On this page13
- Why cost per qualified meeting beats every other conference metric
- What counts as a qualified meeting?
- Which conversations don't count?
- How do you work out the cost side honestly?
- The formula, step by step
- Worked example: a three-person team at a big crypto week
- How do you log meetings during the event without losing them?
- When is cost per meeting the wrong lens?
- FAQ
- How many qualified meetings should a conference produce?
- Should meetings booked before the event count if they could have happened on a video call?
- How long should I wait before judging a conference's ROI?
- Do I need a CRM to run this model?
To calculate conference ROI, add up the full cost of the trip, divide it by the number of qualified meetings you held, then track how many of those meetings reached a second step within 30 days. A qualified meeting is a conversation with someone on your pre-written target list, who can act on what you discussed, that ends with a dated next step. Everything else is networking. It has value, but it isn't the number that justifies the spend.
This is the model I walk founders through before they book flights. It takes about an hour to set up and saves a lot of arguing after the event about whether the week was "worth it".
Why cost per qualified meeting beats every other conference metric
Most teams come home with three numbers: badge scans, new followers and a general feeling that the week went well. None of them connect to an integration, a cheque, a hire or a story.
I've been asking about this for a while:
to the teams spending shit ton of money on events, how do you assess the ROI on events? Shilika Jain on X, 14 Nov 2024
The answers I get, in public and in DMs, mostly describe vibes. Cost per qualified meeting isn't a perfect measure of value, but it has three things the alternatives lack. You can count it on the day. You can compare it across events, so you can rank your conference calendar. And it forces you to decide who you're there to meet before you spend anything.
An older complaint of mine applies here too:
classic problem of web3 startups: don’t understand CAC, LTV, turnover or other basic metrics! Shilika Jain on X, 8 Dec 2022
A conference is an acquisition channel. Treat its output the way you'd treat a paid campaign: units bought, cost per unit, conversion to the next stage. The PR ROI method I use for coverage works on the same logic.
What counts as a qualified meeting?
I use four tests. A conversation counts only if it passes all four.
- Named in advance. The person or company was on your target list before you landed, or fits a category you wrote down in advance, such as "wallet teams building on our chain" or "reporters who cover stablecoin regulation".
- Able to act. They make, or strongly shape, the decision you need. That means signing off an integration, writing a cheque, running a story or accepting an offer.
- Substantive. You discussed the specific thing, not just swapped handles. Ten focused minutes is a fair floor.
- Dated next step. A call booked, a document promised by a day, an introduction with a name attached.
Test four does most of the filtering. "Let's stay in touch" fails it. "Send me the integration docs and I'll loop in our engineer next Tuesday" passes.
Which conversations don't count?
This is where teams inflate the number, usually without meaning to. A quick reference:
| Interaction | Counts as qualified | Reason |
|---|---|---|
| Pre-booked 30 minutes with a partner team's BD lead, ending with a technical call date | Yes | Passes all four tests |
| Dinner chat with an investor on your list who asks for the deck by Friday | Yes | Decision power plus a dated ask |
| Coffee with a reporter on your beat who agrees to a briefing after your announcement | Yes | Press counts when there's a next step |
| Someone off your list who promises a named intro to a target by a set date | Yes, as an introduction | Log it in its own column so it doesn't inflate meetings |
| Badge scan at your booth | No | No conversation, no next step |
| Great talk with a founder outside your category, no ask either way | No | Networking, log separately |
| Photo with a well-known KOL | No | Content, not pipeline |
| Investor on your list who says "keep me posted" | Not yet | Logged, but upgraded only when a date appears |
Keep the networking column. Some of those people will matter in a year. Just don't divide your costs by them.
How do you work out the cost side honestly?
I won't give you price figures here, because ticket tiers, hotel rates and sponsorship packages change by event and by year. What I can give you is the list of categories, so nothing is missing when you fill in your own numbers.
- Passes for everyone who travels, including any side-event tickets
- Flights and hotels, which tend to climb sharply as the week fills up
- Ground transport, which surprises people every year
- Booth, sponsorship or hosting costs, if any
- Merch, printing and shipping
- Meals and meetings you pay for
- Team time away from normal work
- Prep time: the outreach hours in the weeks before
Team time is the line almost everyone leaves out. Three people on a six-day trip is 18 working days that didn't go to shipping, selling or writing. Pick a loaded day rate for each person, even a rough one, and include it. The full line-item version is in my conference budget template.
The formula, step by step
Once you have the cost, the arithmetic is short.
- T = total cost, every category above
- Q = qualified meetings held (all four tests)
- S = meetings that reached a second step within 30 days
- O = outcomes within 90 days: a live integration, a term sheet, a published story, an accepted offer
Then:
- Cost per qualified meeting = T ÷ Q
- Cost per second step = T ÷ S
- Conversion = S ÷ Q
The last step is the one that turns a number into a decision. Ask what it would cost you to get the same second-step meeting another way: a flight to that partner's city, a warm intro through an investor, months of cold outreach. If the conference beats your best alternative, it earned its place on next year's calendar. If it doesn't, you've found budget for something else.
Worked example: a three-person team at a big crypto week
This is a hypothetical, built to show the mechanics. Swap in your own figures.
A three-person team (founder, BD lead, marketing lead) attends a five-day crypto week. Two weeks before, they write a target list of 40 names across partners, investors and press. They pre-book 18 meetings. Three cancel on the day.
On the ground, the team logs 64 conversations. Run against the four tests:
| Source | Logged conversations | Qualified (Q) | Second step in 30 days (S) |
|---|---|---|---|
| Pre-booked meetings | 15 | 14 | 7 |
| Two closed dinners | 11 | 6 | 2 |
| One ecosystem side event | 14 | 2 | 0 |
| Booth and hallway | 24 | 0 | 0 |
| Total | 64 | 22 | 9 |
Whatever T turns out to be, each qualified meeting cost T ÷ 22 and each live opportunity cost T ÷ 9. By day 90 the team has two integrations live, one briefing that became a quote in a trade story, and one angel conversation in diligence.
Now read the table, not just the totals. Fourteen of the 22 qualified meetings came from outreach done before the event. The 24 booth and hallway conversations produced none. In this example, the prep weeks did most of the work, and the expensive, visible part of the week did very little. That's the finding that should shape the next budget: spend more on booking meetings before you land and less on footprint.
How do you log meetings during the event without losing them?
You won't remember who said what by day three. Keep one shared sheet with these columns:
- Name and company
- Target-list category, or "off list"
- Tests passed (1 to 4)
- Next step and its date
- Owner on your team
Fill it in every night before sleep, or in the cab between venues. Voice notes work if you transcribe them the next morning. The owner column matters more than it looks: a next step with no owner is a next step nobody sends. The follow-up itself is its own discipline, covered in my 48-hour follow-up system.
When is cost per meeting the wrong lens?
This model fits teams going to a conference to build pipeline. It fits less well in four cases, and it's better to say so up front.
- Brand or launch moments. If you've timed a mainnet or a keynote to the week, measure it like a launch: coverage, reach, sign-ups.
- Hiring trips. Count candidates who reach a second interview instead.
- Service businesses. If you sell a service, conferences are about relationships, and the payoff can take a year. Track relationships started and revisited, not 30-day conversion.
- Your first time in a new region. Part of what you're buying is context: who matters, how deals get done, what the local press wants. That's real, and it won't show up in T ÷ Q.
Relationships also get undercounted by any tidy formula. After one TOKEN2049 I asked people to tell me about the ones they'd built, because the public story was all parties:
comment if you’ve build one good relationship at @token2049 I’d like to change the perception that everyone was at parties! Shilika Jain on X, 25 Sep 2024
So keep the networking column, revisit it in six months, and move anyone who became a partner, investor or source into the outcome row. The formula should make the week accountable. It shouldn't make you treat people like leads.
If you'd like help building the target list or the reporting around a conference, that's part of how I run Web3 PR campaigns.
FAQ
How many qualified meetings should a conference produce?
There's no universal benchmark worth trusting, because team size, category and event differ too much. Set your own target from your pre-booked count: if you've confirmed 15 meetings before landing, a reasonable goal is to hold most of them and add a handful from curated dinners. Compare against your own previous events, not someone else's.
Should meetings booked before the event count if they could have happened on a video call?
Yes, if they pass the four tests during the week. The conference is often the reason the other side said yes to a meeting at all. If you notice that most of your qualified meetings could have been calls, that's useful too: it suggests a cheaper trip or a smaller team next time.
How long should I wait before judging a conference's ROI?
Read cost per qualified meeting the week you get back, cost per second step at 30 days, and outcomes at 90 days. Partnerships and press often take a full quarter to land. Judging at day seven makes almost every conference look bad.
Do I need a CRM to run this model?
No. A shared spreadsheet with the five columns above is enough for most early teams. Move to a CRM when the same people start showing up across several events and you need their history in one place.
Going to a conference this quarter and want the math done before you spend? Book a 30-minute teardown with me and bring your target list.
