Demo Days That Produce Follow-On, Not Just Photographs
Most demo days are a good evening and a dead end. What to change about the guest list, the format and the fortnight afterwards so something happens.
The event is not the problem
Demo days are usually run well. The room looks good, the pitches are rehearsed, the photographs are fine and everyone says it went brilliantly.
Then, six weeks later, nothing has happened. No follow-on meetings that would not have happened anyway, no customer conversations, and a cohort quietly wondering what the point was.
The failure is almost never the event. It is the guest list, and the fortnight afterwards.
Invite fewer, better
The instinct is to fill the room. A full room looks like success and photographs well.
But the number that matters is not attendance. It is meetings scheduled in the two weeks afterwards, and that is driven almost entirely by whether the right twenty people were there.
The right twenty are:
- Investors who actually write cheques at your startups' stage, in your sectors
- Corporates with a real procurement or partnership route, not just an innovation team
- Operators who could become customers or advisers
- Two or three journalists, if coverage matters to your funder
Two hundred general attendees produce a good atmosphere and no meetings. The reason is simple: someone who cannot act on what they see has no reason to follow up.
Ask each startup who they want in the room
The most underused step in demo-day planning. Every founder can name three companies or people they want to reach. Collect those lists six weeks out and invite them directly, saying which startup asked for them.
The response rate to "a company in our programme specifically asked us to invite you" is dramatically higher than a general invitation, and it fixes the relevance problem at source.
Fix the format
Three rules do most of the work:
Same structure for everyone. Problem, what exists, evidence it works, the ask. Comparable pitches are easier to follow and much easier to remember afterwards, which is when decisions get made.
Hard time limit, enforced visibly. Four or five minutes. A timer everyone can see. Overrunning is not generosity to that founder; it is theft from the later ones and from the audience's attention.
The ask is stated explicitly. Every pitch ends with what the founder actually wants - a specific customer introduction, a pilot, a cheque of a certain size, a hire. "We're open to conversations" is not an ask and produces no action.
Build the follow-up into the night
This is the part that decides whether the event worked, and it has to be set up in advance because nobody organises it afterwards.
During the event, capture interest structurally. A card, a short form, a QR code - anything where an attendee names which startup they want to talk to and why. Do not rely on business cards and memory; both decay by Monday.
Within 48 hours, send the introductions. Not a general thank-you email - specific, double-opted introductions between the named attendee and the named founder, with one line of context. Forty-eight hours matters because interest expressed on the night has a short half-life.
At two weeks, ask both sides whether the meeting happened. This is the step that essentially never happens, and it is how you find out whether the event produced anything.
Measure the right things
Most demo-day reporting is attendance and photographs. Useful measures:
| Measure | Why it matters |
|---|---|
| Introductions made within 48 hours | The mechanism working |
| Meetings that actually happened at 2 weeks | The mechanism producing something |
| Outcomes at 3 months - pilot, customer, cheque, hire | The point |
| Attendees who were named by a startup | Whether the guest list was relevant |
| Repeat attendance next time | Whether it was worth their evening |
Attendance is worth recording only as a denominator.
Prepare the founders, but not too much
Pitch coaching helps. Over-coaching produces a cohort that all sound the same, and audiences notice.
Two sessions is usually right: one on structure and the ask, one on delivery a few days before. What is worth more than a third coaching session is a rehearsal in the actual room, with the actual clicker, in front of two people who will be blunt.
The most common avoidable failure on the night is technical - a laptop that will not connect, a video that will not play, a font that renders differently. A rehearsal in the room removes almost all of it.
Consider not running one
Worth saying, because demo days are treated as compulsory.
A demo day is expensive in staff time, founder time and goodwill with your invite list. If your cohort is not at a stage where anyone in the room can act - pre-product, pre-revenue, no clear ask - the event will be pleasant and pointless, and you will have spent your relationship capital on it.
The alternatives are often better at that stage: a small dinner with six relevant people, individually arranged introductions over a month, or a written cohort brief sent to a curated list. None photographs as well. All of them produce more meetings.
The two-week test
If you want one question to judge your last demo day by:
How many meetings happened in the two weeks afterwards that would not otherwise have happened?
If the answer is unknown, that is the finding, and it is fixable before the next one.
Frequently asked questions
- How many startups should present at a demo day?
- Eight to twelve. Beyond that the audience stops distinguishing between companies and the later slots are wasted. If your cohort is larger, run two events rather than one longer one.
- How long should each pitch be?
- Four to five minutes, hard-stopped, with the same structure for everyone. Longer pitches do not convey more; they convey the same amount less clearly, and they make the event too long for the guests who matter most.
- Who should actually be in the room?
- Fewer people, better chosen. Twenty investors and corporates who genuinely invest or buy in your sectors beat two hundred general attendees. The metric that matters is meetings scheduled afterwards, and that is driven by relevance rather than headcount.
- What is the most commonly missed step?
- The fortnight after. Most demo days end with warm conversations and no mechanism to convert them, so the introductions decay. Capturing who wanted to meet whom on the night, and sending those introductions within 48 hours, is what separates the events that work.
