Why event ROI arguments fail
Every event programme eventually faces the same question from finance: what did we get for that?
The usual answer is a meeting count and some pipeline attributed with a generous hand. It rarely survives scrutiny, and the reason is almost never the arithmetic. It is that the data was never captured in a form that supports the claim.
Three failures recur:
- The denominator is incomplete. The event budget is counted; the cost of the people is not.
- The numerator is unverifiable. Pipeline is attributed to the event because the timing lines up, not because anything links a specific meeting to a specific opportunity.
- The record is reconstructed. Six weeks later, someone assembles a spreadsheet from memory and calendar entries. It is an approximation being presented as a measurement.
All three are fixable, and mostly by deciding in advance what to capture rather than by analysing harder afterwards.
Count the real cost
Most event ROI models understate cost badly, which paradoxically weakens the case — because a suspiciously flattering number invites scepticism about everything else.
A complete picture includes:
Direct event costs. Sponsorship or exhibition fees, stand build, meeting space rental, materials, catering.
Travel and accommodation. For everyone attending, not only the field team.
People time. This is the line most often omitted and frequently the largest. Eighteen people for four days is seventy-two person-days. At a fully loaded cost, that is usually a substantial figure — often comparable to the sponsorship.
Preparation and coordination. The weeks of scheduling work before the event, and the follow-up work after. If one person spends three weeks coordinating, that belongs in the model.
Opportunity cost. What those experts would otherwise have been doing. This is harder to quantify and reasonable people leave it out — but it is worth acknowledging that it exists.
Including people time changes the conversation productively. It reframes the question from "was the sponsorship worth it" to "was this the best use of seventy-two expert-days", which is the question actually worth asking.
Three layers of value
Not all value from event meetings is pipeline, and forcing it all into one number produces claims that do not hold up.
Layer one: pipeline influence. Meetings with prospects that connect to opportunities — new opportunities created, existing ones advanced, deals where the meeting appears in the account history. This is the most scrutinised layer and needs the most rigorous attribution.
Layer two: retention and expansion. Meetings with existing customers. Renewal conversations held, expansion discussions opened, escalations resolved face to face. Frequently the highest-value layer and almost always the most under-counted, because it does not look like new pipeline.
Layer three: market intelligence. What you learned. Which topics dominated, which objections recurred, what competitors were doing, which segments engaged. Genuinely valuable and genuinely not quantifiable — report it separately and qualitatively rather than trying to assign it a number.
Reporting all three separately is more credible than blending them, and it prevents the common distortion where a programme is judged solely on new logos while quietly doing most of its work on retention.
Getting attribution right enough
Perfect attribution is not available and chasing it wastes effort. Defensible attribution is available, and it requires one thing: a durable link between a meeting and an account.
Practically:
- Record the account for every meeting, at booking time, not afterwards.
- Record the participants, both sides.
- Record the topic and the agreed next step, within minutes of the meeting ending.
- Push that record into whatever system holds your opportunities, so the meeting appears in the account history rather than in a separate document.
With that in place, attribution becomes a query rather than an argument. You can state plainly: these forty-one accounts had a meeting; here is what happened to those accounts in the following two quarters, against a comparable set that did not.
Be transparent about the model you use. Whether you count influenced pipeline, first-touch, or a windowed post-event lift matters less than stating which one you chose and applying it consistently. A modest number that is clearly defined survives challenge; a large number with an unclear method does not.
Leading indicators worth tracking
Pipeline outcomes take quarters. These are available within days and predict the later numbers well enough to steer on:
Coverage. What share of your target accounts did you actually meet? This is the cleanest measure of whether the programme did its job.
Completion rate. Meetings held as a proportion of meetings booked. Below about 85% suggests a problem in reminders or scheduling quality.
Match rate. What share of meetings had an expert whose specialism matched the requested topic? This one is rarely tracked and correlates strongly with whether meetings produce anything.
Next-step rate. What share ended with a specific, owned next action? Meetings without one convert far less often.
Expert utilisation. Meetings per expert per day against capacity. Tells you whether to bring more people or fewer next time.
Topic distribution. What attendees actually wanted to discuss. The most actionable data the programme produces, and the input to who you staff next time.
Two or three events of this data and you can forecast an event's outcome from its leading indicators, which is far more useful than judging it retrospectively.
Putting it together
A credible event meeting ROI report has five parts:
- Total cost, including people time, stated openly.
- Coverage: which accounts were met, against which were targeted.
- The three value layers, reported separately — pipeline, retention, intelligence.
- Leading indicators, with comparison to previous events.
- What to change, derived from topic distribution and utilisation.
The fifth part is what turns a defensive exercise into a useful one. A report that concludes "we were under-staffed on security expertise and over-booked on room three" is worth more to next year's programme than any single ROI figure.
The prerequisite for all of it is capturing the data as the event happens rather than reconstructing it afterwards. That is a scheduling-system decision made months earlier, not an analysis decision made in the week the report is due.
CallShark is being built so this data is a by-product of running the programme rather than a separate exercise — accounts, topics, outcomes, and utilisation captured at the point of scheduling. The product is temporarily down.