Predictive Modelling, not magic - FOM Jam slides

RedEye
RedEyeRedEye
Predictive and Data – How
much is actually needed?
Proprietary
Predictive Modelling, not magic - FOM Jam slides
3
Proprietary
How much data do you need?
What is the optimal number of
variables to use in a predictive
model?
What depth of web activity
information do I need to
include in my models?
Should I include all
transactional data or can I look
at individual channels?
I have 10 years’ worth of
transactional data in a
perfectly organised database.
Do I need to have it all
available to my models?
I launched a new website.
Does this mean my tag
information is no longer valid
in my predictive models?
Predictive Modelling, not magic - FOM Jam slides
5
Proprietary
How much data do you need?
What is the optimal number of
variables to use in a predictive
model?
What depth of web activity
information do I need to
include in my models?
Should I include all
transactional data or can I look
at individual channels?
I have 10 years’ worth of
transactional data in a
perfectly organised database.
Do I need to have it all
available to my models?
I launched a new website.
Does this mean my tag
information is no longer valid
in my predictive models?
Consider this example..
Two retailers who sell their goods through a number of channels. Their core
customers are from the same demographic and they have a similar price
points in their product ranges.
7
How many peaks
do you have that
appear cyclically
(for example
consecutive
Christmases)?
How often do your
customers
purchase your
core product
offering?
Have there been
any significant
changes in your
business strategy,
product offering or
wider economic
climate which
need to be tracked
over time?
Retention Period
The length of time that your
data looks at depends
completely on your customers
and business. Some key
aspects to keep in mind are
repurchase windows, annual
peaks and significant events
Proprietary
Proprietary
Bespoke vs
Black Box?
Predictive
modelling
Customer ID
& Cross-device
tracking
Orchestration
Reporting
& Analytics
DataIntegration
Personal
& Demographic
Onsite Behavioural
Data
Engagement
Data
Transactional
Data
Mobile & Device
Data
Data
APP Data
Lifestyle Data
Email
Direct Mail
Paid Social
SMS
AdWords
Web
Push
Store
Multi-channel
& Data
Segmentation
9
10
Proprietary
Can you implement the actions
from your models?
How often do the models get
refreshed?
Does the model account for
current behavioural trends?
How do I use the scores to
segment my customers?
How do I include artificial
intelligence in my day to day
campaign activity?
Can I set up an automated
campaign based on predicted
behaviour?
What role does machine
learning play in helping me
identify my customers along
their lifecycle?
Artificial Intelligence
AI involves machines
that can perform tasks
that are characteristic
of human intelligence
but on a greater scale
Machine Learning
A process for gathering
information and analysing
data to provide learnings for
AI to be successful.
Predictive Analytics
A specific set of algorithms
that provides the likelihood
that something will happen
in the future, eg likelihood
to make a purchase.
11
What's
happening
currently
Predict LearnAct
Proprietary
Customer
response
12
13
Proprietary
How do you measure impact?
14
Is there statistical
validation around
how good the
model is?
Do you have
visibility of score
distributions and
why someone has
high or low
likelihood to
complete an
action?
Is there
significance testing
associated with
reporting on
control cells in
campaigns?
How do you know that the
changes in behaviour you are
seeing are down to your
campaign and not random
chance?
Proprietary
Reporting
Breadth of Data – Pareto Principle or the 80/20 rule
Time Period – Based on the business and length of time the model is covering
Do you build your own or go black box?
Is it actually doing what you intended it to?
How do you know it is working? Are you reporting accurately?
In Conclusion….
Thank you for your time
16
1 of 16

Recommended

Lose the Crystall Ball - FOM Jam slides by
Lose the Crystall Ball - FOM Jam slidesLose the Crystall Ball - FOM Jam slides
Lose the Crystall Ball - FOM Jam slidesRedEye
307 views18 slides
Accelerate Revenue with a Customer Data Platform by
Accelerate Revenue with a Customer Data PlatformAccelerate Revenue with a Customer Data Platform
Accelerate Revenue with a Customer Data PlatformLattice Engines
2.9K views49 slides
Top ten tips for implementing Website Personalisation by
Top ten tips for implementing Website PersonalisationTop ten tips for implementing Website Personalisation
Top ten tips for implementing Website PersonalisationRedEye
593 views18 slides
Connecting the Customer Data Dots by
Connecting the Customer Data DotsConnecting the Customer Data Dots
Connecting the Customer Data DotsTreasure Data, Inc.
5.1K views12 slides
Ai driven Predictive Analytics. Enough theory - let's talk about results! by
Ai driven Predictive Analytics. Enough theory - let's talk about results! Ai driven Predictive Analytics. Enough theory - let's talk about results!
Ai driven Predictive Analytics. Enough theory - let's talk about results! RedEye
438 views17 slides
Big Data - How Marketing Has Revolutionised - by Sean Singleton by
Big Data - How Marketing Has Revolutionised - by Sean SingletonBig Data - How Marketing Has Revolutionised - by Sean Singleton
Big Data - How Marketing Has Revolutionised - by Sean SingletonDigital Annexe
12.9K views16 slides

More Related Content

What's hot

Big Data, customer analytics and loyalty marketing by
Big Data, customer analytics and loyalty marketingBig Data, customer analytics and loyalty marketing
Big Data, customer analytics and loyalty marketingKevin May
6K views32 slides
What is Intent Data? by
What is Intent Data?What is Intent Data?
What is Intent Data?Infer
2.7K views39 slides
855 sponsor gazdak_using our laptop by
855 sponsor gazdak_using our laptop855 sponsor gazdak_using our laptop
855 sponsor gazdak_using our laptopRising Media, Inc.
274 views16 slides
Forrester Sales and Marketing Forum - Creating a Data Foundation for Personal... by
Forrester Sales and Marketing Forum - Creating a Data Foundation for Personal...Forrester Sales and Marketing Forum - Creating a Data Foundation for Personal...
Forrester Sales and Marketing Forum - Creating a Data Foundation for Personal...Lattice Engines
672 views23 slides
Teradata Integrated Web Intelligence by
Teradata Integrated Web IntelligenceTeradata Integrated Web Intelligence
Teradata Integrated Web IntelligenceTeradata
1K views21 slides
Bombora - Intent data - The secret to smarter sales prospecting? - April 2016 by
Bombora - Intent data - The secret to smarter sales prospecting? - April 2016 Bombora - Intent data - The secret to smarter sales prospecting? - April 2016
Bombora - Intent data - The secret to smarter sales prospecting? - April 2016 Bombora
2.4K views16 slides

What's hot(20)

Big Data, customer analytics and loyalty marketing by Kevin May
Big Data, customer analytics and loyalty marketingBig Data, customer analytics and loyalty marketing
Big Data, customer analytics and loyalty marketing
Kevin May6K views
What is Intent Data? by Infer
What is Intent Data?What is Intent Data?
What is Intent Data?
Infer2.7K views
Forrester Sales and Marketing Forum - Creating a Data Foundation for Personal... by Lattice Engines
Forrester Sales and Marketing Forum - Creating a Data Foundation for Personal...Forrester Sales and Marketing Forum - Creating a Data Foundation for Personal...
Forrester Sales and Marketing Forum - Creating a Data Foundation for Personal...
Lattice Engines672 views
Teradata Integrated Web Intelligence by Teradata
Teradata Integrated Web IntelligenceTeradata Integrated Web Intelligence
Teradata Integrated Web Intelligence
Teradata1K views
Bombora - Intent data - The secret to smarter sales prospecting? - April 2016 by Bombora
Bombora - Intent data - The secret to smarter sales prospecting? - April 2016 Bombora - Intent data - The secret to smarter sales prospecting? - April 2016
Bombora - Intent data - The secret to smarter sales prospecting? - April 2016
Bombora2.4K views
Harnessing Data for Better Customer Experience and Company Success by Treasure Data, Inc.
Harnessing Data for Better Customer Experience and Company SuccessHarnessing Data for Better Customer Experience and Company Success
Harnessing Data for Better Customer Experience and Company Success
Jill Kaplan - Customer Analytics and Marketing to Drive Demand by Julia Grosman
Jill Kaplan - Customer Analytics and Marketing to Drive DemandJill Kaplan - Customer Analytics and Marketing to Drive Demand
Jill Kaplan - Customer Analytics and Marketing to Drive Demand
Julia Grosman754 views
Elevating customer analytics - how to gain a 720 degree view of your customer by Actian Corporation
Elevating customer analytics - how to gain a 720 degree view of your customerElevating customer analytics - how to gain a 720 degree view of your customer
Elevating customer analytics - how to gain a 720 degree view of your customer
Actian Corporation2.9K views
Customer and marketing analytics: Integrating multichannel data to gain actio... by Mindtree Ltd.
Customer and marketing analytics: Integrating multichannel data to gain actio...Customer and marketing analytics: Integrating multichannel data to gain actio...
Customer and marketing analytics: Integrating multichannel data to gain actio...
Mindtree Ltd. 1.2K views
Benchmarking Your Online Impact: From Stats to Reputation Management by NSI Partners, LLC
Benchmarking Your Online Impact: From Stats to Reputation ManagementBenchmarking Your Online Impact: From Stats to Reputation Management
Benchmarking Your Online Impact: From Stats to Reputation Management
NSI Partners, LLC1.6K views
Marketing automation best practices for insurance companies by edynamic
Marketing automation best practices for insurance companiesMarketing automation best practices for insurance companies
Marketing automation best practices for insurance companies
edynamic5.3K views
How to use Online Marketing Technology to Improve Campaign Performance - Lowe... by Online Marketing Summit
How to use Online Marketing Technology to Improve Campaign Performance - Lowe...How to use Online Marketing Technology to Improve Campaign Performance - Lowe...
How to use Online Marketing Technology to Improve Campaign Performance - Lowe...
Big Data & Analytics 101: How Customer Lifetime Value Enhances Predictive Mar... by Big Cloud Analytics, Inc.
Big Data & Analytics 101: How Customer Lifetime Value Enhances Predictive Mar...Big Data & Analytics 101: How Customer Lifetime Value Enhances Predictive Mar...
Big Data & Analytics 101: How Customer Lifetime Value Enhances Predictive Mar...
Data Visualization Deck presented at I-COM by Joyce Lin
Data Visualization Deck presented at I-COMData Visualization Deck presented at I-COM
Data Visualization Deck presented at I-COM
Joyce Lin296 views
Big Data Marketing - What You Need To Know by MBA & Company
Big Data Marketing - What You Need To KnowBig Data Marketing - What You Need To Know
Big Data Marketing - What You Need To Know
MBA & Company5.7K views
Driving Growth with Marketing Analytics by Vivastream
Driving Growth with  Marketing AnalyticsDriving Growth with  Marketing Analytics
Driving Growth with Marketing Analytics
Vivastream1.4K views
Analytics is the Bookkeeping of Marketing by Petri Mertanen
Analytics is the Bookkeeping of MarketingAnalytics is the Bookkeeping of Marketing
Analytics is the Bookkeeping of Marketing
Petri Mertanen748 views

Similar to Predictive Modelling, not magic - FOM Jam slides

Monitoring Analytics To Create Customer Value And Experience by
Monitoring Analytics To Create Customer Value And ExperienceMonitoring Analytics To Create Customer Value And Experience
Monitoring Analytics To Create Customer Value And ExperienceeTailing India
85 views3 slides
Six Trends in Retail Analytics by
Six Trends in Retail Analytics Six Trends in Retail Analytics
Six Trends in Retail Analytics Tableau Software
12.7K views26 slides
Cloud and business agility by
Cloud and business agilityCloud and business agility
Cloud and business agilityMike ORourke
620 views22 slides
Why connectcust (1) by
Why connectcust (1)Why connectcust (1)
Why connectcust (1)Rouhin Banerjee
242 views3 slides
What to expect_in_2013 by
What to expect_in_2013What to expect_in_2013
What to expect_in_2013ben_d_walker
452 views25 slides
Digital Transformation in Retail by
Digital Transformation in RetailDigital Transformation in Retail
Digital Transformation in RetailHARMAN Services
23.7K views1 slide

Similar to Predictive Modelling, not magic - FOM Jam slides(20)

Monitoring Analytics To Create Customer Value And Experience by eTailing India
Monitoring Analytics To Create Customer Value And ExperienceMonitoring Analytics To Create Customer Value And Experience
Monitoring Analytics To Create Customer Value And Experience
eTailing India85 views
Six Trends in Retail Analytics by Tableau Software
Six Trends in Retail Analytics Six Trends in Retail Analytics
Six Trends in Retail Analytics
Tableau Software12.7K views
Cloud and business agility by Mike ORourke
Cloud and business agilityCloud and business agility
Cloud and business agility
Mike ORourke620 views
What to expect_in_2013 by ben_d_walker
What to expect_in_2013What to expect_in_2013
What to expect_in_2013
ben_d_walker452 views
Digital Transformation in Retail by HARMAN Services
Digital Transformation in RetailDigital Transformation in Retail
Digital Transformation in Retail
HARMAN Services23.7K views
The Customer Data Platform, the Future of the Marketing Database by RedEye
The Customer Data Platform, the Future of the Marketing DatabaseThe Customer Data Platform, the Future of the Marketing Database
The Customer Data Platform, the Future of the Marketing Database
RedEye3.2K views
Data Driven Marketing: the DNA of customer orientated companies by Good Rebels
Data Driven Marketing: the DNA of customer orientated companiesData Driven Marketing: the DNA of customer orientated companies
Data Driven Marketing: the DNA of customer orientated companies
Good Rebels2.3K views
Applying Data Science Across the Ten Stages of the Retail Lifecycle by Suffiyan Syed
Applying Data Science Across the Ten Stages of the Retail LifecycleApplying Data Science Across the Ten Stages of the Retail Lifecycle
Applying Data Science Across the Ten Stages of the Retail Lifecycle
Suffiyan Syed174 views
Data Science - Part I - Sustaining Predictive Analytics Capabilities by Derek Kane
Data Science - Part I - Sustaining Predictive Analytics CapabilitiesData Science - Part I - Sustaining Predictive Analytics Capabilities
Data Science - Part I - Sustaining Predictive Analytics Capabilities
Derek Kane17.6K views
ai-powered-marketing-and-sales-reach-new-heights-with-generative-ai.pdf by jason668539
ai-powered-marketing-and-sales-reach-new-heights-with-generative-ai.pdfai-powered-marketing-and-sales-reach-new-heights-with-generative-ai.pdf
ai-powered-marketing-and-sales-reach-new-heights-with-generative-ai.pdf
jason66853974 views
10 signs you should invest in Web Scraping by PromptCloud
10 signs you should invest in Web Scraping10 signs you should invest in Web Scraping
10 signs you should invest in Web Scraping
PromptCloud1.5K views
Certus Accelerate - Building the business case for why you need to invest in ... by Certus Solutions
Certus Accelerate - Building the business case for why you need to invest in ...Certus Accelerate - Building the business case for why you need to invest in ...
Certus Accelerate - Building the business case for why you need to invest in ...
Certus Solutions192 views
AGILONE-ACADEMY_The-State-of-Big-Customer-Data-2015-FINAL3 by Angela Sanfilippo
AGILONE-ACADEMY_The-State-of-Big-Customer-Data-2015-FINAL3AGILONE-ACADEMY_The-State-of-Big-Customer-Data-2015-FINAL3
AGILONE-ACADEMY_The-State-of-Big-Customer-Data-2015-FINAL3
Angela Sanfilippo278 views
Digital trends: AI in marketing by asthajain100
Digital trends: AI in marketingDigital trends: AI in marketing
Digital trends: AI in marketing
asthajain10055 views
Top 5 Strategies for Retail Data Analytics by Hortonworks
Top 5 Strategies for Retail Data AnalyticsTop 5 Strategies for Retail Data Analytics
Top 5 Strategies for Retail Data Analytics
Hortonworks2.2K views
Data Science Use Cases in Retail & Healthcare Industries.pdf by Katy Slemon
Data Science Use Cases in Retail & Healthcare Industries.pdfData Science Use Cases in Retail & Healthcare Industries.pdf
Data Science Use Cases in Retail & Healthcare Industries.pdf
Katy Slemon117 views
Enhanced auto shopping experience through analytics path by Marketing Material
Enhanced auto shopping experience through analytics pathEnhanced auto shopping experience through analytics path
Enhanced auto shopping experience through analytics path
Problems And Exercises. 1. Business Analytics Focuses On by Michele Thomas
Problems And Exercises. 1. Business Analytics Focuses OnProblems And Exercises. 1. Business Analytics Focuses On
Problems And Exercises. 1. Business Analytics Focuses On

Recently uploaded

Recently uploaded(20)

AZoNetwork Newsletter Audiences 2024.pdf by Rebecca731061
AZoNetwork Newsletter Audiences 2024.pdfAZoNetwork Newsletter Audiences 2024.pdf
AZoNetwork Newsletter Audiences 2024.pdf
Rebecca73106159 views
DemandMore Example Monthly Deck by WesleyParker10
DemandMore Example Monthly DeckDemandMore Example Monthly Deck
DemandMore Example Monthly Deck
WesleyParker1020 views
Deltaplan - SEO Search by Roy Huiskes
Deltaplan - SEO SearchDeltaplan - SEO Search
Deltaplan - SEO Search
Roy Huiskes47 views
The Lore of Entelect by mike719672
The Lore of EntelectThe Lore of Entelect
The Lore of Entelect
mike71967227 views

Predictive Modelling, not magic - FOM Jam slides

  • 1. Predictive and Data – How much is actually needed? Proprietary
  • 3. 3 Proprietary How much data do you need? What is the optimal number of variables to use in a predictive model? What depth of web activity information do I need to include in my models? Should I include all transactional data or can I look at individual channels? I have 10 years’ worth of transactional data in a perfectly organised database. Do I need to have it all available to my models? I launched a new website. Does this mean my tag information is no longer valid in my predictive models?
  • 5. 5 Proprietary How much data do you need? What is the optimal number of variables to use in a predictive model? What depth of web activity information do I need to include in my models? Should I include all transactional data or can I look at individual channels? I have 10 years’ worth of transactional data in a perfectly organised database. Do I need to have it all available to my models? I launched a new website. Does this mean my tag information is no longer valid in my predictive models?
  • 6. Consider this example.. Two retailers who sell their goods through a number of channels. Their core customers are from the same demographic and they have a similar price points in their product ranges.
  • 7. 7 How many peaks do you have that appear cyclically (for example consecutive Christmases)? How often do your customers purchase your core product offering? Have there been any significant changes in your business strategy, product offering or wider economic climate which need to be tracked over time? Retention Period The length of time that your data looks at depends completely on your customers and business. Some key aspects to keep in mind are repurchase windows, annual peaks and significant events Proprietary
  • 9. Predictive modelling Customer ID & Cross-device tracking Orchestration Reporting & Analytics DataIntegration Personal & Demographic Onsite Behavioural Data Engagement Data Transactional Data Mobile & Device Data Data APP Data Lifestyle Data Email Direct Mail Paid Social SMS AdWords Web Push Store Multi-channel & Data Segmentation 9
  • 10. 10 Proprietary Can you implement the actions from your models? How often do the models get refreshed? Does the model account for current behavioural trends? How do I use the scores to segment my customers? How do I include artificial intelligence in my day to day campaign activity? Can I set up an automated campaign based on predicted behaviour? What role does machine learning play in helping me identify my customers along their lifecycle?
  • 11. Artificial Intelligence AI involves machines that can perform tasks that are characteristic of human intelligence but on a greater scale Machine Learning A process for gathering information and analysing data to provide learnings for AI to be successful. Predictive Analytics A specific set of algorithms that provides the likelihood that something will happen in the future, eg likelihood to make a purchase. 11
  • 13. 13 Proprietary How do you measure impact?
  • 14. 14 Is there statistical validation around how good the model is? Do you have visibility of score distributions and why someone has high or low likelihood to complete an action? Is there significance testing associated with reporting on control cells in campaigns? How do you know that the changes in behaviour you are seeing are down to your campaign and not random chance? Proprietary Reporting
  • 15. Breadth of Data – Pareto Principle or the 80/20 rule Time Period – Based on the business and length of time the model is covering Do you build your own or go black box? Is it actually doing what you intended it to? How do you know it is working? Are you reporting accurately? In Conclusion….
  • 16. Thank you for your time 16