Is your survey bringing in fewer responses than usual, or no new responses at all? The cause is usually a change to your survey settings, your audience, your installation, or your organization's status.
This guide helps you find which change caused the drop and confirm it in your dashboard. Start by checking whether displays or the response rate went down. They have different causes, so knowing which one fell narrows your search.
In this article:
- Step 1: Comparing displays and response rate
- Step 2: Checking why displays dropped
- Step 3: Checking why the response rate dropped
- Tracking your setting changes
- Getting your response volume back
Step 1: Comparing displays and response rate
A display is one showing of a survey to a person. Your response rate is your responses divided by your displays. In your survey's dashboard, compare both numbers for the period before the drop and the period after it.
If both numbers dropped, work through Step 2 and Step 3.
Step 2: Checking why displays dropped
If displays dropped, fewer people qualify to see your survey. Work through these causes in order, and confirm each one before moving to the next.
Your organization is on hold
Your organization goes on hold when it reaches its monthly response limit, when your subscription is canceled, or when a failed payment isn't resolved. While your organization is on hold, every survey stops on every channel: web, mobile, link, and email.
To confirm it, check whether any of your surveys are still collecting responses. If none are, check your monthly response limit and your billing status. See Plans and response limits, How can I cancel my SatisMeter subscription?, and Updating your payment method for details.
Your survey is set to one-time
Frequency controls how often one person can be surveyed by the same survey. A survey without a frequency asks each user once, ever. If your team expected the survey to recur, displays fall sharply once most of your existing users have been asked.
To confirm it, open the survey's settings and check its frequency. If you want to survey the same users again, set a frequency.
Users are in their cooldown
After someone is surveyed, they wait out a cooldown before the survey's frequency makes them eligible again. If you surveyed a large group at the same time, the whole group becomes eligible again at the same time. Displays stay low in between.
Example: You survey most of your users in January. Those users are blocked until July, so displays stay low until then.
To confirm it, compare the date of your last large wave of displays with the survey's frequency. To learn how frequency and cooldowns work, see How SatisMeter decides who sees a survey.
Your userId format changed or users were migrated
SatisMeter identifies each user by their userId. If you change your userId format or migrate users to new IDs, SatisMeter creates brand-new profiles. Each new profile gets a new First Seen date and a new sampling result.
Any targeting based on First Seen or sampling then treats these users as new, so fewer of them may qualify. SatisMeter never merges anonymous profiles into identified ones.
To confirm it, check whether the drop lines up with a release that changed how you send userId, or with a user migration. For more on First Seen, see User traits: First seen vs. Created At.
The snippet was removed or changed
A site release can remove or change the snippet, the installation code that shows the SatisMeter widget on your website. Without a working snippet, the widget can't appear.
To confirm it, check whether the drop started right after a release. Then work through Survey not showing? Troubleshooting checklist or How to test the survey installation.
Step 3: Checking why the response rate dropped
If displays held steady but responses fell, the same number of people see your survey, but fewer of them respond. Look for a recent design or timing change.
The rating question is no longer required
If someone turns off the Required setting on your rating question, people can close the survey without answering it.
To confirm it, open the survey's questions and check that the rating question is still required. To learn how required questions decide whether a response counts, see Partial responses: when does a response count?
People are dismissing the survey
When someone closes the survey with a required question unanswered, SatisMeter records the response as dismissed. A dismissed response counts as a display, not a response, so it lowers your response rate.
To confirm it, compare displays with responses in your dashboard over time. A widening gap means more people close the survey before completing it.
The time-on-page delay increased
A time-on-page delay makes the survey wait before it appears. If someone increased the delay, for example to 10 seconds, two minutes, or five minutes, more people leave before the survey shows.
To confirm it, check the survey's time-on-page delay and compare it with what your team expects.
The survey moved to a different page
If the survey now shows on a page people only pass through quickly, many of them leave before they respond.
To confirm it, check which pages the survey now targets. For more on targeting, see Targeting.
Tracking your setting changes
We can't recover a log of past setting changes, so we can't tell you what changed in a survey or when. Recording changes yourself makes future drops much faster to diagnose.
Tip: Record each change and its date in the survey name or in an internal changelog that your team shares.
Getting your response volume back
Once you find the cause, undo the change or adjust your survey around it. To raise your response volume further, try these steps:
- Turn on email reminders, which are off by default.
- Show the survey later in the user journey.
- Review your sampling to check how many eligible users can see the survey.
- See How can I get the most responses from surveys? for more ideas.