4.4.4 Value Proposition Test - Event

In Brief

Description

An event smoke test asks prospects to do something one-on-one interviews never can: gather as a group, in one place at one time, and prove their interest by actually showing up. You organize a real gathering around your value proposition — a workshop, meetup, webinar, or dinner — and watch whether people register, attend, engage, and, if you charge, pay. It is one of the higher-commitment rungs of the Value Proposition Test family: setting aside an evening and buying a ticket sits high on the commitment ladder, which makes attendance a stronger demand signal than agreeing in an interview that a problem matters.

To run one you have to solve the same problems your eventual product would face: how to reach the audience, what to promise them, and how to deliver an experience they find useful. However, doing that for a single event is far cheaper than building a product. This is particularly true where part of the value proposition is social interaction or trust, which is best generated by real humans interacting.

The format can range widely depending on the depth of signal you want:

• Lecture by one or several speakers
• Peer-to-peer formats like an unconference (e.g., LeanCamp) or a LeanCoffee discussion
• Smaller, intimate formats like a mastermind or dinner
• Networking or drinks
• Conference, online or in person
• Webinar or live video stream

Smaller, intimate formats produce richer qualitative feedback; larger lecture-style events produce a stronger attendance count. In-person events are easiest to fill in a larger city with an existing scene, and they build a closer relationship with attendees than an online interaction does. Online formats cost less and reach a wider geography but give you fewer of the side conversations that carry the real insight.

Because an event is a small product in its own right, it can test more than topic interest: it can qualify leads, test a promotion channel, or lead into a presell of the eventual product. It also combines well with other methods — pair it with a Landing Page Test to gauge interest before you commit to a venue, or follow it with a Concierge Test for the most engaged attendees.

How to

Prep AI Prompt

1. Define the demand hypothesis. Write down the specific belief you are testing: who will show up, why this topic matters to them, what attendance number would convince you (and what number would falsify the hypothesis). If you cannot state a falsifying outcome, you are not running a test — you are running an event.
2. Pick the format. Choose the event format that matches the audience and the depth of signal you need: workshop, meetup, webinar, unconference, dinner, mastermind. Smaller intimate formats produce richer qualitative signal; larger lecture-style events produce stronger quantitative attendance signal.
3. Set price and pitch. Decide what to charge. Free events test topic interest cheaply but produce noisy signal. Paid events (even $10–$25) filter for stronger intent. Draft the pitch using language from your prior prospect interviews — not your internal jargon.
4. Find a venue. Match the venue to the format and audience. Coworking spaces, libraries, and partner offices are often free. Online formats cost nothing and broaden geography but reduce signal quality.
5. Create the event page. Build a registration page on an event registration platform. The description is the smoke test artifact — readers commit (or don’t) based on this copy. Track registration rate as the first signal.
6. Choose the channels. Use what you learned in earlier research to pick promotion channels (Slack groups, newsletters, partner orgs, paid ads, personal network). Note which channel each registrant came from so you can attribute interest later.
7. Set the threshold. Decide before launching what counts as success: minimum registrations, minimum attendance, minimum paid conversions. Write this down. Post-hoc rationalization is the most common failure mode in smoke tests.

Analysis AI Prompt

1. Compare attendance against your threshold. Use the pre-set thresholds, not new ones invented after the fact. Practitioner rules-of-thumb for context: free events tend to see roughly half of registrations show up; paid events convert at noticeably higher rates; virtual events sit somewhere lower than in-person paid events. Treat these as rough anchors, not precise stats — your own channel mix and audience will dominate. Even 10 people who paid to attend a workshop on your specific topic is a meaningful demand signal.
2. Measure engagement quality, not just headcount. Count how many stayed to the end, asked questions, approached you during breaks, or gave contact info voluntarily. An engaged audience of 8 is a stronger signal than a disengaged audience of 40.
3. Read channel attribution. Which promotion channel produced the most registrations? Which produced the most attendees who actually showed up? Channel attendance rates often vary more than overall attendance rates and tell you where to invest if you scale up.
4. Synthesize the conversations. Use AI to cluster transcripts into recurring themes — what problems came up unprompted, what language attendees used, what questions repeated across multiple conversations. The repeated unprompted question is often the real demand signal.
5. For virtual or hybrid events, mine engagement signals. AI-powered event platforms monitor audience attention, run real-time polls, and analyze chat sentiment. Look for where attention drops, which polls produced strong consensus, and which topics generated chat activity.
6. Follow up within 48 hours. Send a short survey while the experience is fresh. Ask what was most valuable, what was missing, and whether they’d attend again or pay for a related product. The “would you pay for a related product” answer is far more reliable when asked within 48 hours than after a week.
7. Make the go/no-go call. Compare results against the falsifying threshold you set in Prep. If the threshold was missed, document it honestly and iterate on the proposition or pivot the audience — don’t lower the bar retroactively.