Qualm — Know why they didn't buy

The problem

of shoppers who leave answer the exit survey.

Everyone else leaves without a word. So stores guess why, the guess is usually “too expensive”, and the fix is a discount.

But someone unsure whether a boot runs small doesn't need 10% off. They need to know their size. A discount spends margin and answers the wrong question.

Figure from the demo store's simulated shoppers.

Two hikers walking toward the camera on a narrow trail through an alpine meadow — Photo: Bob Wick / Bureau of Land Management, public domain

How it works

Clicks are clues.

01
Shopper browses
Every action on the page, like opening the size guide or checking a delivery date, becomes a named clue. No personal data.
02
Qualm works out the worry
A model trained on the few survey answers weighs the clues and gives a probability for each of six worries, with its reasons.
03
The store answers it
Once Qualm is confident, the page shows one card for that worry, built only from the product's own data.
04
The owner sees the pattern
The merchant console shows which products lose sales to which worry, and plans a test to prove the cards work.

Live output from the engine for the scripted shopper “Unsure about fit”: boot shopper checks sizing twice and reads fit reviews.

What's holding this shopper back

Fit>99%

“Will it fit me?”

Value<1%

No clear doubt<1%

Why Qualm thinks so

Changed size selection107× likelier
Re-opened the size guide90× likelier
Filtered reviews by fit24× likelier

What the shopper sees on Cairn Mid Waterproof Boot

Fit Confidence

Runs small: most shoppers go up a half size

Fit feedback: 62% small · 33% true · 5% large
Based on: 157 reviews that mention fit
Exchanges: Free size exchanges for 60 days

Your usual size→ we suggest10.5

Built from: Fit feedback from 157 reviews · Returns policy

Watch it happen, click by click

Six worries. Six different fixes.

A discount only answers one of them. Qualm tells the store which one it's facing on each product, and shows the matching answer to the shoppers who need it. Everyone else sees nothing extra.

Fit
“Will it fit me?”
Clue
Opens the size guide twice
Card
Fit Confidence
Suitability
“Will it work for what I need?”
Clue
Filters reviews by performance
Card
Built-For Answer
Durability
“Will it hold up?”
Clue
Reads the warranty
Card
Durability Proof
Value
“Is it worth the price?”
Clue
Sorts by lowest price
Card
Value Breakdown
Delivery
“When will it arrive, and what will it cost?”
Clue
Checks the arrival date
Card
Delivery Promise
Choice
“Which one should I pick?”
Clue
Bounces between two jackets
Card
Decision Helper

A group of hikers and a dog crossing a wildflower meadow below snowy peaks

One model. Both sides of the counter.

The shopper side

Doubt Lens

Shop a pretend outdoor store while a live panel shows Qualm's guess and reasons. Or press play on one of three scripted shoppers.

Start shopping

The store-owner side

Merchant console

See which products lose sales to which worry, choose when each card appears, and size a real test to check it works.

Open the Doubt Map

Photo: Bob Wick / Bureau of Land Management, public domain

Measured, not claimed.

Scored on 4,000 simulated shoppers, where the true worry is known. Real traffic will be messier, and Qualm makes no claim about sales until a real A/B test runs.

Worry identified correctly: 84.5%; vs 67.1% for “blame the last click”
Error on delivery worries: 0.4 pt; vs 5.1 pt for the exit survey alone
Calibration error: 0.05; “80% sure” is right about 80% of the time

How the model works

A turquoise alpine lake below snow-streaked peaks, with yellow wildflowers in the foreground

See it work in one minute.