Not every conference is worth the booth. That's easy to say and hard to act on, because the case for any given event arrives as a prospectus with an early-bird deadline attached, while the case against it arrives eight months later as a number nobody wants to compute. So teams book on reputation, rebook out of habit, and eventually decide that events don't work for them.
Events do work, and what doesn't work is choosing them one at a time, under deadline pressure, with no written definition of a good one. Forrester's Q1 2025 State of B2B Events Survey, written up by Conrad Mills and Hannah Jachim, found that more than 90% of organizations are focused on getting the right audience to their events, demonstrating ROI, and improving post-event follow-up, while budgets stayed flat or fell for two-thirds of teams. Everyone wants the same 3 things out of a smaller envelope, and the teams that get them decided months earlier which events could deliver.
So what are we buying when we sign a sponsorship, and how do we tell in advance? Here's a framework you can copy, argue with, or bolt onto the one you run now.
Score the event before anyone gets attached to it
A weighted scorecard for events works the way a hiring scorecard works. We agree on what matters, and on how much each thing matters, before we meet anybody at all, because once you're in the room with a charming candidate the criteria quietly bend to fit them. Events are charming. A famous conference with the wrong audience beats a dull one with the right audience in any conversation not governed by weights written down in advance.
The model we built at Visual Layer is called RISC, for relevance, impact, size, and cost. Fourteen criteria, each scored 1 to 5, each carrying a weight that says how much we care.
| Group | Criterion | Weight |
|---|---|---|
| Relevance | Audience fit | 5 |
| Relevance | Content quality | 5 |
| Relevance | Major brands on stage | 3 |
| Relevance | Competitor attendance | 3 |
| Relevance | Industry reputation | 3 |
| Impact | Networking potential | 5 |
| Impact | Engagement opportunities | 5 |
| Impact | Innovation showcase | 4 |
| Impact | Timing | 3 |
| Size | Brand exposure | 4 |
| Size | Audience size | 3 |
| Cost | Sponsorship benefits | 4 |
| Cost | Attendance cost | 3 |
| Cost | Logistics and venue | 2 |
Read the weights and you can read the strategy. Audience fit, content quality, networking, and engagement all sit at 5 while audience size sits at 3, which says out loud that we'd rather be in a small room full of buyers than a hall holding forty thousand people who can't sign anything. Copy the criteria if they fit. Set your own weights, because that argument is the real strategy work, and it should happen months before any prospectus lands.
How do we stop a score from being a vibe?
By writing the bands down before we score anything. Audience size is the easy one: 5 means 100,000 attendees or more, 4 means 20,000 and up, 3 means 5,000, 2 means 1,000, and anything smaller scores 1. Brand exposure runs the same way, 5 for global press and trade coverage, 1 for a closed-door event nobody reports on.
Cost is the interesting one. Teams argue about it most and define it least, so we scored it from a geography baseline, then took a point off each for an expensive host city, a restricted or secure-access venue, and a very small audience, with a floor of 1. Direction matters here: a high score means cost isn't a problem, so a cheap local event scores well and a long-haul week in San Francisco doesn't.
One thing to be precise about, since this is a framework post and not a results post. We designed RISC at Visual Layer and it was accepted internally, and it never ran a full cycle: Camtek acquired the company in April 2026 first. More than 65 events were evaluated as candidates across roughly 12 months, so the pipeline and the scoring were real work, but the outcomes are not something I can show you.
Why does one score never settle it?
Because a score is a forecast, and the cheapest correction available is what happened the last time we went. So a returning event carries a second number too, a 1, 3, or 5 the team assigns in the post-mortem, and the two get read together.
- High potential, proven winner. Accelerate. Buy the bigger presence.
- High potential, bad experience. Investigate, and downgrade the tier.
- Low potential, great experience. A guerrilla sweet spot, the cheapest ROI you own.
- First-timer, no history. Even a top score starts a tier lower. Test before buying.
Let the score choose the size of the bet
A high score doesn't mean we book the booth. It means the event deserves a commitment, and the size of that commitment is a separate decision with three sensible answers.
- Full vendor. Booth, full team, pre-booked meetings. For top scorers, and an outcome you can name.
- Speaker and guerrilla. No booth. A submitted talk, a side dinner, meetings around the venue. Where most deep tech teams should spend.
- Guerrilla only. One person, one pass, a plan. The intel gatherer.
So what does that third tier actually buy? More than the price suggests, because a badge and a flight are the whole cost. It returns competitor booth messaging, session topics, the questions buyers ask from the floor, and a list of partners worth a call in January.
What does the plan look like for one event?
Scoring tells you which events and at what tier. It doesn't tell you what to do once you've committed, so the playbook we built at Visual Layer to set the following year's engagements, budget, and priorities carries an execution plan too. Small team, tight budget, and a hall full of louder companies. It runs in three phases, and the third is where most of the value sits.
- Pre-event, T-90 to T-1. Listen on LinkedIn, Reddit, Substack, and Luma for who's already talking about the event. Send any invite-only dinner invitations early, because senior calendars fill fast. Pre-book demos, and get your founders onto podcasts while there's still time to matter.
- On site. Put the team in something visible across the hall and work the floor rather than standing behind the booth. Qualify before the expensive swag, and book the next meeting before they leave the stand.
- Post-event, T+1 to T+30. Write the follow-up templates before anybody boards a plane, answer hot leads the same day using someone who stayed back at the office, then run a post-mortem that sets next year's score.
Then comes the arithmetic, which is the part teams skip: add up the true cost, including booth, travel, swag, the dinner, and staff days, then divide it by the meetings you can realistically hold. That's your cost per meeting. It tells you whether the tier you bought was right, and it only works if you tag every lead to the event, hold the attribution for a year, and compare closed contract value against total cost before you rebook.
Does any of this guarantee pipeline? Of course not. And the model has a limit worth stating plainly: it scores events, it doesn't design your booth, and it won't rescue an event that nobody follows up on. Speed in that first week is a separate muscle. Forrester's survey suggests it's the one most teams already know they're missing.
The same weighting logic sits under most go-to-market allocation calls, so it'll look familiar next to the GTM framework work case.
Key takeaways
- Write the weights down before the prospectus arrives.
- Define the 1 to 5 bands, so a score is a measurement.
- Guerrilla passes buy competitive intelligence for the price of a flight.
- Cost per meeting and a year of attribution decide whether you go back.
If you're staring at a shortlist right now and the deadline is closer than the consensus, bring it to a free 30-minute call. We'll score three of them together, write the weights down, and you'll leave with a model your team can run without you.
