informational advantage. It is used in a wide range of profiling practices, such as marketing,
surveillance, fraud detection and scientific discovery.
1
■ Business users of activity-based
costing information gain valuable
insights as to which products,
service lines, channels and
customers are relatively more
or less profitable. They also see
why – by observing the visibility
and transparency of the internal
process and activity costs that
yield each customer’s contribution
profit margin layers. But the pricing,
marketing and sales functions often
struggle with determining which
actions to take to create increasing
profit lift for customers. This paper
describes analytical techniques
that can identify which drivers best
explain the differences between
high-profit and low-profit (or
negative-profit) customers. Knowing
these drivers can lead to the most
profit-lifting actions.
5. Increase Customer Profitability Using Data Mining and Advanced Analytics
The goal is to accelerate the identification of the differentiating drivers so that actions –
or interventions – can be made to achieve that high-payback profit lift from customers.
Analysts using ABM have benefited from applying online analytical processing (OLAP)
multidimensional cubes to slice and dice data. Even greater benefits and better
decisions can come from applying data mining and advanced analytics.
Three Key Trends Affecting ABM
Activity-based management involves calculating how expenses (e.g., salaries or
supplies) are converted into the costs of work activities that in turn are traced into the
costs of outputs such as products, services, channels and customers. The calculation
method is activity-based costing, and it is done with modeling. ABM then exploits the
ABC information for insights, analysis and decisions.
The three current trends affecting ABM are:
• The shift in attention from product-centric to customer-centric costs.
• The explosion of available data.
• Diminishing returns from process and productivity improvements
The Suppliers’ Shift from Product-Centric to Customer-Centric
Before diving deep into the role that data mining and analytics can play when
combined with managerial accounting, let’s first get some context to help us better
appreciate the problem suppliers face in increasing profitability from various customers.
A primary reason that companies are increasingly calculating and measuring customer
profitability is because of a shift in the sales, marketing and operations functions from
being product-centric to being customer-centric. This shift results from customers
increasingly viewing all suppliers’ products and standard service lines as commodities
(i.e., having little differentiation). In response to this trend, suppliers are shifting their
attention toward differentiating services for different types of customers. That is, rather
than mass selling giving the sales force incentives to “push” products, suppliers are
working backward by starting with their customers and tailoring unique offers and
deals based on the distinctive preferences and tastes of customer microsegments (and
even individual consumers, at the extreme).
But what deal, discount, special service, etc. should potentially be offered to which
type of customer in order to get the maximum profit lift?
2
6. Increase Customer Profitability Using Data Mining and Advanced Analytics
Answering that question is a challenge. Customers should be viewed as investments
rather than as something a supplier spends money to serve. With this “customers
as investments in a portfolio” view, the challenge becomes determining which deals,
offers, special services, etc. will maximize the return on investment (ROI) for each
customer microsegment (and potentially for each individual customer). That is, how do
we determine which actions will yield the largest financial profit lift – and from which
individual customers?
The Availability of Detailed Data
The progression toward transactional ABC models has been fostered by the
availability of systematic processes, technologies and customer data now that
most major organizations have introduced enterprise resource planning (ERP)
and customer relationship management (CRM) systems. This has meant that it
is more practical to define work activities at a more detailed level, and provide
direct cost driver data to support translating the activity costs into outputs.
This has inevitably led to an increase in the number and sophistication of work
activities and activity cost drivers in the model, presenting even more candidates
to investigate to understand what is and is not important.
Historically it was very difficult to build models of a scale that could produce
individual customer profitability models; so models tended to stop at a segment
level (e.g., all customers from a given standard industry code, geographic area
or other arbitrary category). We have tended to rely on the traditional rather
than arbitrary groupings used within a business, and this potentially disguises
important information on trends that cross customer segment boundaries.
Today’s software computing power, particularly transactional costing for
individual customers, removes that restriction.
However, when more product variations than ever before are factored in,
including more distribution channels, the complexity of costing models is beyond
the level at which basic reporting or even OLAP can be applied to find the most
important insights.
Declining Emphasis on Process and Productivity Improvement as a
Way to Improve the Bottom Line
In the early days, activity-based cost management (ABC/M) was very much focused
on process improvement, and could be seen as part of the whole BPR/Six Sigma/
TQM movement. After 25 years of these improvement initiatives, it is probably fair to
assume that most companies have reasonably efficient processes. While there may
still be productivity gains to be made in this area, they are unlikely to be substantial.
3
7. Increase Customer Profitability Using Data Mining and Advanced Analytics
In other words, ABC/M literature has largely focused on the internal efficiency of
business processes as a whole. It has yet to really address how processes relate to
individual customers, and how their varying applications affect profitability. Yet this is
where we are now more likely to find the opportunity for dramatic gains in profitability
of the firm.
Whither ABM?
With these points in mind, a strong business case can be made that the major
benefit from applying the principles of activity-based costing is not just from product
profitability reporting but also from the more encompassing customer profitability
reporting. The latter profitability reporting is inclusive of product and standard service-
line costs, and it also includes the “below the gross profit margin line expenses,” such
as distribution, channel, customer service, selling and marketing-related expenses.
These nonproduct and nonstandard service-line expenses are commonly called
costs-to-serve. ABC (combined with direct costing) solves the problem of not reliably
knowing which products or service lines make or lose profits, or which customers are
more or less profitable – and by how much. ABC also measures the cost elements for
each customer that yield the level of profit.
But as with many other fields, solving one problem creates a new problem. In the case
of ABC, the new problem for a company is to understand what actions to take to
improve profit generation from customers.
What? So What? Then What?
The three trends affecting ABM reveal moving beyond just knowing what outputs
cost to understanding the relevance of what causes those costs (so what?) – and
then investigating, testing and validating what the financial consequence (then what?)
will be from decisions based on insights gleaned from the ABC information. This is
also a good reason for the ABC reporting to be a permanent, repeatable and reliable
production reporting system. This is in contrast to its use as only a one-time study
or project to learn an answer and be done. Effective ABM creates benefits through
frequent short-interval refreshing of the ABC data to monitor progress and see
emerging insights for further investigations.
4
8. Increase Customer Profitability Using Data Mining and Advanced Analytics
From Seeing Costs to Understanding Them
Companies that have successfully implemented ABC and can successfully report
customer profitability as a permanent and repeatable production system deserve
to congratulate themselves and celebrate. They have provided better visibility,
transparency and accuracy for reporting profit margin contribution layers of their
customers. With this information, the pricing, sales and marketing functions can see
things they previously could only speculate or guess about. And much of what they
might see may not be pretty or may come as a surprise. For example, they may
realize that their highest-sales-volume customer may not be a very profitable customer
due to the substantial extra services that customer requires, and associated high-
maintenance behavior. Under certain conditions, some customers may be outright
unprofitable. But the celebration of this robust reporting should be temporary. There is
much more to do to increase the customers’ profitability to the company.
With customer profitability reporting, companies can gain insights of all kinds. But there
is eventually a limit. As mentioned before, in the grand scheme of decision making,
good ABC information reporting only answers the first of three critical questions:
“What?” That is, what do things cost? What products, service lines, channels and
customers are more or less profitable? But that is only reporting. More is needed to
increase profits.
Analysis and decision making requires answers to two more questions: “So what?”
and “Then what?” The “so what?” question begs to know what about the profit margin
information is relevant and could be acted upon. The “then what?” question begs to
know – to validate – if an action is taken, what will be the likely financial effect?
What Differentiates More-Profitable From Less-Profitable
Customers?
Figure 1 displays a popular profit contribution-ranked deciles histogram that groups
customers by measuring and viewing them. The source of the data is the profit
generated by ABC for each customer.
5
9. Increase Customer Profitability Using Data Mining and Advanced Analytics
Figure 1: Customer profit contribution deciles.
Profitability reports like that in Figure 1 are often shocking and disturbing to
executives and managers when they are seen for the first time. This is because the
reports reveal their misconceptions – that there are substantially higher financial
profit and greater losses in certain customers than they suspected. (ABC reporting
overcomes these misconceptions by replacing accuracy-suppressing cost
allocations that use broadly averaged overhead expense allocation factors with
cause-and-effect cost-driver tracing assignments.)
To answer the “so what?” question related to determining how to increase a customers’
profitability, a supplier could look at its customer profit contribution-ranked histogram
decile diagram (as in Figure 1) and ask this question: “Excluding the obvious profit effect
from sales volume, what one characteristic, trait, behavior or transaction of a customer
differentiates highly profitable customers from the rest?” That is, what is the most
prominent and explanatory driver among all those that are possible?
There are challenges to answering this question. How should the analysts determine
what and where to investigate? Is it with guesswork, luck, speculation?
This is where data mining, statistics and analytics play a role: to reveal what dominant
and secondary drivers explain the differentiation between high- and low-profit customers.
What most drives profitability across an organization? If this were known, could pricing,
marketing and sales actions be more focused, and yield greater certainty?
6
10. Increase Customer Profitability Using Data Mining and Advanced Analytics
7
How Can a Supplier Determine Differentiating Drivers
of Its Profits from Customers?
Let’s start simple. Imagine the supplier’s business analysts speculate that the residential
location of a customer may be a major driver explaining the differentiation between
high- and low-profit customers – the first and last profit contribution decile in Figure 1’s
histogram. Since the analysts have access to both of these data items (i.e., profit and
home address), a correlation2
(i.e., the explanatory value level) can be measured.
With a very simple examination of just the most and least profitable (10 percent)
customer histogram deciles, the correlation measure may confirm the analysts’
hypotheses that the most profitable customers live in affluent neighborhoods and the
unprofitable customers reside in low-income neighborhoods. There is, however, a
remaining question – how strongly do these newfound facts support the conclusion?
If the correlation is extremely high, then potential “so what?” actions – like knowing
where to advertise and where not to – become obvious. But let’s imagine that in this
case the correlation measure is relatively low – meaning that residential location does
not strongly support the analysts’ hypothesis.
What next? Which other driver might explain the customer profit differentiation?
The Explanatory Investigation Continues … and Continues
Imagine the supplier’s analysts next speculate that it is the customer’s age, not their
residential location, that may be a major explanatory driver differentiating high-profit
from low-profit customers. Again, both data records for all customers are accessible
(i.e., profit, age). The correlation is again measured. A possible outcome might reveal
that older customers (e.g., senior citizens) are much more profitable, and younger
customers (e.g., teenagers) are much less profitable.
However, the outcome could have been the reverse, with young people (e.g.,
spendthrifts) being most profitable and older people (e.g., frugal) not. But similar to
the residential location hypothesis, let’s imagine that the strength of the correlation
measure is again low – meaning there is not clear evidence that age is a differentiating
driver.
How about the product mixes that customers purchase? Figure 2 displays what
the analyst could see. However, imagine again that the correlation score does not
demonstrate sufficient evidence that this is a differentiating driver.
2 In statistics, dependence refers to any statistical relationship between two random variables or two
sets of data. Correlation refers to any of a broad class of statistical relationships involving dependence.
Familiar examples of dependent phenomena include the correlation between the physical statures
of parents and their offspring, and the correlation between the demand for a product and its price.
Correlations are useful because they can indicate a predictive relationship that can be exploited
in practice. For example, an electrical utility may produce less power on a mild day based on the
correlation between electricity demand and weather. In this example there is a causal relationship,
because extreme weather causes people to use more electricity for heating or cooling; however,
statistical dependence is not sufficient to demonstrate the presence of such a causal relationship.
11. 8
Increase Customer Profitability Using Data Mining and Advanced Analytics
Figure 2: Product mix deciles.
How about the region of the country the customer lives in rather than the type of
neighborhood within a metropolitan area, as the analyst first speculated? Figure 3
displays this view. But again, let’s imagine that this driver does not provide clear or
sufficient evidence.
Figure 3: Region decile
12. Increase Customer Profitability Using Data Mining and Advanced Analytics
9
Where do the analysts go next? What other driver or trait could they test? That is,
what other customer driver or trait could the supplier’s analysts consider as the high
versus low customer-profit-level differentiator? Customer weight? Hair color? Type of
credit card? Number of brothers and sisters? Sibling age rank (e.g., oldest, youngest)?
Model year of their car? Car manufacturer and model? Which traits can you think of?
The point here is that the possibilities appear to be unlimited, especially if you have
a big imagination. Does the pursuit need to continue to be somewhat trial-and-error
as in the examples above? Possibly – however, experienced analysts do apply some
common sense in speculating which drivers to consider. But in a complex world, even
experienced analysts need some assistance to shorten their investigation time and
help them quickly focus on what matters most.
In reality, the number of single customer behaviors or traits that is “most explanatory”
is not limitless. It is restricted by the amount of data a supplier has about each of its
customers. But with the massive amount of customer information in storage, the list of
driver choices could be fairly extensive.
So, what driver should the supplier’s analysts test next? Selecting the first few traits
may be relatively easy – as with the residential location and age. However, as in our
example, assume that the correlation values are low. Then do you test other traits
that are less obvious and may be more challenging to hypothesize? What should the
analysts do to reduce the time and effort of this investigation? This research should not
be like looking for the single needle in a haystack, or the single diamond in a coal mine.
Let’s Try a Different Approach
At this point it is clear that customer profitability reporting is not the same thing as
customer profitability analysis. What is needed is an approach that will crystallize
insights gained from customer profitability reporting – and generate meaningful insight
into which characteristics and behaviors of customers and products separate the
relatively more- and less-profitable customers.
Analyzing large-scale customer-profitability models is the sort of challenge ideally
suited to SAS®
software’s advanced data mining and analytical capabilities. These
techniques allow a business analyst to increase the value of the model by:
• Simplifying complexity and identifying what is most important for the business to
focus on.
• Discovering hidden patterns that cross arbitrary customer segment boundaries.
• Allowing the business to predict how profitable a customer is likely to be now and
in the future.
Applying data mining and analytics to cost and profitability reporting will enable the
business analyst to answer the “so what?” question. Performance management
methodology modeling can solve the “then what?” question.
13. 10
Increase Customer Profitability Using Data Mining and Advanced Analytics
The next section shows how data mining solved that earlier problem of finding which
drivers were critical ones in our model.
Applying the Computing Power of Data Mining
and Advanced Analytics
Let’s discard the hypothetical supplier analysts’ quest and get more directly to the
point. By combining data mining and advanced analytics (in this case a statistical
technique called a decision tree) with today’s enormous computing power and
its access to massive amounts of stored data about customers, one can gain
tremendous insight. Decision trees are a simple but powerful form of multiple variable
analysis. Produced by algorithms that split data into branch-like partitions, decision
trees are developed and presented incrementally as a collection of one-cause, one-
effect relationships calculated in a recursive form. The appeal of decision trees lies in
their relative power, ease of use, robustness with a variety of data types, and ease with
which they can be understood by non-experts.
Figure 4 displays the initial “branching” of the most statistically significant explanatory
differentiating driver. For this particular supplier’s 22,161 customers’ profit rank ordered
for 2010, the correlation analysis calculated “average transaction quantity” as the most
explanatory driver.
Figure 4: Decision tree - the average transaction quantity.
14. Increase Customer Profitability Using Data Mining and Advanced Analytics
11
The figure displays other potentially useful information:
• It calculates that 5.3 is the average transaction quantity that divides the more-
and less-profitable customers into two subsets of the whole population
of 22,161 customers.
• It calculates that 6,551 customers are the “less profitable” (with their own average
transaction quantity of 1.08) – and that 15,610 customers are in the “more
profitable” subset (with their own average transaction quantity of 7.07).
• It calculates that 14.69 is the dividing amount, with customers above that number
being x and those below it being y.
OK. So what?
Where Does the Analyst Go from Here?
Based on the initial partition, the marketing and sales functions can begin to
brainstorm how to alter the behavior of customers in the “less profitable” segment
so that they move in the direction of customers in the “more profitable” segment. For
example, the supplier could provide customers with a menu of service-level prices to
encourage them to increase or decrease transaction quantities with offered price levels
derived from the ABC information by assuring that an incremental change in price (up
or down) will always exceed the incremental change in cost to deliver that service level.
This way, the supplier gains a higher profit by altering the behavior of the customer to
select a service-level offer in either direction.
But while that brainstorming is occurring, the analysts can delve deeper. After the
average transaction quantity is revealed as the most prominent factor, each “more/less
profitable” segment will be recursively partitioned. Following one branch of the decision
tree down, Figure 5 reveals that the factor that most differentiates the “more profitable”
customers is “% cash”; and subsequently, further down the tree, a third critical factor –
“days with a negative balance” – applies.
At this point, an uncomfortable fact is uncovered. Within the “high average transaction
quantity customers,” there exists a distinct microsegment who use a lot of cash and
frequently run overdrafts. Consequently, they are the least-profitable customers. Now
the marketing and sales functions can focus on this particular microsegment and
brainstorm ideas to change this customer segment’s behavior or their commercial
terms, and move them toward profitability. Figure 5 displays the expanding the
decision tree diagram.
15. 12
Increase Customer Profitability Using Data Mining and Advanced Analytics
Figure 5: Three-level tree diagram.
You get the idea. Why speculate when the computer can do the heavy lifting?
And There’s More …
Does this mean that the analysts’ work is done? By no means. This is just one
technique that can be applied to a model, to solve one particular question: What
are the typical behaviors that distinguish our most and least profitable customers?
Advanced data mining and analytical techniques give the business analyst both the
time and capability to gain ever more insight into their customers. The analyst plays a
critical role in this process, defining the business problem, understanding how it can be
answered (and therefore which analytical technique to use), and finally, how to structure
the analysis.3
This enables a business to tackle a range of other issues that include
using analytics to:
• Better understand the implication of nonrecurring events.
• Understand the nature of process failure.
• Predict which prospects are likely to be most profitable.
• Understand customer lifetime value.
• Develop customer strategy based on profitable behaviors.
3 Appendix 1: The Importance of Ratios
16. Increase Customer Profitability Using Data Mining and Advanced Analytics
13
Using analytics to better understand the implication
of occasional (intermittent) events4
When customer segments are disaggregated down to an individual level, one may
encounter occasional (i.e., intermittent) activities associated with a customer (what
we might call “risk” incidents) that are unlikely to reoccur on a regular basis, but still
have a dramatic impact on the potential profitability of that customer. An example
of this is a customer moving to a new home, which is a significant expense for an
energy company, but one that probably does not happen frequently for the majority of
customers. For a correct appreciation of the profitability of a customer, one needs to
not just understand what it costs to process this incident, but also the likely probability
of it occurring in any given period.
Using analytics to understand the nature of process failure5
Any ABM model makes the immediate costs of failure in internal processes extremely
visible. But in addition to the direct impact of the cost of recovery activities, there
may be secondary impacts that are less visible in an ABM model. These can still
be identified, and the implications can be quantified. These secondary impacts can
materialize in a number of ways over time, including: increased cancellation rates;
selection of more costly but less risky business channels; reduced reorder frequencies
and volumes; and elevated customer churn rates.
Predicting which prospects are likely to be most profitable6
Once an existing customer’s behavior is known, it becomes relatively easy to predict
whether that customer is likely to be profitable, even without a detailed profit model. Of
course, the potential behavior of prospective customers is an unknown – but it is not
necessarily unpredictable. An analysis linking customer profitability to geo-demographic
characteristics allows an analyst to identify prospects with certain determining
characteristics. We can theorize that these customers are likely to behave in a similar
fashion to similar existing customers, and become similarly profitable.
4 Appendix 2: Risk Incidents: Accounting for Occasional Events in Profitability Analyses
5 Appendix 3: Risk Incidents: Better Understanding the Impact of Process Failures
6 Appendix 4: Profitable Customer Acquisition
17. 14
Increase Customer Profitability Using Data Mining and Advanced Analytics
Using analytics to understand customer lifetime value7
One of the significant insights discovered early when customer profitability models
are data mined is that customer profitability does not tend to follow a nice, smooth
incremental path. Instead, it tends to cluster around customers at different life stages
and steps of their relationship with a business. To provide a real picture of the potential
long-term value of a customer – showing the customer’s longevity and their likely future
– analytics incorporates the possibility of significant life-stage changes.
Using analytics to develop customer strategy based on
profitable behaviors8
Where ABM has been applied to strategy, it has typically been focused on structural
issues, such as how to organize departments to achieve economies of scale, or
what markets to continue pursuing. But for a marketer thinking about customer
strategy, customer profitability and behavioral analysis should be foundation stones.
This calls on two aspects previously discussed – the characteristics of profitable
customers, and where a business can find more customers like them. However,
strategy will tend to focus less on the small and unusual customers, and more on large
groups of customers with similar characteristics.
Conclusion
All things considered, why speculate and guess at the rank-ordered drivers that
differentiate between relatively more- and less-profitable customers? Why not apply
computer power to do the heavy lifting? An additional message is to not perform the
analysis as a one-time study, but to produce the information at frequent time intervals
as a permanent, repeatable and reliable production information system.
It is true that experienced analysts typically suspect and hypothesize that two or more
things are related, or that some underlying behavior is driving behavior seen in the
data. They then search for confirmation and understanding of the relationships. In
other words, the application of analytics is usually more confirmatory than exploratory.
It is not like finding diamonds in a coal mine. One does not simply flog the data until
it confesses! However, in the case of attempting to identify the differentiating traits
between more- and less-profitable customers, the major traits may not be intuitively
obvious to an analyst.
The goal is to accelerate the identification of the differentiating drivers so that actions
– interventions – can be considered as a way to get high-payback profit lift from
customers. The analysts using ABM have benefited from applying online analytical
processing (OLAP) multidimension cubes to slice-and-dice data. Even greater benefits
and better decisions can come from applying data mining and advanced analytics.
7 Appendix 5: Customer Equity Analysis
8 Appendix 6: Customer Strategy
18. Increase Customer Profitability Using Data Mining and Advanced Analytics
15
APPENDIX 1
The Importance of Ratios
One of the critical first steps when analyzing an ABC model is to understand the
data, and how it needs to be transformed to correctly answer your questions. The
biggest issue we face is the impact of volume effects (the amount a customer buys)
on our analysis, and how they can override any other potentially more important
analyses. This is why one of the first actions often taken is to “normalize the data”
through the derivation of key ratios that remove the size effects and allow deeper
insights to be surfaced.
To illustrate this, we will use the case of a wholesale business that has some large and
some small customers, with varying return rates. We could consider four customers –
two large (A and B) and two small (C and D) – with two of the customers having high
return rates (A and C) and two of the customers having low return rates (B and D).
See the table for an illustration.
If we then plot this data on two charts, one showing net total contribution versus
number of returns, and the other showing percent of contribution versus percent of
return rates, we see a dramatically different picture. The volume effect overwhelms the
return rate effect in the first graph where we plot absolute values, giving a potentially
misleading message that the number of returns has a positive correlation to profit.
In the key ratios analysis, with the size effects removed, the rate of returns can be
correctly seen to have a negative impact on profitability.
19. 16
Increase Customer Profitability Using Data Mining and Advanced Analytics
Indeed, the beloved whale curve diagram (more properly called the profit margin cliff
curve) has a tendency to provide a relatively false picture of profitability. Because
it orders customers on the basis of absolute profit, it tends to group all the small
customers in the center and place large customers at each end. This over-emphasizes
these few significant but anomalous customers, and leads us to miss potentially critical
trends and patterns that can be found in the population at large, and which can have a
dramatic impact on profitability.
By sizing and ordering our customers based on revenue, and by showing profit versus
revenue, we can transform a relatively benign-looking whale curve into a much more
dramatic hook curve. As shown below, we can now see that we have a whole set of
customers who are in fact more profitable than our largest customer, even though they
generate the most absolute profit. And similarly, we have a host of customers who are
significantly less profitable than our customer who generates the greatest loss. It is also
quite clear that clearing out our unprofitable customers will not have a dramatic impact
on our top line, so we will feel free to attack them without worrying too much about the
impact on the share price.
20. Increase Customer Profitability Using Data Mining and Advanced Analytics
17
APPENDIX 2
Risk Incidents: Accounting for Occasional Events
in Profitability Analyses
One of the unfortunate side effects of the periodic nature of an ABC model is that it
captures occasional events against a customer in the period that they happen, then
registers the impact on profitability in that period, but provides us with little information
about whether that event is likely to recur frequently or infrequently.
Examples include things as varied as home moves for energy companies, insurance
claims, warranty claims, issuing of new credit cards or mobile handset renewals.
When designing our data exploration model, we need to adopt a different strategy for
these sorts of costs. Specifically, we need to replace the occasional event or behavior
with a marker indicating the probability of the event occurring. Typically this would
be a direct calculation of the probability of the event occurring for each customer in
the period. This may be calculated using a number of techniques, including logistic
regressions, neural networks or a decision tree.
21. 18
Increase Customer Profitability Using Data Mining and Advanced Analytics
Alternatively, we may find a proxy indicator for the risk of the occasional event. These
are typically geo-demographic indicators, but they could also include products (for
example, car model affecting the probability of a warranty claim), or channel (returns
are more likely for mail order than store-purchased products).
Exact choice of approach depends very much upon:
• The needs of the analysis.
• The availability of data.
• The tools available.
APPENDIX 3
Risk Incidents: Better Understanding the Impact of Process Failures
One class of occasional events often has to do with process failures. One of the huge
benefits of an ABM model is that it makes the true costs of such failures extremely
visible. But in addition to making the cost visible, analytics also gives us the ability to
better understand the impact of such failures, from the likelihood of them occurring to
the long-term impact on customer retention.
One example of this occurred at a classified advertising company, where a segment of
customers were identified who had many advertisement amendment costs, but also
unusually low revenue due to cancellations.
22. Increase Customer Profitability Using Data Mining and Advanced Analytics
19
Beyond the initial identification of this unusual set of behaviors, a technique called
survival analysis was applied to the problem. This clearly revealed that there was a
critical turning point in the customers’ relationship with the business at which point
they became frustrated enough to cancel their advertisement. (See the diagram
for more information.) With this insight, we knew when to set a warning marker on
transactions, allowing us to review them and determine if a recovery effort was worth
engaging in.
APPENDIX 4
Profitable Customer Acquisition
While much of this paper has been focused on how we identify which behaviors
customers exhibit that make them profitable or not, it is not possible to understand
how a prospect will behave if we acquire them, and therefore whether they are likely
to be worth the effort. Much work has gone on in sophisticated companies to review
expected sales revenues for different demographics. But these analyses can be
improved even further by moving from analyzing and segmenting based on expected
revenue, to applying the additional insight available from a customer profitability model.
This type of model not only has the potential to tell a company what level of profit they
may expect from a particular prospect segment, but it can also show:
• How that group is likely to behave.
• The product mix they may prefer.
• The channels they may prefer to purchase through.
• Their typical order sizes.
• Whether they are likely to have payment problems, which would cause
potential impact on the company’s resources (call centers, order processing,
warehouses, etc.).
23. 20
Increase Customer Profitability Using Data Mining and Advanced Analytics
To achieve this analysis, a decision tree is often the most useful tool. But rather than
applying behaviors to segment the customers, we apply demographic indicators – and
once a segment is identified, we overlay that initial analysis with a behavioral analysis.
APPENDIX 5
Customer Equity Analysis
The final stage in this process is to move to some form of lifetime value analysis.
Much of the literature assumes that customers advance on an incremental basis,
gradually growing over time to buy more and newer products, right up to the point
at which they leave.
However, with the much greater depth of knowledge we have on customer behaviors
and which ones are significant, one of the clear findings is that customers are not
generally incremental in nature. Instead, they tend to be relatively static until they go
through some form of state change transformation (such as leaving school, getting
married or losing a job). With our ability to identify how customers typically behave in
each of these states, and the propensity of them to move between states, we can
develop a more realistic approach. That approach would be to not look at individual
customer lifetime value, but to look at the potential value of a particular segment –
including how it will change over time as new customers are acquired through transfer
or acquisition, and how they are lost either through transfer or churn. The diagram
below shows this more realistic model of customer equity analysis.
24. Increase Customer Profitability Using Data Mining and Advanced Analytics
21
APPENDIX 6
Customer Strategy
Where ABM has been applied to strategy, it has typically been focused on structural
issues: how to organize departments to achieve economies of scale, and what
markets to continue pursuing. But for a marketer thinking about their customer
strategy, customer profitability and behavioral analysis should be foundation stones.
However, this need differs from that of the person looking to identify and understand
the sorts of unusual customers who are identified by a decision tree.
To understand customers, it’s important to understand a broad sweep of customer
behaviors, and to identify large segments of customers with similar patterns of
behavior for which they need to develop a strategy. For this purpose, a technique
called cluster analysis becomes invaluable. In cluster analysis, all business drivers
are considered equally important for the segmentation. This is unlike a decision tree,
where there is a clear target variable (typically profitability) and explanatory variables
(the key ratios). In cluster analysis, all variables are tested, and the significant ones that
indicate customers with similar patterns of behavior (including things like profitability)
are identified and used to segment the customers.
With the members of each cluster identified, other information can be overlaid on the
analysis to deepen the picture, and appropriate strategies can be developed. In the
case of a technology wholesaler, six clusters were identified; of this, four represented
the core of the business for which strategies were deployed.
25. 22
Increase Customer Profitability Using Data Mining and Advanced Analytics
The first group (PG) was happy to pay for a relatively high service level and
consequently was very profitable, so a strategy of “cuddle” was developed. The
UNeg group was very similar to the PG group in many ways, but a significant portion
of their purchase mix involved redundant technology that was sold at a loss, but
should probably never have still been in stock. The strategy here was to “cure”
this stock management problem and return these customers to profit. The core of
the business came from those in the PNorm group; low-effort customers with an
OK margin who we needed to “keep” as customers. Finally, there was a class of
customers called UNorm, who asked for the earth but did not want to pay; and for
those, an active “cull” strategy was developed. Under this strategy, they were offered
terms that would make them profitable if accepted – but if the terms were rejected,
the wholesaler would no longer supply them.
26. Increase Customer Profitability Using Data Mining and Advanced Analytics
23
About SAS
SAS is the leader in business analytics software and services, and the largest
independent vendor in the business intelligence market. Through innovative solutions
delivered within an integrated framework, SAS helps customers at more than 50,000
sites improve performance and deliver value by making better decisions faster. Since
1976 SAS has been giving customers around the world THE POWER TO KNOW®
.