# Post-Purchase Fit Surveys: What Bracketers Tell You When You Ask | BracketFence

Source: https://bracketfence.com/blog/2026-10-10-post-purchase-fit-surveys-bracketing.html

## The gap in your returns data

Returns data is a record of actions, not reasons. Two orders with identical patterns, three sizes ordered and two returned, can have completely different causes. One shopper was unsure about fit. Another feared a stockout and ordered a backup. A third was buying for someone else and using their own account. The dashboard treats all three as bracketing. The fixes for each are different, which means a fix built on the dashboard alone is a guess.

This is the blind spot of even the best bracketing analytics. Scoring models are excellent at finding who brackets and how often, but they infer intent from behavior, and behavior is ambiguous. The cheapest way to resolve the ambiguity is to ask. Not every shopper, and not with a long questionnaire. Just the multi-size orderers, with three questions, at the right moment.

## Three questions, not thirty

Length is the whole game. A survey that takes thirty seconds gets answered. A survey that takes three minutes gets closed. Three questions cover nearly everything you need: which size did you expect to keep, what made you order more than one size, and did the product page answer your sizing question. Offer multiple choice with an optional write-in box. The multiple choice gives you clean segments. The write-in gives you the phrasing shoppers actually use, which is gold for product-page copy.

Resist the urge to add a fourth question about satisfaction or NPS. Every additional question cuts completion, and the completion rate is what makes the data trustworthy. If the survey only gets answered by your most motivated shoppers, the answers skew toward complaints. Keep it short enough that the average bracketer answers on a phone while waiting for coffee.

## Where to put the survey

Timing and placement decide the response rate. Do not put the survey at checkout. The shopper is trying to buy, and anything that slows that down is the enemy. Do not put it on the return request form either. By then the mood is wrong and the answers skew negative. The sweet spot is post-delivery and pre-return: trigger it from the delivery confirmation email or the order tracking page, limited to orders with more than one size of the same SKU.

That window matters because the shopper has the garments in hand but has not yet decided what goes back. Their memory of why they ordered multiple sizes is fresh, and they have no incentive to shade the answer. A simple link, one screen, three taps, done. Expect single-digit response rates and be happy with them. On a few thousand multi-size orders a month, even five percent is a large, honest sample.

## The three buckets the answers fall into

Once the responses accumulate, they sort into three buckets with almost no overlap. Bucket one is fit uncertainty: the size chart did not answer the question, the model photos did not include their body type, the reviews disagreed. The fix is fit content, better charts, garment measurements, fit notes. Bucket two is inventory anxiety: they ordered a backup because they feared a stockout or a long restock wait. The fix is stock messaging and fit-first filtering, not fit content at all.

Bucket three is proxy buying: the order was for a partner, a kid, or a parent, and it looks like bracketing in the data but is not. The fix is household profiles or gift-mode checkout, and the analytical fix is to stop scoring these orders as bracketing. Most brands are surprised by the size of buckets two and three. They have been treating a large share of their bracketing as a fit problem when it was a messaging problem and a data problem.

## Closing the loop

The survey earns its keep when the answers flow back into the systems that score and act on bracketing. Tag SKUs where fit uncertainty dominates for product-page fixes. Feed the bucket shares into the bracketing model so proxy orders stop inflating customer risk scores. Put one number in the monthly returns report: the share of multi-size orders explained by each bucket, and how that share moves as you fix things.

Run the survey continuously, not as a one-off project. The bucket mix shifts with seasons, categories, and assortment changes, and a standing survey catches the shift while it is still cheap to act on. Asking costs almost nothing. Guessing costs margin on every bracketed order you misdiagnose.

## Frequently asked questions

### Do post purchase fit surveys reduce bracketing?

They help when the answers are used. A short survey after delivery that asks which size was kept and why feeds size guidance and fit notes that future shoppers actually see. Surveys collected and never wired into product pages change nothing.

### When should a fit survey be sent?

A few days after delivery, while both sizes may still be in the house and the fit decision is fresh. Surveys sent weeks later get low response and unreliable memories.

### What should a fit survey ask?

Which size was kept, whether the kept size fits as expected, and one question about the dimension that decided it, such as length or waist. Short surveys with specific questions outperform long generic ones on both response rate and usefulness.