Experiments that improve every inch of your store.
From product placement to promotional mechanics to checkout flow — Decision Process helps retail operators answer the questions that move revenue, not just website metrics.
Experiment Templates
Ready-to-run experiments
Product Placement A/B
“End-cap placement increases weekly unit sales vs. standard aisle shelf.”
Conditions
- Aisle shelf (control)
- End-cap display
Metrics
- Units sold / week
- Revenue per SKU
Pricing Tier Comparison
“A price reduction of 10% improves margin-weighted volume.”
Conditions
- Standard price (control)
- −5% price
- −10% price
Metrics
- Units sold
- Gross margin / unit
- Revenue
Promotional Mechanics Test
“BOGO promotions outperform percentage discounts for basket-size growth.”
Conditions
- No promotion (control)
- 20% off
- Buy 2 get 1 free
Metrics
- Basket size ($)
- Units per transaction
- Promo redemption rate
Shelf-Height Experiment
“Eye-level placement increases grab rate for impulse items.”
Conditions
- Bottom shelf
- Middle shelf (control)
- Eye-level
Metrics
- Units grabbed per 100 passes
- Revenue per facing
Loyalty Reward Structure A/B
“Points-per-dollar rewards drive more repeat visits than tiered status rewards.”
Conditions
- Tiered status (control)
- Points per dollar
Metrics
- Visit frequency / month
- Avg spend per visit
- Churn rate (30d)
Worked Example
End-cap vs. aisle placement across 20 stores
A grocery chain tests two placement strategies for a new snack SKU across 20 stores. 10 stores get end-cap placement, 10 keep the standard aisle shelf. Weekly unit sales are recorded for 8 weeks.
Results: units_sold (weekly)
Aisle shelf (control)
mean: 147 units/wk
95% CI: 131–163
End-cap display
mean: 168 units/wk
95% CI: 152–184
P(end-cap better) = 96%
End-cap placement produces a +14% lift in weekly unit sales (d = +0.71, large effect). At 96% posterior probability, this exceeds the retailer's 90% decision threshold. Recommendation: roll out end-cap placement chain-wide.
Explore other domains
Run your first retail & grocery experiment
Private beta — tell us about your use case and we'll get you set up.