CRO Is Not a Design Opinion Poll

Your conversion rate doesn’t care what you like. It cares what your customers understand.

Every CRO client says the same thing upfront:

“We don’t do those silly button color tests.”

And yet…

Many still treat conversion optimization like a lottery ticket. Spin the wheel. Hope for green a green number.

The truth?

Most superficial tests aren’t wrong. They’re just left to random chance.

Color tweaks, button sizes, changing a headline, these can work, but only if they address the real friction points: the ones killing confidence, clarity, and momentum.

That’s where meaningful uplift lives.

That’s where the revenue is hiding.

Win of the Week:

Problem: New visitors were highlighting product titles on PDPs, an often-overlooked behavior we’ve come to associate with users searching for discount codes or better prices.

Discovery: Our team spotted this pattern in session replays and quantified it using behavior flags. It was a subtle but powerful signal of hesitation.

Hypothesis: We believe showing an exit intent modal when a visitor highlights the product title on a PDP reinforcing the existing price or offering them a discount to buy it now will increase cart adds.

Test: When a desktop user highlighted the product title with their mouse, a persistent red banner appeared reminding them of the discount they were already getting.

Result:

  • +3.65% lift in add-to-carts

  • +4.65% lift in transactions (new visitors only)

But, the real win was more nuanced.

New visitors responded with:

  • +57% lift in add-to-carts

  • +90% lift in cart visits

  • +83% lift in checkout visits

  • +112% lift in transactions

And, Returning visitors with:

  • -18% drop in add-to-carts

  • -14% drop in cart visits

  • -16% drop in checkout visits

  • -26% drop in transactions

Lesson: Behavior signals often whisper, not shout. Conversion opportunities live in the in-between moments, like highlighting text.

But don’t stop there, this is something you don’t want to show to your returning or logged in visitors, but if you’re leveraging discounts your returning visitors might love it.

How to Read Session Recordings Like a Strategist

Session recordings are gold, but only if you are starting from a place of curiousity.

Don’t just watch to confirm your assumptions. Watch to spot the invisible friction, signals of hesitation, confusion, or unmet needs.

Look for these patterns:

Hesitation & Looping Behavior

  • Hovering between two elements

  • Re-reading the same section

  • Scrolling up and down in a tight zone

  • Repeated visits to shipping, size, or returns info

  • Back-and-forth PDP to category flow

Clicks That Go Nowhere

  • Clicking non-clickable product names or images

  • Clicking static trust badges expecting info

  • Double-clicking CTAs (expecting something to happen faster)

Highlighting & Copying

  • Product titles = likely price shopping

  • Return policy = seeking risk reduction

  • Copying size info = searching for clarity or comparing

  • Opening nav or filters multiple times without choosing

  • Footer scrolling just to find policies or links to alternative pages

  • Using site search to find content already on the page

Dead Ends

  • Add to cart but no progression

  • Zero interaction followed by bounce

  • Clicking a button, then doing nothing

When you see these behaviors across multiple sessions, you’re not watching anomalies. You’re observing friction.

Quote of the week:

It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts.

Arthur Conan Doyle, Sherlock Holmes

Stop trying to optimize what you think is a good idea.

Start fixing what’s actually breaking the sale.

Most “CRO” work is just surface-level tinkering.

But conversion lives deeper, in the confusion, hesitation, and missing clarity your team stopped noticing.

Go find that.

How valuable was this week's newsletter?

Login or Subscribe to participate in polls.

P.S. Want to turn confusion into conversions? Let’s talk.