The conference room had good intentions and bad lighting. An engineer I'll call Marcus had spent six weeks building a prototype that could cut his company's invoice-matching time from four days to twenty minutes. He opened the demo with the settings page. Fourteen minutes later the CFO was reading email on her phone, and the COO asked whether this was "basically a spreadsheet." The budget request died of natural causes, right there between the dropdown menus.

Marcus didn't have a product problem. He didn't know how to demo software to executives, which is the most common way good prototypes go to die: the builder shows what they built instead of what the business gets. Executives don't fund features. They fund endings, the kind where a painful thing stops happening.

So here's how to demo software to executives so the check actually gets signed: tell one story, with real data, in seven minutes or less. Open on their pain, walk a single workflow end to end, land on a dollar figure, make a specific ask, and then stop talking. Feature tours lose rooms. Stories with arithmetic win budget.

Why demos die in the first two minutes

Most demos fail before anyone has finished their coffee, and they fail in predictable ways. If you recognize your own rehearsal in this list, good news: it's fixable before the meeting, not after the "we'll think about it."

Every one of these comes from the same instinct: showing the software as it exists instead of the problem as it feels. Flip that, and the whole demo changes shape.

How to demo software to executives in seven minutes

Seven minutes is not arbitrary. It's about as long as a busy executive will watch anything before the question "what does this cost or save" becomes physically unbearable. Structure the demo like a flight, not a museum tour.

Minute 0:00 to 0:30 — the pain, in their words

Open with the problem as the business describes it, ideally quoted from someone in the room. "Accounts payable touches every invoice four times. Four. We counted." If they nod in the first thirty seconds, you've already beaten 80% of vendor demos they've sat through this quarter.

Minute 0:30 to 2:00 — the hero moment

Show the single most impressive thing the prototype does, immediately, with no setup. The invoice drops in, the match appears, the exception gets flagged. You're not proving completeness. You're proving plausibility.

Minute 2:00 to 4:30 — one workflow, end to end

Now rewind and walk one complete path: intake, processing, the exception, the resolution. One path, not five. You're telling a story with a beginning and an ending, and the ending is a Tuesday that takes twenty minutes instead of four days.

Minute 4:30 to 5:30 — the money slide

The money slide gets its own section, because it's the one that actually wins the budget.

Minute 5:30 to 7:00 — the ask, then silence

State exactly what you need: budget, a decision date, a sponsor. Then stop talking. Silence is a feature of the demo, not a bug in it.

One story, real data

Pick one persona and one day. "This is Dana in AP. This is her Tuesday." Then run Dana's actual invoices, or a scrubbed copy of them, through the workflow. Real data does two things fake data never can: it makes the room lean forward because they recognize the vendor names, and it quietly proves the prototype survives the messiness of reality.

A typical mistake is using clean demo data because the real data is embarrassing. Use the embarrassing data. When the prototype chokes on a weird edge case and recovers gracefully, you've demonstrated more competence than any polished happy path could. If it chokes and doesn't recover, better to learn that in rehearsal than in front of the CFO.

This is also why the demo belongs at the end of a disciplined prototyping process. If you've been building against real workflows the whole time, like the approach in the rapid prototyping playbook, the demo is just the last week of work shown in order. Teams that practice demo-driven development never have to scramble for a story, because they've been telling it every Friday.

The money slide nobody taught you

Executives think in dollars and risk, so give them both, in one slide, with math a middle schooler could check. Keep it illustrative and say so: "These are our numbers, rounded."

Say the prototype saves AP roughly 12 hours a week. At a loaded cost of $48 an hour, over 46 working weeks, that's about $26,500 a year. The project you're asking for costs $40,000 and ships in ten weeks. Payback lands somewhere around month 18 conservatively, or month 9 if the invoice volume grows the way it has the past two years. Two sentences, and the room is doing your ROI for you.

Notice what's missing: the architecture diagram, the framework choice, the word "AI" used as a noun. Nobody funds those. If a technical question comes up, answer it in one breath and get back to the story. The details live in the appendix and in your v0 to v1 checklist for what happens after they say yes.

Handling the live demo gods

Live demos are a negotiation with chaos, and chaos negotiates hard. Stack the deck:

And accept now that something will go sideways. The senior person who recovers calmly from a glitch wins more trust than the one who never glitches. Rooms fund composure.

The ask, and the silence

The worst ending in demo history is "so... yeah, that's it." Close with a specific, dated ask: "We're asking for $40,000 and one decision by the 15th. In return you get this in production by the end of the quarter." Then stop. Count to seven in your head. The first person to speak after a clear ask is usually the person holding the budget, and what they say next tells you everything.

If the answer is "not yet," you still won: you now know the objection, and objections are just requirements wearing a disguise. Write it down, address it, and ask for the follow-up slot before anyone leaves the room.

Marcus, by the way, got his second chance two months later. Same software, different seven minutes. Dana's Tuesday, one workflow, one money slide, one ask. The COO approved it before the meeting hit its halfway mark, then spent the remaining time asking when the next department could get one. That's the real measure of a demo: not applause, but a calendar invitation for the sequel. Be honest about what the prototype still owes, because prototype tech debt doesn't forgive anyone, then book the room. Seven minutes. One story. Real numbers. Go get the budget.