Not every conference is worth the booth. 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. 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 works the way a hiring scorecard works: we agree on what matters, and on how much, before we meet anybody, because a charming candidate quietly bends the criteria. Events are charming.
The model I built and used at a deep tech company is called RISC, for Relevance, Impact, Strategic alignment, and Cost. Eight weighted criteria, each scored 1, 3, or 5, with weights that add up to 32. The weighted sum divided by 32 gives a score out of 5.
| Criterion | Weight | What earns a 5 |
|---|---|---|
| Industry and vertical fit | 7 | Explicitly a defense or manufacturing event |
| Audience fit | 5 | Decision-makers, on data, infrastructure, AI, or digital topics |
| Partner and ecosystem fit | 5 | Tier 1 partners such as AWS and NVIDIA present, or every key competitor attending |
| Networking and guerrilla access | 4 | Invite-only or curated 1:1s, strong investor presence |
| Startup opportunities | 3 | A formal, high-value startup program |
| Audience size | 3 | 40,000 attendees or more |
| Brand exposure | 3 | Major global press and analyst coverage |
| Cost and logistics | 2 | A local event with no flight or hotel |
Read the weights and you can read the strategy. Industry fit carries 7 of the 32 points and audience size carries 3, which says out loud that we'd rather be at a small defense show full of buyers than in a hall holding forty thousand people who can't sign anything. Copy the criteria if they fit, and set your own weights, because that argument is the real strategy work.
How do we stop a score from being a vibe?
By writing the bands down before we score anything. Audience size alone uses the full 1 to 5 scale: 5 means 40,000 attendees or more, 3 means 5,000, and 1 means under 1,000. Cost and logistics measures the trip, not the invoice: a local event scores 5, a short or free-lodging trip scores 3, and a long-haul flight plus hotel scores 1.
The sponsorship price stays out of the score on purpose. The score says whether the event is a high-value target. Only then do we ask whether the price is justified.
One thing to be precise about, since this is a framework post and not a results post. I designed RISC at a deep tech company and it was accepted internally, and it never ran a full cycle: the company was acquired 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 is what happened the last time we went. Before scoring we check our master event database, including events we once skipped. A returning event carries a second number, a 1, 3, or 5 the team assigns in the post-mortem, and the two get read together.
- High score, experience of 5. Accelerate. Repeat or expand the full presence.
- High score, experience of 1. Investigate. Don't rebook the same sponsorship; send a smaller team and find out what went wrong.
- Low score, experience of 5. Protect. A guerrilla sweet spot, so repeat the cheap plan and add no spend.
- Low score, experience of 1. Don't attend.
A first-timer has no experience score. At 4.0 or above we send a speaker and a guerrilla rep rather than buying a booth. From 3.0 to 3.9 it's a guerrilla pass only. Below 3.0 we skip it. A booth on a first visit needs a score above 4.5 with a 5 on both audience fit and partner fit, or an invitation funded by a delegation.
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 answers. For 2026 we allowed up to six full-vendor events and 6 to 10 of the cheaper two.
- Full vendor. Booth, a team of sales, an exec, and a technical lead, plus branding and lead capture. For must-attend events where audience and partner fit are both close to perfect.
- Speaker and guerrilla. No booth. A high-prestige speaking slot, plus one exec and one technical speaker taking 1:1 meetings. The goal is brand leadership and C-level conversations rather than volume.
- Guerrilla only. One pass and one sales or BD rep with a pre-booked schedule. The intel gatherer.
What does that third tier buy? A badge and a flight return competitor messaging and collateral, notes from partner sessions, 10 to 15 meetings, and an answer on whether the event deserves more money next year.
What does the plan look like for one event?
Scoring picks the events and the tier. The playbook's execution plan, built for a small team with no ad budget, runs in three phases, and the third is where most of the value sits.
- Pre-event, T-90 to T-1. The attendees post on LinkedIn at T-30, T-14, T-7, and T-1, with marketing drafting. The website gets one Events page and one call to action, not a landing page per event. Outreach starts three to four weeks out: an email on day 1, a value-add email on day 3, a LinkedIn note on day 5. The goal is 10 to 15 meetings booked before arrival. No attendee list? Hunt the sponsor list, the speaker list, and the event's LinkedIn page.
- On site. Ask, don't pitch. Open with "What brings you to the event?", connect on LinkedIn on the spot, log every lead that night, and post once a day. Swag follows the tier: none for a guerrilla pass, a QR code on the speaker's last slide, and notebooks for hot booth leads only.
- Post-event, T+1 to T+30. The playbook's rule is that a lead loses half its value after 48 hours. Consolidate every list the next day, sort it into hot, warm, and cold, and have whoever met a hot lead phone them within one or two days. At T+30 marketing reports cost per lead, cost per MQL, and pipeline against goals, and leadership spends 15 minutes assigning the final 1, 3, or 5.
Does any of this guarantee pipeline? Of course not. The model scores events, and it won't rescue one that nobody follows up on. The weighting logic sits under most go-to-market allocation calls, as in the GTM framework work case.
Key takeaways
- Write the weights down before the prospectus arrives.
- Define the bands, so a score is a measurement.
- Keep the sponsorship price out of the score.
- Guerrilla passes buy competitive intelligence for the price of a flight.
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.



