The counting trap
Partnership programmes drift towards counting activity because activity is available, unambiguous and flattering. The number of signed partners goes up. The number of joint sessions goes up. Neither number tells anyone whether the arrangement produced anything, and both can be improved by doing more of what is not working.
The corrective is not to abandon counting but to be clear about what each number is for. A number used to steer is a different thing from a number used to judge, and confusing the two is what makes measurement political.
Leading and lagging indicators
A lagging indicator confirms an outcome after the fact: revenue recognised through the arrangement, work delivered, cost avoided. It is the number that matters, and it arrives too late to act on.
A leading indicator moves earlier and correlates with the outcome: how many of a partner's staff can actually explain the product; how many joint opportunities are live; how long it takes each side to respond to the other. These let a programme be steered, and they share a weakness — they are easy to inflate, because they measure inputs that someone controls directly.
The practical rule is to steer on leading indicators, judge on lagging ones, and never let the same person be measured solely on a number they can produce at will.
The attribution problem
When two parties both contributed to a result, deciding whose it was is not a technical problem with a technical answer. It is a negotiation, and pretending otherwise causes trouble.
A buyer heard about a product from one party, was persuaded by material the other produced, and bought through a third route. Each party's system will record this differently, each in its own favour, and each will believe its own records. The dispute that follows is usually about money, and it is unwinnable on the evidence because the evidence was never designed to answer the question.
Arrangements that handle this well do two things. They agree the rule in advance — which recorded event counts, and whose system holds it — accepting that any rule will be wrong in individual cases. And they keep a separate, deliberately rough measure of joint influence that nobody is paid on, so that the parties can talk about what is really happening without every conversation being a claim.
Counting the cost side
Partnerships are usually measured on what they produce and rarely on what they consume. The consumption is real: the time of senior people in reviews, the effort of the operating team, the support given to a partner's staff, the internal cost of building whatever the arrangement needed.
A partnership that generates modest results at trivial cost may be a better use of attention than one that generates larger results at enormous cost, and no programme can see that difference unless someone bothers to hold a rough figure for effort. It need not be precise. It needs to exist.
The measures that do not become numbers
Some of the most informative signals are qualitative and should be recorded in words. Does the partner bring problems early, or only after they are unavoidable? Do their people describe the joint work accurately to buyers? When something goes wrong, does either side reach for the contract first? Would each side enter this arrangement again knowing what it now knows?
Judgements like these predict the second and third year of a relationship far better than the first year's numbers. They should be written down, briefly, at each review, precisely so they cannot be reinvented afterwards to fit whatever the parties have decided to do.
The honest review
A measurement system's real purpose is to make one decision possible: continue, change or stop. If the numbers collected could not, in any realistic circumstance, produce the answer "stop", then they are not measurement. They are reporting.
Deciding in advance what result would justify ending an arrangement, and by when, is the single most useful discipline available. It is also the least popular, because it turns an open-ended relationship into something that can visibly fail.
Where this leads
Most partnerships end without a formal decision. The patterns behind that are consistent enough to be worth studying directly.