Forty trials, two paid, and I have no idea what the other thirty-eight thought - how do you find out?

Before asking anybody, find out where they stopped. You probably already have this and have not looked.

For each of the thirty-eight, what was the last thing they did? You will typically find a small number of clusters:

Never got past signup. Never returned after the first session. That is not a product problem, it is a positioning or expectations problem - they arrived expecting something else.

Got stuck at a specific step. Connected nothing, imported nothing, never reached the point where the product does its thing. This is the big one and it is an onboarding problem.

Used it properly and left anyway. These are the interesting ones and usually the smallest group. They understood it and decided no, which is the only group whose opinion on value is worth anything.

The reason to do this first: the three groups need different questions, and asking all thirty-eight the same thing produces mush. Somebody who never returned cannot tell you about your pricing.

Most people discover the second cluster is the majority, which means the answer is not why did you not buy - it is that they never got far enough to have an opinion, and the fix is in the first ten minutes rather than in the product.

Then fix the timing, because a month late is the real problem with your plan.

Ask at the moment they leave, automatically, from now on:

When a trial expires without converting, one email, one question. Not a survey, a single question you would answer yourself.

When somebody cancels, one question on the cancellation screen. Free text, optional. People are surprisingly forthcoming at that exact moment because they have just made a decision and it is fresh.

When somebody stalls at a known step for a few days, one message asking whether they got stuck. This one frequently converts people as well as informing you.

On the wording, since that is what is stopping you: the version that gets replies is short, personal, and asks for help rather than feedback. Something like, you signed up a couple of weeks ago and did not carry on; I am trying to work out why, and one line would genuinely help. Nothing to sell in it.

Expect a low reply rate and useful replies. Five out of thirty-eight is a normal outcome and five is plenty.

And do email the thirty-eight now. Late is worse than immediate and much better than never, and you will not have another cohort for months.

Small thing that doubled my reply rate: send it from a person, to one person, with no unsubscribe footer and no images. The moment it looks like a campaign it gets treated as one.

Thirty-eight people is small enough to read every session recording individually, which is grim but fast.

When the replies come, read them carefully, because the first reason people give is usually the polite one.

Too expensive is the default answer and it is frequently not true. It means the value was not obvious enough to justify the price, which is a positioning problem rather than a pricing one. The follow-up worth asking: what would it have needed to do for the price to feel reasonable.

No time right now usually means not a priority, which means the problem is real but small for them.

Missing feature X is worth taking seriously only if several people name the same one, and even then check whether the ones who named it would actually have paid.

Silence is data too. Thirty-eight people who could not be bothered to complain is a different signal from thirty-eight people who wrote angry emails.

One more thing worth doing: compare the two who paid against the thirty-eight who did not, on whatever you know about them. Company size, role, how they arrived, what they did first. If the two have something in common, that is the most valuable finding available from this cohort - it tells you who to go and find more of, which is a better use of the next month than trying to convert people who were never going to buy.