Talk to Customers for Six Months Before You Pitch a VC

The Insight

Most founders validate backwards: raise first, then find out what to build. Alon Talmor did the reverse. Before pitching a single investor for Ask-AI, he spent six months getting into enterprises and asking what they actually needed. One of those conversations handed him the product, and the customers who shaped it then vouched for him in the seed round.

His reasoning is blunt: "a startup is bound to die," and the number one risk is money. So from day one, the question is how the idea makes money. Enterprises won't pay for nothing, which makes them the most honest validation machine available. The decision rule: don't pitch capital until a real buyer has described a problem they'd pay to solve.

How He Did It

1. Target companies, not investors. He went after bigger startups (Yotpo, Monday.com, and similar 700 to 2,000-person companies) through any door available: a friend of a friend, cold attempts, asking simply to hear their problems.

2. Expect brutal hit rates. His math from doing it: try 10 to 20 companies, 2 to 3 take the call, and if you're lucky one tells you what they need. That's the price of real signal.

3. Ask one question and shut up. "What do you really need right now? What would really change how you work?" Yotpo's VP of delivery answered: knowledge scattered across support tickets, the CRM, Confluence, Jira and Slack, and a go-to-market team that couldn't answer anything from it. That answer became the product.

4. Convert the conversation into a design partner. He built a prototype, let Yotpo test it internally for free for about a year, and got a commitment they'd pay around $30K annually. Then he raised: "we raised the seed with the customers vouching for that."

What Trips Up Founders

Pitching the idea instead of hearing the problem. You come in with a deck and the enterprise politely declines. Alon came in asking to listen, with just enough of an idea to anchor the conversation.

Treating a friendly meeting as validation. Taking a call isn't a signal. Describing a painful problem unprompted and agreeing to test a prototype is.

Fearing the free year. A year of free design-partner work feels like lost revenue. It's actually the cheapest enterprise sales cycle you'll ever run, and it compounds into the fundraise.

When This Doesn't Work

If your buyer is a consumer or a $29/month self-serve customer, six months of enterprise listening tours is the wrong motion; ship and measure instead. And if investors are literally your only path to surviving the next 90 days, you've already skipped the window where this works. The signal it's failing: three months of conversations and no one has named a problem they'd budget for. That's the market telling you the answer.

The Question

Before booking a VC meeting, ask: which paying customer would vouch for this product today? If the answer is nobody, you're raising on a story you haven't tested. Spend the next month finding the person with the problem instead. You'll know within weeks whether the pitch just got easier.