The vendor a client almost hired, and why we said no

The scenario below is a composite built from patterns across different engagements, not a specific client. Details are changed accordingly.

A business I advised was two days from signing with a marketing automation vendor. The pitch was polished, the demo was clean, the price was fair. On paper, an easy yes.

What the pitch didn’t answer

Before signing off, I asked the vendor three questions directly: what does data export look like if we leave, what’s the actual uptime history not the marketing page number, and who owns support after the sales team hands us off. The answers were vague on all three, in the specific way that means “we haven’t been asked this before, or we have and don’t like the answer.”

That vagueness was the actual signal, not the product itself.

Why vagueness beats a bad demo every time

A bad demo is obvious. A vague answer to an exit question is not, because it hides behind confidence and good design. Most vendor evaluations stop at “does it do what we need,” and skip “what happens when we need to leave, or when something breaks at 2am.” Those are the questions that predict the relationship two years out, not two weeks out.

What we did instead

We didn’t walk away from the category, the automation problem was real. We asked for a written data-export guarantee and a named support contact before signing anything, not after. The vendor could provide both, once asked directly, which told us something too: the gap wasn’t malice, it was that nobody had pushed on it before.

The lesson, generalized

A slick demo tests the product. It doesn’t test the relationship. The questions worth asking in any vendor evaluation are the ones about exit, ownership, and what happens on a bad day, not the ones about features on a good one. Most businesses only learn to ask these after getting burned once.