Like many firms that are serious about manufacturing, our firm is a longtime member of APICS, The Association for Operations Management, participating regularly in meetings, training classes and even serving on their various local boards when asked. We believe APICS makes manufacturing people smarter and more knowledgeable in their jobs. We are also regular subscribers to their excellent bimonthly magazine, which this issue highlighted some good points regarding inventory KPIs, and which we’ll reprise here.
Key Performance Indicators, or KPIs, were part of the original “balanced scorecard” model developed in the 1990s. The approach begins with the answer to a high-level business question: To achieve your company’s vision, what should success look like in the areas of finance, customer experience, business process, and learning and growth? Then, once those questions have been answered, ideally by a team of managers and staff, the organization develops goals for each, including specific objectives, targets and measures — KPIs.
In inventory, for instance, KPIs typically fall into three broad categories:
- Raw facts and figures: like fill rate, turns, stockouts
- Percent complete: applied to goals with a clearly defined finish state
- Trend line scores: or directional scores, related to projects or goals without a defined finish state. These move up and down – things like production efficiency, productivity and safety stock.
Classic inventory KPIs are inventory turns (the number of times per year on average that you turn over, or sell, your inventory)… inventory aging… average days on hand… excess inventory average… month-end inventory value… average carrying cost per unit of inventory, and so on.
Users settle on sources for their raw data, and then define, track and refine KPIs as needed, looking to squeeze out waste and redundancy, identify laggards, and create gradual improvement through incremental change in improved KPIs.
Sometimes, settling on standards or a common set of facts can be difficult if some players lack contextual understanding of the issues, leading to potential ‘apples to oranges’ analysis. Ideally, these are avoided when stakeholders seek out a single, objective truth, not one depending on one’s department or perspective.
The more directly measured a KPI is, the easier this is. It gets much harder when KPIs need to be calculated or derived from models.
But developing even good KPIs in isolation is not enough. “KPIs must combine and serve ongoing strategic business needs” as Jonathan Thatcher points out. That’s where KPIs start to lose some of their simplicity, a topic we’ll review in our next post.