Benchmarking Without Lying to Yourself: A Practical Method
Benchmarking has a bad reputation, and it earned it. Too many organizations 'benchmark' by finding a number that makes them look good, framing it on a slide, and declaring victory. The pandemic recovery made this temptation worse: with so many baselines disrupted, it became easy to compare this quarter's mess against last year's worse mess and call it progress. Honest benchmarking does the opposite. It compares comparable things, fairly, to learn where you actually stand and where the gap worth closing really is.
In Lean Six Sigma terms, benchmarking is part of how you understand the voice of the process and set a credible target. Done well, it sharpens a project charter and grounds the Measure and Analyze phases. Done dishonestly, it produces confident targets built on sand.
Decide what you are actually comparing
The first discipline is choosing the right type of benchmark. Each answers a different question, and confusing them is where most self-deception starts.
Internal benchmarking. Compare similar processes across your own sites or teams. It is the cleanest data you will get because definitions and context are mostly shared — start here.
Competitive benchmarking. Compare against direct competitors on the same metric. Useful but hard, because you rarely see their true internal numbers, only what they choose to publish.
Functional benchmarking. Compare a specific function — say, order fulfilment — against the best performers in any industry. A hospital learning intake from a hotel is a classic example.
Generic benchmarking. Compare a broad process pattern, like complaint handling, irrespective of sector. Lowest comparability, but it can break loose fixed thinking.
Make the comparison fair before you trust it
A benchmark is only as honest as the normalization behind it. Two organizations can report 'cycle time' and mean wildly different things — one starts the clock at order receipt, the other at production release. Before you compare a single number, agree on the operational definition, the unit, the scope, and the conditions.
Match the operational definition: where does the measurement start and stop, exactly?
Match the unit and the denominator: per order, per line item, per customer — pick one and hold it.
Match the conditions: peak versus average, in-scope exclusions, and how rework or returns are counted.
Match the time window: comparing a stable quarter against a disrupted one tells you about the disruption, not the process.
Resist the urge to exclude your worst data 'because it was unusual.' If unusual events are part of how your process actually behaves, they belong in the picture. Cherry-picking the favourable window is the most common way a benchmark quietly becomes a lie.
Turn the gap into a target, then into a project
Once the comparison is fair, the gap between you and the benchmark becomes a candidate for improvement — but not automatically a goal. A best-in-class performer may operate at a scale, or with constraints, you cannot replicate. The honest move is to ask what is genuinely transferable: which practices, not just which numbers, produced their result. Use the gap to set a stretching but defensible target for your DMAIC project, and document the assumptions so a sceptical reviewer can follow your reasoning. A benchmark you cannot defend in a room is not an asset; it is a liability waiting to be exposed.
Done with discipline, benchmarking stops being a slide-decoration exercise and becomes what it should be: an honest mirror that tells you where the real opportunity is, and a credible anchor for the targets you are about to commit to.
Choosing the right comparisons and turning them into defensible improvement targets is rarely a solo exercise. XNM's strategic advisory helps leaders benchmark honestly and act on what they find.