· Men's grooming DTC We rebuilt the funnel for Particle's Meta listicle traffic. The new path beat the existing one at 99.53% confidence.
Same ad. Same listicle. Same traffic. We changed the page they landed on.
The only variable in the test was the destination. On the left, the existing product page. On the right, the offer page we built. Scroll either panel to see the full page.
Controlled A/B test in VWO. Same SKU, same traffic source, 12,893 visitors. The offer page won on revenue per visitor at 99.53% confidence.
We've worked with Jacob for over a year and a half. The biggest win was a dedicated offer page he built to replace what we were sending our Meta traffic to. It beat our existing path at 99.5% confidence. Beyond that, more than 35 test wins across the site. He comes up with the ideas, builds almost everything himself, runs the tests, and shows us what's working. Hands off for us, which is exactly what we needed. If you're on the fence, just do it.
Real Meta spend. The funnel that caught it was not built for the visitor.
Particle was buying meaningful Meta traffic, warming the visitors through a listicle, and then routing every one of them through a funnel that ended on a generic product page.
The architecture had been inherited, not designed. The product page at the end of the path was built to merchandise the catalog, not to convert one specific visitor who had just read one specific story about one specific product.
The team suspected the funnel was leaking revenue. But a suspicion is not a number. Nobody wanted to rebuild the funnel that caught the brand's most expensive traffic on a hunch, because if the hunch was wrong they would have torched conversions across paid Meta.
A purpose-built funnel for one specific kind of visitor.
We designed and built a new funnel for the Meta listicle traffic. The listicle stayed. The destination changed. Instead of dumping that visitor onto a generic product page, we routed them into a dedicated offer page, in the style of the high-converting offer pages used by brands like Javy.
Copy, layout, offer framing, proof, pricing logic, upsell. Built end to end by Shivook. The brand did not have to wireframe, scope, or chase deliverables. We shipped the funnel.
New funnel head to head against the old one. The visitors decided.
A controlled A/B test in VWO. Same product. Same SKU. Same traffic source. Half the visitors went through the new funnel. Half stayed in the existing path.
At 12,893 visitors the new funnel was a winner on both revenue per visitor and conversion rate, each at its own confidence interval. The brand rolled the new funnel out to all of that traffic.
Put plainly: for every dollar the existing funnel earned per visitor, the new funnel earned a dollar and twenty one cents.
A new funnel is not automatically better than the old one. The only way to know is the test.
We did not guess which funnel would win. We built the alternative and proved it head to head on real traffic, before the brand bet another dollar of ad spend on it.
That is the whole model. The funnel your paid traffic flows through is usually the highest leverage and least tested asset in your store. Sometimes the existing path wins. The only way to know is a controlled test on live traffic. That is also what removes the risk. You keep what wins. You cut what loses. You never bet the store on an opinion.
Running meaningful paid traffic and not certain it is landing on the page that converts best?
That is the exact question we answer. With a controlled test, not a hunch.