Every area of your site, and every team, tests differently.
By site area: what to test there, then the demo store’s tests, ideas, and lessons. By profile: what Probant changes in each team’s work.
- Free shipping progress barCartConclusive. Proven gain61,240 visitors+6.0%, range +2.2% to +10.0%
- One-page checkoutCheckoutSample ratio mismatch.19,800 visitorsHidden until diagnosed
- Simplified mobile menuNavigationNot conclusive yet…day 9 of 13+1.9%, range -3.7% to +7.8%
- Pay-in-3 option highlightedProduct pageNot conclusive yet…day 6 of 17+3.5%, range -3.4% to +11.0%
- Customer reviews at the top of the pageProduct pageNo detectable effect.64,100 visitors-0.4%, range -4.0% to +3.3%
- “You may also like” recommendationsProduct pageStopped by a guardrail.stopped 3 h agoReturns +1.2 pts
- Orbe product page: title, photo, and badgeProduct pageConclusive. Proven gain80,512 visitors+26.6%, range +13.9% to +41.0%
- Sort by popularityCategory pageNot conclusive yet…day 3 of 14+1.5%, range -2.8% to +6.1%
Maison Lune, a fictional demo store · data computed by the engine
Each area of the site has its own tests and pitfalls.
Judge the cart on margin, the checkout on its traffic split, and category pages beyond the click. Each page starts from the demo store.
Cart and shippingFree-shipping threshold, progress bar, fees.Demo store · most advanced testFree shipping progress barConclusive. Proven gain+6.0%, range +2.2% to +10.0%61,240 visitors · 3 pending ideas · 8 completed testsSee this area Product pageReviews, installment payments, recommendations.Demo store · most advanced testOrbe product page: title, photo, and badgeConclusive. Proven gain+26.6%, range +13.9% to +41.0%80,512 visitors · 7 pending ideas · 12 completed testsSee this area CheckoutSteps, trust signals, one-page checkout.Demo store · most advanced testOne-page checkoutSample ratio mismatch.Hidden until diagnosed19,800 visitors · 3 pending ideas · 5 completed testsSee this area Mobile navigationMenu, filters, and search on small screens.Demo store · most advanced testSimplified mobile menuNot conclusive yet…+1.9%, range -3.7% to +7.8%day 9 of 13 · 2 pending ideas · 7 completed testsSee this area Category pagesSorting, filters, and featured products.Demo store · most advanced testSort by popularityNot conclusive yet…+1.5%, range -2.8% to +6.1%day 3 of 14 · 3 pending ideas · 4 completed testsSee this area
Each team comes looking for a different answer.
Founders, CRO teams, developers, or agencies: the same reports, read with different questions.
Store owners and foundersKnow what pays off, without a statistician.
- A one-sentence verdict, with revenue and margin in euros
- Test length known before launch
- Variants without a developer
- Planned checkpoints instead of peeking
- Gains corrected for the winner’s curse
- Ideas ranked from past tests
- Server-side assignment, no flicker
- Events API
- Rollout steps with automatic rollback
- Checks listed in every report
- Published methods with their sources
- A demo your clients can open
- A hidden field links each lead to its visitor
- Qualified and signed leads sent back by your CRM
- Lead contact details stay in your CRM
- Stripe connected directly
- MRR and retention at day N
- Follow-up flags when a decision reverses
The examples come from a fictional store.
Maison Lune, an online lighting store, doesn’t exist. Its input data is made up; everything derived from it (lifts, ranges, checks, durations, amounts) is computed by the Probant engine, as it would be for a real store.
The Demo page describes what the demo app contains. No sign-up needed.
Conclusive Or not yet…
Create your workspace, or explore the demo first with data from a fictional store.
- Open demo, no sign-up
- Free to start, no overage charges
- Database and compute in Amsterdam, with Scaleway