On 11/04/2026, your surveys will start remembering where each customer left off. If a customer begins your survey and comes back later, they pick up at their first unanswered question instead of starting over at question 1.
In short: you will finish more surveys, and you will stop collecting second copies of answers you already have. Nothing you have already collected is deleted or changed.
Your numbers at a glance
What you look at | Direction | Typical brand |
Survey views | No change | 0% |
Responses collected | No change | under 0.3% |
Response rate | No change | under 0.2 points |
Completed surveys | Up | about +1% overall, more on longer surveys |
Answers on your first questions, in a date range | Down slightly | -0.4% |
"How did you hear about us?" answers, in a date range | Down slightly | -0.4% |
Events sent to Klaviyo, Shopify Flow, Shopify tags, Yotpo | Down slightly | -0.8% |
Measured across every brand on the platform over a full month, alongside live results from surveys that have already been running this behavior since July.
What does not change
How often your survey is shown, and to whom. Views, targeting and matching are untouched.
Your response rate. Views stay the same and responses stay the same, so the rate does not move. This is the number most often misread at a change like this. If it looks lower on a given day, that is normal day-to-day variation, not this change.
Your stored data. Every customer still has one saved answer per question. Nothing is deleted, and all-time totals per question are identical before and after.
Changing an answer mid-survey. Customers can still go back within a visit and correct an answer, and that correction syncs to your tools exactly as it does today. That accounts for about half of all answer changes.
The order attached to a response.
Short surveys. On surveys of 1 to 3 questions, customers almost never return mid-survey, so you will see no measurable difference at all.
What goes up: finished surveys
Returning customers bring the same effort to every visit, but today most of it is spent re-answering questions you already have, so the survey stalls.
On the surveys already running this behavior, return visits produced nearly twice as much new progress (visits that produced nothing new fell from 54% to 9%), and the completion rate rose by 1 to 3 percentage points. The longer your survey and the more often customers return to it, the larger the gain. Your deepest questions benefit most, because they are the ones customers rarely reach today.
Why a few counts dip
Today, a returning customer is walked back through questions they already answered. If they re-submit one, it is saved with today's date and pushed to your integrations again. Your dashboard replaces the old copy, but any data you already exported keeps both, so the same answer is counted twice in your records. Starting customers at their first unanswered question stops the re-walk, so those second copies stop being created.
Every record that disappears from a date range is a duplicate, or a newer copy of an answer you already have. No new information is lost.
How large the dip looks depends on how often you pull data:
Two specifics:
Attribution mix does not shift. When returning customers did change their "how did you hear about us" answer, the changes ran evenly in both directions between channels, about as many Facebook to Instagram as the reverse. Keeping the first answer does not favor any channel.
Integration events. About 6 in 10 of the pushes that stop were byte-for-byte identical to one you had already received: duplicate flow triggers, re-applied tags, re-written profile properties. The rest carried an edited value, which will now stay at the customer's original answer unless they navigate back and change it. Nothing already sent is retracted.
Could my store see more than this?
Most brands see nothing noticeable. A minority see a real step down in those three counts on the day of the change:
You are more likely to be in that group if you run surveys of 6 or more questions with low completion, if your customers are frequently shown the survey again (subscription and repeat-purchase brands, link or email surveys that re-match), if your questions push data to Klaviyo, Shopify Flow, Shopify tags or Yotpo, or if you sync in real time. Roughly 1 in 10 brands sees 3% or more on at least one of these counts, and a small number of brands see considerably larger drops.
Want your own numbers? Contact support and we will send you the projected change for your store before [DATE], survey by survey. We would rather show you the figure in advance than have you find it in a dashboard.
What you need to do
Nothing, if the above sounds good. The change applies to your surveys automatically on 11/04.
If you would rather keep today's behavior on a survey, open it in the survey builder, go to the Behavior tab, and switch off Resume at the first unanswered question before 11/04. This is a per-survey setting, so turn it off on each survey you want to keep starting from question 1 every time.
For surveys you create later, you can set your preferred default under Settings > Surveys > Miscellaneous. That default applies to newly created surveys only. It never changes surveys you already have.
Either way, your customers' answers are saved. The setting only controls where a returning customer starts. You can change it at any time after 11/04 and once the change is live we can tell you what it actually did to your data.
How we measured this: a full month of data across every brand on the platform, using the complete answer-change history rather than a sample, plus a before-and-after comparison on surveys that have been running this behavior since 22 July. Figures are typical values. Your own store's numbers are available from support.