How to find which signup survey question is costing you completions
Remove one question at a time, compare what finishes, and keep your answer options stable so the segments you already use still work.
- Completion drops at specific questions, not evenly across the survey.
- Personal questions with no obvious reason, like age, gender or purchase plans, cost the most finishes.
- Remove one question at a time and compare completion between versions to find the culprit.
- Keep answer option names identical across versions so your existing segments keep working.
By the end of this guide you will have removed one suspect question from your post-signup survey and set up a simple comparison to see whether more people finish without it. You need two things before starting: a live post-signup survey that is already collecting responses, and access to whatever completion tracking your form tool already shows you, whether that is raw responses or a completion count.
Every question you ask up front costs you completions. That is not a flaw in your survey design, it is how attention works. A new subscriber just gave you their email address and is willing to answer a few things, but each additional question is another chance to close the tab. The drop is not spread evenly. It clusters at specific questions, and the questions that cost the most are the personal ones with no obvious reason behind them: age, gender, income, whether they plan to buy something. When a reader cannot see why you are asking, they stop.
Decide what you actually need to ask
Before you touch anything, write down the test you are running. The question to ask of every question on your survey is simple: will I actually use this answer in the newsletter or in the next campaign? If the answer is no, the question is a candidate for removal regardless of what the comparison shows. If the answer is yes, keep it and let the comparison tell you what it costs.
- 01Pick one suspect question
Choose the question you suspect most, usually a personal one with no clear use, and leave the rest untouched.
- 02Count today's completions
Write down how many people finished the current survey version over a fixed window, for example the last two weeks of responses.
- 03Remove that one question
Publish a new version of the survey with only that question deleted, changing nothing else.
- 04Wait for a comparable sample
Let the new version run until roughly as many people have started it as started the old one, so the comparison is fair.
- 05Compare completion
Compare the share of starters who finished the new version against the old one. A meaningful gap points at the removed question.
- 06Decide what to ask instead
If the question was useful, either drop it, move it later in the survey where fewer people see it, or reword it so the reason for asking is obvious.
- 07Repeat with the next suspect
Run the same one-question removal on the next weakest question until the survey holds only questions you will use.
Keep your segments working
When you change questions, keep the answer option names identical across versions. If one version of your survey records a reader as living in "Downtown" and the next records "downtown area," you have split one segment into two, and the next campaign built on that segment will miss people. Rename options only when you are prepared to rebuild the segment on top of them.
Where people get stuck
The survey is short already and every question feels essential. Test the one that made you slightly uncomfortable to add. If you cannot name the campaign or newsletter decision that uses the answer, that question is your first removal, and the comparison will usually confirm it.
You removed a question and completion did not move. That is a result, not a failure. It tells you the question was not the bottleneck, and the drop likely sits at an earlier question or at the survey's sheer length. Move your test to the first question readers see, since early exits remove everyone behind them.
You cannot tell where readers quit because your tool shows only finished responses. Compare starts against completions if your tool reports both. If it reports neither, a shorter survey tested against the same signup volume is the honest fallback, and it is a reason to check what your form tool actually records before you rely on it.
A note on tooling
If you run your post-signup survey on greetform, every answer a subscriber gives is stored against their session, including the question, the option chosen, and the branch the survey took based on conditional logic. That record is what makes the removal comparison possible without guessing. Every answer also syncs to the subscriber record in Beehiiv, ActiveCampaign or Mailchimp as fields you can segment on, so keeping your option names consistent protects segments you have already built there. There is a free forever plan with no card needed.
