Managing Supplier Performance with Advanced Analytics
- 1. March 26 – 29, 2017 Reno, NV Managing Supplier Performance with Advanced Analytics Judy Smith Assistant Director, Strategic Sourcing Howard Hughes Medical Institute Dan Traub Founder and CEO FinVantage Solutions LLC
- 2. Great Ideas. Great Expectations. Great Rewards. At your next supplier meeting, are you: Able to compare price movement to CPI? Able to illustrate incentives for the supplier? Able to discuss pricing variation data? Able to bring specific data to the table that improves accountability and delivers added value to the relationship?
- 3. Great Ideas. Great Expectations. Great Rewards. • HHMI and Strategic Supplier Management Background • Project Vision and Objectives • Approach and Solution Walkthrough – Focus Area: Pricing Accuracy / CPI – Focus Area: Market Share Analysis • Reactions and Results as we go • Future Opportunities • Getting Started at your Institution Today’s Session
- 4. Great Ideas. Great Expectations. Great Rewards. • Overview – A science philanthropy, HHMI’s mission is to advance biomedical research and science education for the benefit of humanity – Headquartered in Chevy Chase, MD – 2,600 employees – Self funded; $663M in medical research, $86M science education in 2016 – Labs are located on more than 60 Universities and hospital campuses across the US • 300 investigators, 180 are members of the National Academy of Sciences and 25 are Nobel laureates – Janelia Research Campus an HHMI based research facility in specializing in neuroscience and imaging • Strategic Supplier Program – Started in 2012 – Built using internal procurement staff, located across the U.S. – HHMI did not have formal contracts in place for any suppliers, no primary supplier relationships, no supplier management HHMI / Strategic Supplier Management Background
- 5. Great Ideas. Great Expectations. Great Rewards. • Basic Supplier Relationship Management started at HHMI in 2013 • Most of the SRM efforts relied upon Excel-based analytics that were created one supplier at a time by a Procurement Analyst (1 for the sourcing team) • In order to expand the SRM program to our 53 catalog suppliers and 20+ high spend equipment suppliers we needed a tool to expand and enhance the effort • HHMI partnered with FinVantage Solutions to create a supplier scorecard using Tableau’s Software business intelligence tools to enhance HHMI’s supplier analytics, now sourcing team members can invoke and update the scorecard throughout the year • The team went from sharing high level spend information and catalog pricing consistency to a robust platform that evaluates pricing changes, market shares, campus market shares at a lab level, freight costs changes and more Project Vision and Objectives
- 6. Great Ideas. Great Expectations. Great Rewards. Approach and Solution Walkthrough
- 7. Great Ideas. Great Expectations. Great Rewards. • HHMI partnered with performance analytics consulting firm FinVantage Solutions to enhance and expand previous analytics environment • HHMI IT makes Tableau Software’s Server and Desktop tools available to access PeopleSoft and Ariba data • Augmented data with in-house Access tool for market shares, expanded supplier attributes • Solution designed to empower strategic sourcing team members to produce, monitor, and deliver supplier-specific scorecards throughout the year • Additional capabilities: deliver a rapid analysis environment where data can be easily “explored” and “visualized” to pursue new insights Expertise and Tools
- 8. Great Ideas. Great Expectations. Great Rewards. Project Implementation Activities 2013-16 2016 2017+ Deliver Excel-based prototype Design and build current solution with FinVantage Solutions Publish first scorecard – 16 Suppliers Deliver revisions based on feedback Continue/expand use in quarterly reviews • Key roles: HHMI management, analyst role, IT, FinVantage team • Respect the importance and effort involved in collecting / mapping supplier performance data • Address the domain knowledge gap of internal IT (Tableau owners) with focused outside consultants • An important justification of this project: automating the repetitive work of our in-house analyst
- 9. Great Ideas. Great Expectations. Great Rewards. Solution – Spend Summary Summary of total spend with 4 plus current FY trend Shows ordering method trend Provides aggregate spend by unit, the ultimate customer known by supplier
- 10. Great Ideas. Great Expectations. Great Rewards. Solution – Ordering Details Type of order (AN=Ariba Network) with forecast trend and confidence bands Freight percentage trend provided as percent of spend total, if available Item-level details provide important insight to what is important to our users and relationship
- 11. Great Ideas. Great Expectations. Great Rewards. • Objective was to develop key metrics about pricing variation (consistency over time) and to compare with CPI. • Using reconciled invoice SKU-level pricing (not PR, PO, or catalog data). Use low matching tolerance to catalog pricing. • Measure pricing “spread” over time: Coefficient of Variation = Std. Deviation (unit prices for a SKU) Average (unit prices for a SKU) • Determine a Cost Index basket of goods – Automatic, based on SKUs purchased in both periods • Example: Analysis Period 1/1/2016 – 12/31/2016 Base Period 1/1/2015 – 12/31/2015 – Dynamic to reflect changes over time, with selectable date ranges Solution – Pricing Accuracy
- 12. Great Ideas. Great Expectations. Great Rewards. Calculate a supplier’s Cost Index specific to your procurement history: • Laspeyres technique, similar to CPI calculation • Also known as “base year quantity weight” technique. where: P=unit pricing new/old=analysis vs base period Q=quantity purchased • Compute for goods basket of all purchased SKUs both in the base period and in the analysis period • Key assumption is that quantities purchased will remain static Solution – Pricing Accuracy Example: Extend this calculation for all SKUs in the supplier cost index basket Base “old” 2015 Analysis “new” 2016 Po Qo Pn Po × Qo Pn × Qo SKU1 $100 5 $ 110 500 550 SKU2 $125 15 $ 135 1875 2025 Total: 2375 2575 Cost Index = (𝑃 𝑛 × 𝑄 𝑜) (𝑃 𝑜 × 𝑄 𝑜) = 2575 2375 × 100 = 108.42 Cost Index = (𝑃 𝑛𝑒𝑤 ∙ 𝑄 𝑜𝑙𝑑) (𝑃 𝑜𝑙𝑑 ∙ 𝑄 𝑜𝑙𝑑)
- 13. Great Ideas. Great Expectations. Great Rewards. Solution – Pricing Accuracy Automatically pulls CPI All Urban Consumer data Popular Items Analysis provides depth for SKUs most important at HHMI Challenge : UOM variation
- 14. Great Ideas. Great Expectations. Great Rewards. • Objective was to give suppliers an incentive by showing opportunity to grow share of Institute category spend • Data Development & Capture approach • Focused on “Top 3” Markets provided by supplier to our constituents specifically • Example: Gloves (conceptual data shown) – Overall spend on gloves category: $1.0M Solution – Market Share Supplier 12 Mo. Spend (All) Est. % of Spend with Supplier on Gloves Est. Spend with Supplier on Gloves Supplier’s Market Share of Gloves Category Supplier A $ 7.0M 10% $ 700K 70% (1st) Supplier B $ 400K 25% $ 100K 10% (3rd) Supplier C $ 200K 100% $ 200K 20% (2nd)
- 15. Great Ideas. Great Expectations. Great Rewards. Solution – Market Share Ranking over time – “Bump Chart” Only current supplier’s name is shown Configurations: - Markets of interest - Show/hide actual category spend
- 16. Great Ideas. Great Expectations. Great Rewards. Solution – Packaging Title Page Notes, Glossary, and Contact Information Not shown: - Configuration settings - Interactive drilldown/ export
- 17. Great Ideas. Great Expectations. Great Rewards. • Suppliers are surprised at the level of detail Sourcing has at its fingertips – Many suppliers have indicated that “it’s better data than they have access to” without going to a reporting group for a special report • Pricing issues can be found at a PO level very swiftly – this results in investigation by the supplier, in some instances refunds, and communication with field sales reps to correct erroneous quoting (greater than contract price) • The data helps us to understand improvement opportunities for contract compliance at a lab level • Market share data shows suppliers opportunity to grow their market share, this data provides HHMI leverage because of our ability to move spend • Comparisons to CPI have been very valuable in negotiations for price changes that the supplier presents to HHMI – Suppliers are not thrilled with using CPI, they would rather use a specific PPI, but we continue to push for the CPI do to the simplicity; thus far it is a very helpful gage Reactions and Results
- 18. Great Ideas. Great Expectations. Great Rewards. • We still see opportunity in developing more market data at a commodity level • HHMI has recently allowed end users to have Pcards – Need compliance reports to monitor use of the card • My staff is still learning the full capabilities of the Scorecard and building an analytics mindset/culture – Drill-down / interactive mode is very helpful – Some spin-off analysis have resulted – and we need to get those into Tableau • Supplier-specific versions – Would be helpful especially for Equipment suppliers Future Opportunities
- 19. Great Ideas. Great Expectations. Great Rewards. • First, determine the key metrics that matter most based on the goals of the strategic supplier relationship • Consider the types of data you have available • Build an ally with executive sponsor for holding suppliers accountable through performance measurement • Approach: Excel is not a bad way to begin, but consider the effort involved in: – Repeating the measurements over time (e.g. QBRs) – Scaling to your desired number of suppliers Getting Started at your Institution • FinVantage Solutions can assist: – Building out your in-house analytics / Tableau environment, or – Software-as-a-Service analytics solution (including pre-built scorecards and an extensive library of KPIs)
- 20. Great Ideas. Great Expectations. Great Rewards. Please provide your business card, and we will be happy to forward the presentation and a sample scorecard to you. Judy Smith smithj@hhmi.org Dan Traub dan@finvantage.com Thank You!