Think Build Implement Repeat
London, UK +44 7367 067226
WhatsApp FOLLOW f in X
  1. Home
  2. Blog
  3. Predicting Returns at the Point of Sale
AI & Machine Learning

Predicting Returns at the Point of Sale

Returns are a large and largely accepted cost. What can be predicted at checkout, and the uses that help customers rather than penalise them.

Updated 2 min readBy SpiderHunts Technologies

Free estimateNo obligation

Get a free estimate

Tell us what you need. A senior engineer reads every enquiry.

Takes under a minute. We never share your details.

  • Free consultation
  • No commitment
  • NDA on request

Prefer to talk? Book a free 30-minute call →

Quick answer — TL;DR

Return likelihood is predictable from product, customer and basket signals. The productive uses are fixing the causes - sizing guidance, better imagery, description accuracy - rather than restricting customers, which damages the relationship for modest savings.

A cost treated as weather

In several categories returns run at a level that would be treated as a crisis anywhere else in the business, and they are accepted as a cost of trading. The processing, shipping, inspection and markdown on returned goods are all real.

Some of it is genuinely inherent - customers buying two sizes intending to return one. A meaningful share is not, and is caused by something fixable.

What predicts a return

  • The product itself - some lines return far more than their category average
  • Size and fit, where the product runs differently from the customer's usual
  • Customer history - past return rate is a strong signal
  • Basket composition, particularly multiple sizes of one item
  • Discount depth, where deep discounts attract more speculative buying
  • Delivery time, where a long wait increases second thoughts

Product-level patterns are the most actionable. A line returning at several times its category average usually has a specific cause - the colour differs from the photograph, the sizing is off, the description is misleading - and that cause is fixable once identified.

Use it to fix causes, not to restrict customers

UseAssessment
Flag products with unusual return rates for reviewBest return on effort
Improve sizing guidance on problem linesDirectly reduces the cause
Add imagery or detail where descriptions misleadDirectly reduces the cause
Plan reverse logistics capacityOperationally useful
Forecast net sales for stock planningUseful and uncontroversial
Restrict or charge high-return customersHigh risk to relationships and reputation

That last row deserves care. Customers with high return rates are frequently high-value customers who buy a lot, and restricting them can cost more than the returns. It also generates the kind of publicity nobody wants.

Returns and demand are different numbers

For planning, what matters is net sales. A line selling a thousand with three hundred coming back needs stock planned on seven hundred, and the returned units re-entering stock have their own timing.

Businesses that forecast gross sales and treat returns as a separate write-off consistently mis-plan. Modelling the return rate per line and the return timing makes the stock position considerably more accurate.

The reason codes are worth fixing

Return reason data is usually poor - a dropdown where 'no longer wanted' absorbs everything because it is the quickest option. That destroys the most useful signal available.

Improving reason capture, even slightly, pays for itself. Fewer options, better wording, and asking at the right moment produce data that identifies fixable causes. Without it you know products come back but not why, which is the question that matters.

A product returning at three times the category average is telling you something specific. The reason codes are how you find out what.

FAQ

Frequently asked questions

The questions readers ask us after this guide.

Still have a question?

Ask us directly — a senior engineer will get back to you.

Ask about your project

How much history is needed?

Enough returns per product type to see patterns. Product-level analysis needs reasonable volume per line; customer-level needs repeat purchasers.

Should we ever refuse a return?

That is a policy question with legal constraints depending on jurisdiction. Prediction is better used to reduce the cause than to police customers.

Can we predict returns in store as well as online?

Where transactions are linked to a customer, yes. Anonymous cash sales carry much less signal.

Does free returns policy affect this?

Substantially - it changes buying behaviour. Any model trained under one policy will not transfer cleanly to another.

Keep reading

More on AI & Machine Learning

Start here

Want machine learning project details from us?

Tell us what you are trying to predict and roughly what data you hold. We will come back with an honest view on whether machine learning is the right tool, what the work would involve and a realistic cost range. If a spreadsheet would do the job, we will say so.

  1. You tell us what you needTwo minutes on the form, or a message on WhatsApp.
  2. A senior engineer reviews itAnd comes back with questions, a realistic range and an honest view on fit.
  3. Free 30-minute scoping callWe talk through scope, options and a realistic estimate — with no obligation.
Free estimateNo obligation

Talk to someone who builds this

Send a short brief and we will come back with an honest view and a realistic range.

Takes under a minute. We never share your details.

  • Free consultation
  • No commitment
  • NDA on request

Prefer to talk? Book a free 30-minute call →