Sourcing & Procurement Analytics for the modern enterprise


Published on

BRIDGEi2i has frameworks to establish Analytics CoE for Supply Chain functions within organizations. Demand planning solution of BRIDGEi2i aims at using advanced statistical forecasting coupled with real-time decision engines models for demand planning, inventory optimization.

Published in: Data & Analytics
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Sourcing & Procurement Analytics for the modern enterprise

  1. 1. @ 2015 BRIDGEi2i Analytics Solutions Pvt. Ltd. All rights reserved Sourcing & Procurement Analytics for the modern enterprise Arun Krishnamoorthy Director - Supply Chain & Pricing Analytics Practice
  2. 2. Our Core Supply Chain Offerings 2 COMMODITY INTELLIGENCE DEMAND FORECASTING QUALITY ANALYTICS FRIEGHT SPEND REDUCTION INVENTORY MODELLING EXCESS & OBSOLETE CONTROL S O U R C I N G & P R O C U R E M E N T C o E S U P P LY & D E M A N D P L A N N I N G C o E M A N U FA C T U R I N G O P E R AT I O N S C o E Understand your commodity landscape and stay in-the-know of factors that affect prices Develop better statistical demand forecasting models to match market dynamics Improve utilization/ yield and reduce failures by employing a predictive control process Analytical control of Freight and other non- material spend Continuous tracking & optimization of inventory to improve SC agility Control Excess & obsolete costs by bringing predictability into demand BRIDGEi2i has frameworks to establish Analytics CoE for Supply Chain functions within organizations INDIRECT PROCUREMENT PLAN TRACKING DASHBOARDS ORDER FULFILLMENT Identify opportunities to reduce indirect spend through supply base optimization Track revenue, bookings and builds along with backlogs and inventory – Real-time Build an “analytical control tower” that alerts delayed orders & bottlenecks before time
  3. 3. 3 ValueRealization Timeframe Low High Commodity Intelligence Component & Commodity Cost Forecasting Forward-Buy & Contract Recommendations Design Solve Implement Track Value & Learn • Understand cost drivers of Memory, Resins and base metals • Provide monthly intelligence to Commodity managers Commodity Dashboards Procurement Risk Mgmt Forward-buy Opportunity Smart Contracting Build analytics capacity at affordable cost • Automate commodity intelligence • Scale to HDD, Panels, Rare metals, batteries, power supplies • 80% spend had should- costs 6 Months 12 Months 24 Months 36 Months > $1bn cost savings by leveraging analytics • Ability to predict commodity prices using drivers understanding • Predict inflexion points • Mature process for intelligence • Leverage price forecasts for large forward buys • Identify opportunities to bake price intelligence, rebate mechanisms and share-of-wallets into a smart contract offering for suppliers +10% Price Forecast Accuracy +25% Portfolio Price Forecast Accuracy +25% Spend Forecast Accuracy Net Impact >100X ROI The Sourcing Analytics CoE in Action – a Case Study Client : A global Fortune 500 PC and printing company Length of Relationship : 3+ years
  4. 4. How does it work? 4 Identify Imperatives Accelerate Solutions Realize Impact • Identify a business challenge • Employ data analytics to address the challenge in a smaller set-up • Scale and Build analytics solution into systems • Make it accessible to operations • Ensure expected impact is realized • Identify new gaps in process efficiency BRIDGEi2i partners with businesses to form an Analytics Center of Expertise (A CoE) Our CoE will  Learn your business from an analytical standpoint  Embed the knowledge within analytical solutions  Make analytics accessible, actionable and operational  Ensure sustained impact
  5. 5. A few case studies on Commodity Intelligence
  6. 6. Commodity Intelligence Solutions Commodity Profiling Develop a deep understanding of commodity landscape by corroborating information in various intel reports. Develop KPIs for the commodity specific to business Commodity Profiles corroborated & created from multiple industry reports Commodity Intelligence for Precious Metals For a Fortune 100 High-Tech company Commodity Intelligence for Memory (DRAM) For a Fortune 500 Networking and Storage company Commodity Intelligence for Plastics For a buyer of commodity plastics (ABS/ HIPS) in Singapore Tracking & Monitoring Commodity KPIs Detailed profiling of key target accounts to assess financial performance, future objectives, potential technology spending to build better customer understanding Case Studies Correlate factors such as demand, supply, inventory and prices to draw a holistic picture of the commodity
  7. 7. Cost Forecasting Solutions Fundamental Factors Demographics, weather, trade flows, production quotas and export controls influencing the demand and supply of commodities. Understand Drivers of Commodity Costs Value at Risk Measurement Providing beforehand visibility into the risk associated with future prices outlook and specific purchase commodity price over the time horizon. Continuous Tracking Continuously track the performance of the price forecasting model to ensure the minimal divergence from the actual commodity prices and immediate intervention in the forecasting model. Hope For The Best But Plan For The Worst Price Forecasting model Build a mathematical model to accurately predict the future commodity prices depending on the historical data decompositions and driver impact. Macroeconomic Factors Demand & supply side economic factors like investments, savings, labour indices etc. Other Factors Substitute material prices, global political situations, price speculation determines commodity prices. Scenario Forecasting Stochastic forecasts based on low probability and high impact exceptional scenarios to plan for the worst situations. Develop Spectrum Of Mutually Beneficial Contracting Terms Supplier Collaboration Develop different types of contracts (buy-back, revenue sharing, quantity flexibility) to create negotiation friendly environment and positively engage supplier for conversation. RFQ Design Historical supplier performance analysis on the contract attributes to develop preferred list of potential suppliers to participate in the RFQ Attribute Selection Identification of negotiable contract attributes price, rebates, share-of-wallet, lead times etc. for the negotiations. Contract Design Analysis of contract attributes to lay down terms of contract make the deals interesting for suppliers and cost efficient for buyers Case Studies BRIDGEi2i’s Bachelier Commodity Price Prediction Tool A unified analytics platform for cost management Advanced Cost Forecasting model developed For a Fortune 10 high-tech company
  8. 8. Case Study : Memory Procurement Risk Management 88 • Corroborate and validate info from multiple market reports • Metricize market demand sufficiency • Understand impact of macro variables – PC demand, DDR2-DDR3 transition, confidence indices etc. • Set-up the multi-variate forecasting models for buy- price with identified drivers • Add an innovation effect due to spot market speculations • Develop price forecasting models using VAR, VECM and Bayesian models (available in SAS) • Automate the modeling process • Profile price forecasting accuracy and track based on REACT (recursive accuracy testing) framework • Track the drivers’ influence regularly to estimate model maintenance schedules  An accurate memory price forecasting model – especially to predict inflexion points in prices  ~93% accuracy 3 months out and >85% 6 months out  Low-touch, self- learning models Data Key Features Outcome Driver Identification Multi-variate forecasting models Profiling & Automation •Historical buy-price data for commodity •Spot market prices from DRAM Exchange •Market reports from multiple industry watchers – inSpectrum, Market View, Gartner etc. •Planned demand volumes • To accurately forecast prices of memory (1gb equivalents) based on true drivers of prices • To create a repeatable process to give strategic sourcing and commodity managers proactive insights on the commodity Objective  BRIDGEi2i’s Bachelier Tool has a suite of forecasting models configured for commodity price forecasting  Ability to run what-if forecasts designed for self-driven insights designed for commodities designed for actionability
  9. 9. Embedding Analytics back in Client Systems BACHELIER – Commodity Price Forecasting Engine DATA MANAGEMENT & TREATMENT FORECAST FORECAST EVALUATION DECISION ENGINE & WHAT-IF FORECASTS Easy & intuitive interface for data management and treatment Ensemble forecasts made from strong & advanced forecasting models Make “What-if” forecasts and forward-buy decisions while understanding risk involved Rigorously tested forecasts to ensure maximum confidence in numbers
  10. 10. A few case studies on Indirect Procurement
  11. 11. Our Procurement Analytics Solution 11 Data Enrichment Define matching attributes Integrate data across sources Augment data from other sources like contracts text, websites Business Objectives Outcome Process and frameworks to proactively identify and minimize cost Data driven frameworks to establish contract terms and ensure compliance Improve ability to capture information, analyze for insights and enable informed decision making Identify opportunities to minimize cost in each category Category Supplier Identify supplier consolidation, rate rationalisation opportunities In and Across categories Identify opportunities to optimize contract terms Leverage transaction data to segment spend into categories, analyze supplier distribution, spend coverage etc. Within category & sub categories analyse spend type, supplier performance, rate variation, dependency and presence across categories Analysis of rebates and payment terms to lay down terms of contract that make the deals interesting for suppliers and cost efficient for buyers Contract Check completeness of key fields Cleanse data – consistent names, abbreviations, units etc. Scorecard metric or KPI to measure progress toward a goal Analytics Segmentation Text Analytics Variance Drivers Behavior Analysis Forecasting Dashboards and Alters
  12. 12. Approach to Identify Opportunities of Cost Minimization 1212 • Develop process for accurate mapping product and item to defined spend categories • Segment each spend categories based on recency, frequency and value of transactions • Identify similar categories using attribute analysis • Concentration of buyers by category • Understand buyer behavior and opportunities to aggregate spend across buyers • Analyze transaction channels and associated cost • Identify top suppliers in each category • Understanding commodity- business-supplier mapping to reveal overlaps Prioritize categories with opportunities to minimize cost Develop data enrichment to identification of cost minimization opportunities Destination-> AdelaideBrisbane Bulwer Darwin Kurnell Kwinana Lytton Perth Sydney Adelaide 0% 1% 2% 6% 2% 3% 5% 4% 7% Brisbane 1% 0% 0% 0% 1% 0% 0% 2% 1% Darwin 5% 0% 2% 0% 0% 0% 5% 4% 1% Geelong 0% 0% 0% 0% 0% 0% 0% 0% 1% Kurnell 1% 0% 0% 0% 0% 0% 4% 0% 0% Lytton 2% 0% 0% 0% 2% 0% 0% 0% 2% Melbourne 2% 0% 0% 0% 1% 0% 2% 0% 1% Perth 2% 0% 0% 8% 0% 0% 0% 0% 2% Sydney 7% 1% 1% 1% 1% 0% 4% 3% 0% Adelaide 0% 1% 0% 2% 1% 2% 0% 4% 10% Brisbane 0% 0% 0% 1% 1% 0% 0% 0% 1% Darwin 1% 0% 0% 0% 0% 0% 0% 1% 0% Geelong 0% 0% 0% 0% 0% 0% 0% 0% 0% Kurnell 1% 0% 0% 0% 0% 0% 0% 0% 0% Lytton 2% 0% 0% 0% 1% 0% 0% 0% 3% Melbourne 0% 1% 0% 0% 0% 0% 0% 0% 4% Perth 2% 1% 0% 5% 1% 0% 0% 0% 7% Sydney 4% 2% 6% 1% 2% 11% 4% 16% 0% CY 2010 CY 2011 Leveraged BRIDGEi2i text mining solution to appropriately augment missing data Segmentation of categories using RFM technique to identify top spend segments Data Approach Outcome CATEGORY ANALYSIS BUYER BEHAVIOR SUPPLIER CONCENTRATION Categories details (UNSPSC) Supplier Firmographics Buyers details Transaction details Our EXPERIENCE
  13. 13. Approach to Drive Cost Efficiency In & Across Categories 1313 • Identify aberrations in pricing across region/period in same category • Identify large variances in rates across suppliers for the same category & similar supplier performance score • Identify perceptible fee deviations from agreed rate card • Scorecards to rank suppliers based on predefined metrics • Business inputs to validate preference of suppliers • Develop list of preferred suppliers with presence in multiple or across categories • Build aggregate demand forecast across buyer groups • Develop standardized discounted rate cards in exchange for volume commitment & a larger share-of-wallet Consolidated list of suppliers and contract terms to enable YoY deflation of spend Enabled Implementation of processes and frameworks to minimize procurement • Developed and list of preferred suppliers taking consideration buyer preferences • Standardized rate cards to minimize rate aberrations Identification of aberrations in pricing across region for the same category Overall and Category wise preferred list of suppliers and suggested rate cards to minimize spend Data Approach Outcome CATEGORY SPEND PATTERNS SUPPLIER CONSOLIDATION DEMAND FORECASTPricing details Business inputs/needs Transaction details Supplier Firmographics Contract terms details Our EXPERIENCE
  14. 14. Approach To Drive Cost Efficiency In Tier 3 Supply Base 14 Isolate Tier 3 vendors • Identify and separate Tier 3 vendors • By non-ASL • By absence of contracts, catalogues etc. Statistical indexing • Identify category-wise spend concentration • Gini or HH index • Relative importance of vendor in category based on statistics Textual analysis • Identify what is actually purchased • Extraction of object in text of line item description • Extraction of context of purchase from text Product mapping • Identify similar products/ services in Tier 1 & 2 universe • Mapping of object to contracted purchase list • Identify better suppliers of same product/service Supplier dependency • Identify how and why the Tier 3 supplier is used • Price competitiveness • Region/ business/ unique product dependencies Initiate consolidation • Confirm analysis with buyers and initiate consolidation • Conversation with Tier 3 supplier for potential contracting • Conversation with best alternate supplier for rate card discussions 1 2 3 4 5 6 The Long Tail Problem in Indirect Sourcing Tier1~X suppliers;Yspend Tier2~5Xsuppliers; Y*0.2spend Tier 3 ~ 35X suppliers; Y*0.2 spend30% 60% 90% %CumulativeSpend # of Vendors ----> Challenge is with Tier 3 suppliers where 1. Supplier spread is high – hard to identify the big ones 2. Supplier products or service are misclassified – hard to identify what is purchased 3. Supplier mapping is unknown – hard to map their products/ services to capabilities of Tier 1 & 2 4. Supplier dependency in unknown – low spend concentration implies less insight into why the supplier exists 15% IP Spend 0 2 4 6 8 1 22 43 64 85 106 127 148 169 190 211 232 253 274 295 316 337 358 379 400 421 442 463 484 505 526
  15. 15. Thank You