Synthetic data
All sessions

s-0491

Simulated shopper

Oct 25, 2026, 2:55 PM UTC · desktop · email · Ridgeback 38L Pack · 10 actions

abandoned

How Qualm's guess changed

With the current settings, the store would have shown Decision Helper after action 5 (dashed line). Hover the chart to see each step.

FitSuitabilityDurabilityValueDeliveryChoiceNo clear doubt
Chance of buying (separate abandonment model)10% → 13%

Step by step

Each action, the clue Qualm took from it, and its guess afterwards.

  1. 123sBrowsed catalogBrowsed the catalogNo clear doubt 26%
  2. 244sViewed Cairn MidViewed a product pageNo clear doubt 34%
  3. 31m 17sImage 1 · Cairn MidViewed a product imageNo clear doubt 56%
  4. 41m 43sSearch “best backpacking pack”Searched to compare optionsChoice 78%
  5. 52m 20sOpened comparisonOpened the comparison→ the store shows Decision Helper hereChoice >99%
  6. 62m 43sViewed Ridgeback 48LViewed a product pageChoice >99%
  7. 72m 57sViewed Ridgeback 38LViewed a product page · Switched to a similar productChoice >99%
  8. 83m 33sExpanded specs · Ridgeback 38LExpanded technical specsChoice >99%
  9. 93m 37sImage 2 · Ridgeback 38LViewed a product imageChoice >99%
  10. 3m 47sLeft the store

Final guess

Choice>99%
Fit
<1%
Suitability
<1%
Durability
<1%
Value
<1%
Delivery
<1%
Choice
>99%
No clear doubt
<1%
Why choicehow much likelier
  • Opened the comparison57.4×
  • Searched to compare options39.5×
  • Switched to a similar product8.6×
  • Viewed a product page ×31.4×

What the shopper told us

Their survey answer, if any. For simulated shoppers you can also reveal what was really true.

Survey answer
Choice
exit survey · held out (calib)
Reveal synthetic ground truth
Primary doubtChoice
Secondarynone
Resolved on its ownno

Only available because this session was simulated. Real sessions have no ground truth.