Every asset gets better, automatically.
A/B test layouts and content, let statistical significance declare the winner, and start the next iteration — all without a developer or a manual check-in.
Six ways the platform optimises itself.
Testing, personalisation, and revenue attribution — built into the same widget runtime, not bolted on via a third-party script.
A/B testing of media assets
Run two widget variants simultaneously — different layouts, different content selections, different product orderings. Traffic splits automatically and both variants track the same conversion events so the comparison is apples-to-apples.
Automatic winner selection
Idukki applies Welch's z-test for statistical significance. Once a variant clears the confidence threshold you configure, the system auto-declares a winner and routes all traffic to it — no manual check-in required.
Personalisation by segment
Serve different galleries to returning versus new visitors, to mobile versus desktop, or to shoppers who have already added to cart. Segment rules slot into the same widget configuration — no separate personalisation tool.
Continuous optimisation
Once a test concludes, you can immediately start the next one from the winning variant as the new baseline. Over successive cycles, the widget configuration converges toward what actually converts for your audience.
Revenue per variant
Every test tracks revenue attribution at the variant level via UTM auto-stitching and GA4 plus Meta Pixel forwarding. You see exactly which version of the widget earned which revenue — not just clicks or impressions.
No-code test setup
Duplicate an existing widget, change the layout or content selection, toggle "A/B test" mode, and set a minimum impression count. The test runs against your live store without a developer or a new deploy.
Statistical test used for significance. No cherry-picked confidence windows — significance is calculated continuously.
Widget layouts that can be A/B tested against each other within the same Idukki account, all on one flat price.
Result polling interval. The dashboard reflects live conversion data as each impression arrives.
One test cycle, from setup to winner.
No engineering sprint to launch a test. No data analyst to call the winner. The platform handles both.
Pick your control and variant
Choose any live widget as the control. Clone it into a variant and change whatever you want to test — layout style, content source, product ordering, CTA copy. The rest of the configuration is inherited automatically.
Set the rules and let it run
Define the traffic split, the minimum impressions before a winner can be declared, and the primary metric (add-to-cart, revenue, click-through). The test starts on publish and runs until significance is reached.
Winner ships, next test starts
Idukki auto-promotes the winner and archives the loser. Open the test report, review which content drove the result, and kick off the next iteration — the winning config becomes your new baseline.
Part of the Idukki platform
Testing included. No extra seat.
A/B testing, segment personalisation, and revenue attribution are built into every Idukki plan. You do not pay per test, per variant, or per conversion event tracked.
- Revenue per variant via UTM auto-stitching and GA4 plus Meta Pixel forwarding
- Segment by device, visit count, cart status, or custom URL parameter
- No minimum contract — start testing on the trial plan
Stop renting six tools. Ship one.
Spin up your first widget in 4 minutes. Migrate from Bazaarvoice, EmbedSocial, Tolstoy or Videowise in a day. Your CFO will love us. Your CRO already does.
- No credit card
- Cancel anytime
- SOC 2 + GDPR
