Designing a priority points formula that rewards the behaviour you actually want

Your category director comes back from a renewal meeting annoyed. One of her accounts has gone from 200 square feet to 1,000 in four editions, brings buyers with them, and picks 190th. Another account has shrunk in three of the last four editions, pays late every year, and picks 41st.

She wants to know who decided that. Nobody decided it. Priority points formula design happened once, years ago, in an afternoon. The formula was correct in the sense that it did what it said, and it has been quietly ranking loyalty above growth ever since.

Most points formulas pay for size and tenure. Both of those are measurements of how much space an exhibitor took in the past, which makes them measurements of inertia, and inertia is the one behaviour you never need to pay for.

Every clause has a price and a half life

Two properties describe any clause in a points formula and most organisers have only ever thought about one of them.

The price is what a point costs the exhibitor in that currency. If your formula gives one point per 100 square feet and your rate is $32 a foot, a point costs the exhibitor $3,200 of contracted space. If it gives one point per $1,000 of sponsorship, a point costs $1,000. Converting every clause to a common currency is the first audit anyone should run on a formula, and it usually finds that the cheapest point is not the behaviour the organiser most wants.

The half life is how long the point keeps paying. In almost every published formula the answer is forever. A point earned in 2013 has exactly the same weight in the 2027 standings as a point earned in 2026, which means the formula ranks histories, and the exhibitor standing in front of you inherits a position earned by decisions made a decade ago.

Forever is a choice, and it is the choice that produces the situation your category director is annoyed about. It is also the easiest clause to change, because unlike the price of a point it does not require you to argue about what sponsorship is worth. Left alone, it is what freezes the front of your floor for a decade.

What does decay do to the standings?

Take a formula of the kind most organisers actually run: one point per 100 square feet per edition, plus two points for each consecutive edition exhibited.

Your incumbent has taken 400 square feet for twelve consecutive editions. That is four points for the space plus two for continuity, so six a year, and twelve years of it is 72 points.

Your riser has four editions behind them and has gone 200, 400, 800 and 1,000 square feet. Their yearly totals are 4, 6, 10 and 12, which sums to 32. The gap is 40 points and under the current formula it closes slowly or never.

Now add one clause. Points from the five most recent editions count in full. Points from any edition older than that halve for each additional year.

The riser is unaffected. All four of their editions sit inside the window, so they still hold 32.

The incumbent keeps their last five editions in full, which is five times six, or 30 points. Their earlier seven editions are one to seven years beyond the window, so those points are multiplied by 0.5, 0.25, 0.125, 0.0625, 0.03125, 0.015625 and 0.0078125. Those multipliers sum to 0.992, and six points times 0.992 is 5.95. The incumbent's total is 30 plus 5.95, which is 35.95.

The gap has gone from 40 points to 4.

Look at what that clause did and did not do. It did not punish the incumbent for being large, because their recent editions still count in full and they are still ahead. It removed the value of history the incumbent is no longer earning. Twelve years of 400 square feet is now worth about 36 points and four years of aggressive growth is worth 32, which is roughly the ranking anyone in the room would have produced by hand.

The cheapest point you will ever award

The second clause worth adding costs you nothing and pulls your cash flow forward by weeks.

Award one point for every edition booked before the published deadline. A size point costs the exhibitor $3,200 of contracted space in the example above. An early booking point costs them a signature four weeks earlier than they would otherwise have given it, which is worth real money to you in forecasting confidence and nothing at all to them in cash.

Put it into the standings above. If the riser books early in all four of their editions and the incumbent books late, as large incumbents usually do because they know their position is safe, the riser gains four and the incumbent gains none. The riser is now at 36 and the incumbent at 35.95, and the tie break ladder decides it.

The obvious objection is right. A point that is cheap for the exhibitor is a point every exhibitor will take, and once everybody books early the clause stops discriminating between them. That is a real failure mode and it has a fix. Make the point conditional on the deposit clearing, not on the signature, which is the part exhibitors actually find inconvenient. Or cap the accumulated early booking points at, say, five, so the clause pulls new behaviour out of a lapsed booker and stops paying an exhibitor who was always going to be early.

Either way you should want the clause to stop discriminating eventually. If in three years every exhibitor books before the deadline, the formula has done its job and you can retire the clause. The deadline in question is usually the selection appointment itself, which is why the draw and the formula have to be designed together.

Simulate the new standings before you publish anything

The change above looks small on two exhibitors. Run it on 420 and you find out what you have actually done.

Rebuild the standings twice from the same contract history, once under the current formula and once under the proposed one, and produce a table with both ranks side by side and the difference. Then look at three specific things: how many of the current top 30 leave it, how far the largest single fall is, and where in the order the churn concentrates.

The result is usually counterintuitive in a helpful way. Decay does not reshuffle the top of the queue, because if the median tenure in your top 30 is eleven editions then decay hits all thirty of them by a similar proportion and their ranks relative to each other barely move. The movement happens at the boundary, somewhere between positions 30 and 80, where a four edition riser crosses a twelve edition incumbent who has been flat. That is exactly the part of the floor where you want movement and exactly the part where the phone calls are survivable.

If your simulation instead shows your top ten reordering, your decay rate is too aggressive and you should widen the full value window from five editions to seven before you show anybody.

What are your exhibitors already telling you about cost?

A formula that pays mostly for size is a formula asking for more spend from a group whose spend is under review.

Explori published Exhibit Leader Insights in 2023 with Exhibitor Group and The Exhibitor Advocate, drawn from 255 exhibit marketers running programmes at organisations of a range of sizes. Of those planning to exhibit at fewer in-person shows, 82 per cent said exhibit related costs were a factor in the decision. Ninety four per cent expected some level of change to their exhibit programme within three years. Thirty per cent of senior leaders were described as no longer convinced that exhibitions are essential.

Those numbers change what a points formula should be buying. Paying only for square feet asks an exhibit marketer to justify a larger stand to a leadership team that is already sceptical, in return for a queue position that pays off in a decade. Paying for an early commitment, a completed post show survey, a hotel booking inside the block, or a category referral asks for behaviours that cost the exhibit marketer very little to defend internally.

One large organiser already prices absence without calling it decay. IAAPA added a reset to its Expo seniority rules from the 2022 Expos: a company that elects not to exhibit for three consecutive years has its exhibiting years history revert to zero. Seniority there is otherwise determined by membership status, then exhibiting years, then consecutive years of membership, and finally receipt of contract date and receipt of deposit. That is a blunt instrument beside a half life, because it does nothing until the third missed edition and then takes everything at once. It settles the principle worth borrowing. Tenure can be made perishable, and exhibitors accept it when the rule is published well in advance of the edition it bites on.

Where this stops

You cannot simulate the behavioural response, and the behavioural response is the entire point of the exercise.

The standings you produce from a rebuilt formula assume every exhibitor books exactly as they did before. They will not. A decay clause tells a coasting incumbent that their position is now perishable, and some of them will respond by taking more space, which is the outcome you wanted, and some will respond by treating the show as a place they no longer have status and leaving. Nothing in your historical data distinguishes those two groups in advance.

The second limit is political and it is the reason most of these formulas never change. Every exhibitor who loses rank is by construction one of your longest tenured accounts, and they will hear about the change from a competitor before they hear it from you. Publish the new formula at least two editions before it takes effect, publish both sets of standings during the transition, and accept that the first year will be spent explaining it.

There is also a data limit. Decay needs points attributed to the edition that earned them, and a large number of organisers hold only a running total in a single field. If your points column is one number with no history behind it, you cannot apply a half life to it, and rebuilding the history from old contracts is the actual first task. An edition by edition points ledger is the same object exhibitor analytics needs for almost everything else, so the work is not wasted if you abandon the decay clause.

One more thing the formula cannot do is decide how long each exhibitor gets in front of the map. That is set by the appointment grid, and a formula that promotes forty new names into the middle of the order without lengthening their slots will produce a slower draw and a worse floor.

Start there this week. Pull the contract file for every edition you still hold, recompute each exhibitor's points edition by edition, and check that the sum matches the total you have been publishing. Where it does not match, you have found the years somebody adjusted by hand, and those are the years your formula was already being overridden.