Skip to content

Portfolio segmentation for event groups that decides where capital goes next

Portfolio and M and AUpdated 2026-08-238 min read

In short

Portfolio segmentation for event groups sorts every show into invest, hold, fix or exit using two computed inputs: the contribution trend across the last three editions, and the health of the exhibitor base measured as retention plus replacement. Thresholds are published before the sort runs, so next year's classification reproduces without a fresh argument.

Every group I have seen runs the same budget round. Twenty four shows, twenty four business cases, and a discretionary pot that covers maybe a third of what was asked for. The cases that get funded are the ones argued best, which correlates with the seniority and confidence of the person arguing rather than with anything about the show.

Portfolio segmentation for event groups exists to break that correlation. Sort every show into invest, hold, fix or exit using inputs computed the same way for all of them, publish the thresholds before you run the sort, and the budget round becomes a conversation about the thresholds instead of a competition between show directors.

The sort is worth doing even if you disagree with where a specific show lands. A classification you argue with is more useful than a ranking nobody wrote down.

Two inputs, both computed from editions

The first input is contribution trend across the last three editions. Take contribution after direct costs for each of the last three editions, compute the per edition compound rate, and use that. Three editions damps out the single bad year caused by a venue move or a clashing date, and it is short enough that the number still describes the show as it is now.

The second is exhibitor base health, and it needs two components because retention alone hides the shape of the decline.

Retention is the count of paid exhibiting companies at the current edition that also exhibited at the previous one, divided by the previous edition's paid exhibitor count. Replacement is the count of new paid exhibitors divided by the count of exhibitors lost since the previous edition. A show at 0.68 retention and 1.2 replacement is healthy and churning. A show at 0.68 retention and 0.5 replacement is quietly emptying, and the two look identical on a revenue chart for about two editions.

Both inputs come from data you already hold, and both survive the change of a show director, which matters because the classification has to outlive the people who ran the shows when it was first built.

Where do the thresholds come from?

Somebody has to choose them, and the choice should happen before anyone knows which show lands where.

A workable starting set puts a show in invest when contribution trend is at or above plus 4 per cent per edition and base health passes, meaning retention at or above 0.62 and replacement at or above 1.0. Hold covers trend between minus 2 and plus 4 per cent with base health passing. Fix is any show where base health fails but trend is no worse than minus 8 per cent. Exit is trend below minus 8 per cent with base health failing.

Those numbers are a starting point rather than a standard, and I would expect a group to move them once after the first run. What matters more than the levels is that they are written down, dated, and applied to all twenty four shows in the same afternoon.

Listed groups publish the equivalent discipline. Informa's 2025 full year results define underlying revenue and underlying adjusted operating profit as results adjusted for acquisitions and disposals, the phasing of events including biennials, accounting standard and policy changes, and the effects of currency, with acquisitions included on a pro forma basis from the first day of ownership in the comparative period. Every one of those adjustments is arguable, and the value of the definition comes from it being fixed in advance and applied to a portfolio the company describes as more than 800 specialist brands across more than 30 market categories. A private group with 24 shows has less to reconcile and no excuse for not writing its rules down.

The 24 show portfolio, sorted

Total contribution across the 24 shows is 31.6 million. The six largest produce 6.4, 4.9, 3.8, 3.1, 2.6 and 1.6, which is 22.4 million between them, or 71 per cent of the total. The remaining 18 shows share 9.2 million, averaging 511,000 each.

Run the sort and the boxes come out at five in invest, nine in hold, six in fix and four in exit.

The distribution of contribution across those boxes is where it gets interesting. The five invest shows carry 14.8 million. The nine hold shows carry 12.3 million. The six fix shows carry 3.4 million. The four exit shows carry 1.1 million, which is 3.5 per cent of group contribution.

Now count the people. Show facing capacity across the group is 34.4 full time equivalents. The four exit shows consume 5.2 of them and the six fix shows consume 8.9, so 14.1 of 34.4, which is 41 per cent of the group's show facing capacity, works on 4.5 million of contribution, or 14.2 per cent of the total. That single ratio is usually the finding that makes a board act, and it is invisible in any report organised by show.

Note also what the sort does to the biggest shows. Two of the six largest land in hold rather than invest, because their contribution trend is positive but under 4 per cent. That is the right answer and it is unpopular, because the shows that carry the group tend to attract capital by gravity. A hold classification for a 4.9 million show means maintenance funding and no growth project, which frees the money for a 1.4 million show growing at 11 per cent per edition with a replacement ratio of 1.4.

What the sort does to the capital allocation

Set the split as policy before the sort runs. Take a discretionary pot of 3.0 million. Give invest 70 per cent, which is 2.1 million across five shows. Give fix 25 per cent, 750,000 across six shows, released against a dated milestone rather than at the start of the year. Hold gets maintenance only. Exit gets nothing beyond running the final edition properly, because a badly run last edition damages the brand you are trying to sell or fold. The remaining 5 per cent, 150,000, stays unallocated until the half year, which is the only honest way to leave room for something nobody forecast.

The dated milestone on the fix pot is the part that gets dropped and should not be. A fix show receives its money on condition that its replacement ratio reaches 1.0 by a named edition. If it does not, the show moves to exit at the next sort, automatically, without a fresh debate. Writing that consequence into the allocation at the moment the money is granted is far easier than arguing it eighteen months later.

The exit box then hands off to a separate decision about form and timing, since closing outright is only one of the ways out, and that sits with the closure rule and its alternatives.

Why is the founding show almost always in the wrong box?

Because it is old, and old shows in event portfolios have low contribution trends and long serving exhibitor bases that churn slowly, so they tend to land in hold or fix on the arithmetic while carrying disproportionate weight in how the group sees itself.

Feldman studied exactly this in Organization Science in 2014, looking at legacy divestitures, meaning the sale or spinoff of a company's original business. The finding runs against intuition. Firms that divested their legacy business had weaker post divestiture operating performance than comparable firms that retained theirs, and the effect was strongest where the divested unit operated in the same industry as others of the divesting firm's businesses. Feldman also found that newer chief executives were more likely to undertake legacy divestitures, with the most recently appointed undertaking the most costly ones.

Read across to an event group, that says two things. The founding show is probably more entangled with the rest of the portfolio than the classification can see, through shared exhibitors, shared sales relationships and the credibility it gives the group in that sector. And a proposal to sell it is more likely to arrive in the first two years of a new chief executive's tenure than at any other time, which is worth knowing about the proposal.

So the rule I apply is that the sort classifies the founding show like any other, and the exit classification for it requires an explicit written case covering exhibitor overlap with the rest of the group. Not a veto. A higher evidential bar, applied to one named show, for a reason published in the policy.

Where this stops

The sort is only as good as the contribution figure underneath it, and contribution is where allocation policy hides.

If central costs are spread by revenue, large shows carry a heavier load than they cause and small ones carry a lighter one, and the sort will systematically favour small shows. If central costs are excluded entirely, the sort will favour shows that consume enormous amounts of central attention for modest direct cost. Neither error is visible in the classification itself, which is why the allocation basis belongs in the published thresholds document alongside everything else. Building the underlying view of how shows sit against their sectors is the job of the two axis portfolio map, and the sort inherits whatever that map got wrong.

The second limit is that four boxes cannot hold a show whose problem is timing rather than performance. A show that would be healthy on a different date, or biennially, or co-located with a larger event in the group, fails the same tests as a show with no future and lands in the same box. The classification is a filter that tells you which shows need a decision, and the decision itself still needs a person who knows the sector.

The sort also says nothing about shows that do not exist yet, and a group that prunes without launching shrinks. Pricing those belongs to the launch pipeline valuation and sits outside this exercise entirely.

This week, compute the replacement ratio for your five smallest shows. New paid exhibitors divided by lost paid exhibitors, from the last two editions, nothing else. Any show below 0.7 is emptying regardless of what its revenue line says, and that list is the honest starting point for a first pass at the portfolio and synergy picture.

Questions people ask about portfolio segmentation for event groups

What inputs should decide how an event portfolio is segmented?
Two are enough if both are computed the same way every year. Contribution trend across the last three editions gives the direction of travel with the noise of a single bad edition damped out. Exhibitor base health, measured as paid exhibitor retention combined with a replacement ratio of new logos to lost logos, gives the underlying condition.
Why publish segmentation thresholds in advance?
Because a classification produced after the numbers are known is an argument dressed as analysis. Fixed thresholds mean the sort reproduces next year, the shows that move between boxes moved for a measurable reason, and anyone contesting a classification has to contest the threshold rather than the outcome for their own event.
How much capital should each segment receive?
Set the split as policy, not per show. A workable starting point gives the invest group the large majority of discretionary capital, gives the fix group a smaller pot released against a dated milestone, funds the hold group at maintenance only, and gives the exit group nothing beyond the cost of running the last edition well.

Related reading

All portfolio and m&a articles