Six Sigma Sourcing (and Procurement) – A Bridge Too Far?
Categories: Guest Post, Procurement Strategy & Planning, Sourcing | Tags: Process and Best Practice, Zycus
Spend Matters welcomes a guest post from Richard Waugh, vice president of corporate development at Zycus Inc.
Six Sigma, the process measurement methodology originally developed by Motorola and popularized by GE, has far-reaching applications beyond its manufacturing roots. Applying the discipline of Six Sigma to improving sourcing and procurement processes will undoubtedly produce systemic and sustainable business process improvement. But the bigger questions for procurement leaders to address in considering Six Sigma are these: What level of performance is realistic, which projects would produce the most benefit, and what enabling technologies will be required to apply the DMAIC principles (Define, Measure, Analyze, Improve, Control) of Six Sigma effectively?
When is a Six Sigma performance achievement attainable, or should the goal be to achieve some other level of acceptable, even superior performance?
As a statistical measurement of process variability, actually achieving Six Sigma level performance of only 3.4 defects per million opportunities might not be an appropriate goal for all procurement projects. Imagine for example, a procurement department that issues 100,000 purchase orders annually. Undertaking a project focused on measuring “perfect orders,” – accurate, properly approved and delivered POs which do not require re-work – would not allow for even one incorrect PO – a lofty goal but not necessarily an attainable one. On the other hand, achieving 93% accuracy – three sigma level – would actually exceed the current standard for world-class performance of 90% POs that do not require post-issuance rework, according to Hackett Group benchmarks.
What are some potential procurement projects that would benefit most from applying Six Sigma?
Regardless of the most appropriate level of process quality being targeted, any number of potential procurement projects could be considered. The guiding principle should be that in the first step of applying DMAIC – Define – the process being targeted is specific enough to be addressed within a realistic timeframe and has the potential to produce significant return on the organizational resources to be invested in the process improvement project. Some notable examples of procurement processes/projects, for which Six Sigma has been successfully applied, include the following:
Cycle time reduction – whether it’s requisition-to-order, invoice approval, contract creation, supplier on-boarding or sourcing cycle times, applying Six Sigma to fundamental procurement processes, to remove waste or non-value added activities by focusing on those items which are truly Critical To Quality (CTQs in Six Sigma speak), is a great place to start.
Transaction Quality – in addition to measuring POs that do not require rework, other projects to consider include:
- first-pass invoice match rates
- spend with line item visibility
- requisitions that are correctly coded and classified
- orders generated and delivered electronically
While achieving Six Sigma may not be the most appropriate standard to apply here either, just getting north of 90% across the board would signify world-class performance.
Technology Adoption – tracking adoption rates of both internal users, as well as suppliers, is a best practice for ensuring an ROI from technology investments. Sample projects include:
- Percentage internal user adoption of new tools (Note: expect much higher adoption rates for tools primarily used within procurement, such as Supplier Performance Measurement, as in the case of AMD, which reports 100% adoption by Supplier Managers, as compared to enterprise-wide deployments such as Procure-to-Pay where, as long as end-users can find another means of making a purchase (Expense Reports, P-Cards, Telephone, the Internet, Bitcoin, really the possibilities seem endless), at least a few probably will
- Percentage of suppliers e-enabled (suggest applying the 80/20 rule here – go after 20% of suppliers that account for 80% of spend and leave the long tail for another day)
- Percentage of total indirect RFQs through e-Sourcing (85% for indirect is world-class, although only about half as much for direct)
Compliance – recognizing that the goal of any process is a desired, or in fact, required output, particular focus on those deviations that result in non-compliance are especially important from a regulatory and risk standpoint, as well as a bottom-line perspective. Some areas that merit attention would include tracking:
- contracts issued, suppliers on-boarded, POs issued, or invoices paid, without proper approvals
- spend with non-preferred, or worse yet, blacklisted vendors
- RFQs not competitively bid or without including diverse suppliers in the bid process, if required
- contracts or supplier certificates expiring, lapsing or auto-renewing
Which enabling technologies will be required to effectively apply the DMAIC (Define, Measure, Analyze, Improve, Control) principles of Six Sigma?
Of course, the technology implementation may itself be the “I” of DMAIC – the Improvement which has been identified to address the process deficiency – so this may be a case of “Chicken or the Egg.” However, some minimal technology enablers are almost perquisites for undertaking this statistically driven methodology. Backbone financial systems such as ERP, will be a minimum requirement to generate transactional data, but may also need to be augmented by other tools if the data quality is suspect. Consider, for example, the extent to which the success of the Six Sigma initiative, which is predicated on the ability to measure and analyze processes, might require advanced spend analytics to cleanse and classify purchase transactions by supplier and category; supplier information management to remove duplicate or inactive suppliers from the supplier master, or capture missing profile and certifications; and contract management to aggregate enterprise-wide supplier agreements in a single, searchable repository.