How to research a prospect before a call in 10 minutes

Most advice on this question is a list of places to look. The company site, their profile, recent news, their investors, the tech stack, the org chart.

Follow it and you will spend forty minutes and arrive at the call with a folder of facts and no idea which one to say first.

The problem is not the sources. It is that nobody tells you when to stop, or what the output is supposed to be.

The output, defined first

Research is finished when you can write these three sentences:

1. The problem I think they have, in language they would recognise.
2. Why I think that, pointing at something specific and dated.
3. The one thing I want to find out on this call.

That is it. Three sentences, and you should be able to say the first two out loud in under fifteen seconds.

Defining the output first is what makes the timebox possible. Without it, you are reading with no criterion for relevance, which is why you cannot stop.

The ten-minute pass

In this order. The order is the method, because each step is only worth doing if the previous one came up short.

If the timer runs out with the three sentences unwritten, you did the research wrong, regardless of how much you learned.

The stopping rule

Stop when the next thing you learn will not change what you say in the first thirty seconds.

That is the whole rule, and it fires much earlier than people expect. Their headcount will not change your opening. Their last announcement probably will not either. The thing they wrote about a broken process three weeks ago will.

Everything past the stopping rule is comfort. It feels like diligence, and it is really deferral, because reading is easier than deciding what you think.

Turning notes into the first thirty seconds

Here is where most preparation dies. People do good research and then open with a generic pleasantry, because the notes were never in a form you could say out loud.

The three sentences map directly:

Phrase the opening as a question, not a diagnosis. You have an inference, not a fact, and stating it as a fact invites correction rather than conversation. Being visibly willing to be wrong about it is what makes the other person willing to tell you what is actually going on.

And say the evidence in your own words. Reading their sentence back to them lands as surveillance even when it is not, for the same reason covered in how to say why you are reaching out.

Where the line is on personal information

Use what they put into their professional presence deliberately. Do not use what you merely found.

Something they published about their work is fair, and referencing it is normal professional behaviour. Their personal life, their location, their family, anything assembled by cross-referencing sources: not fair, and the tell is that you would feel odd explaining how you know it.

The practical test is simple. If you would not be comfortable saying where you found it, do not use it. That test also protects you from the thing that actually kills calls, which is a prospect quietly deciding you are strange in the first minute.

A worked pass, minute by minute

Shape of a real one, with the company anonymised. The point is the decision at each step, not the findings.

Minute 0. Open their profile and sort their own writing by date. Nothing in the last six months. Two posts last year, both re-shares with no comment. Step one is a dead end and I know it in ninety seconds.

Minute 2. Move to step two rather than digging further into a source that has already told me it is empty. This is the discipline that saves the pass: people spend eight minutes on a silent profile because leaving feels like giving up.

Minute 3. Their remit changed four months ago. They now own a function that used to sit under someone else. That is specific to them, dated, and directly upstream of the problem I sell into. It is an inference, not a statement, so it is a situational reason.

Minute 6. One level out. Two other people would be in the room, and one of them has written publicly about exactly this problem. That is more valuable than anything about my actual contact, and I would not have found it if I had spent the whole pass on one person.

Minute 8. Write the three sentences. The problem I think they have, phrased in their language. Why I think it, which is the changed remit plus the colleague's public description. What I want to learn, which is whether they own the budget or whether the colleague does.

Minute 10. Stop. I know less about this company than I could. I know exactly what I am going to say in the first thirty seconds, which was the goal.

The general lesson: the pass often finds its best material one step away from the person you were researching, and only if you leave the first source quickly enough to get there.

Why this is hard to sustain

Ten minutes a person, done properly, is real work. Twenty calls a week is over three hours of preparation, and it is the first thing that gets cut in a busy week.

That is worth naming honestly, because the failure mode is not doing it badly, it is doing it for the first five people and generic-ing the rest. Two well-prepared calls beat six unprepared ones, so the honest fix is usually fewer calls rather than faster research.

If you want the reading half done for you, that is roughly what Trendle exists to do: it runs the pass every morning and hands you the shortlist with the reason attached, so the ten minutes goes into deciding whether you agree rather than into looking. Related: how to prioritize sales leads covers deciding who is worth the ten minutes in the first place.