When to send it
Send this three to five days after someone signs up, once they have logged in a couple of times but before they have decided whether your product earns a place in their week. Earlier than that and they have nothing to say; much later and the people who struggled have already drifted off without telling you why. Two places work well: a short prompt inside the app on their second or third visit, and a line in your day-three onboarding email.
If your app knows the customer's first name and plan, add them to the link as ?name= and ?plan=. The survey greets people by name, with a plain greeting when the name is missing, and the plan lets you compare answers from free and paid accounts in the CSV export.
Why these questions
The survey starts with the job, not the product. Asking what someone signed up to get done, in their own words, tells you whether your marketing attracts the people your product is built for. If the answers describe a job you don't do well, that's a positioning problem rather than an onboarding one.
Next comes the question that matters most for activation: have they done that job yet? Anyone who says "partly" or "not yet" sees a checklist of common blockers, from importing data to waiting for a teammate. Ticking boxes is quicker than writing, so more people answer, and each option maps to something you can change.
The effort question uses customer effort wording on a 1–7 agreement scale. Track its average month by month. If you ship a new setup flow and the number doesn't move, the flow didn't help.
Finally, people pick what would help them most. Anyone who asks for a call is shown an optional contact question, so you can book it while they still care.
Why there's no coupon
A discount during onboarding pulls attention toward price at the very moment you want people focused on getting value. Most new users will answer a short, specific question from a founder anyway, especially when it's clear you'll act on it.
What to do with the answers
Read the "not yet" answers every week while your product is young. Count which blockers come up most and fix the top one before touching anything else. If importing data is the main complaint, a spreadsheet template or a ten-minute hands-on import may do more than any new feature. Reply to every call request within a day.
Compare the goals people describe with the copy on your homepage, and borrow their words. When the blockers shrink, move on to Feature feedback for specific parts of the product, and use Trial ended without upgrading for the people who never got going.