Cart Drawer Rebuild Case Study: A Simpler Drawer With an Order Recap Lifted Revenue per Visitor 4.7%

Cart Drawer Rebuild +4.7% revenue per visitor

[ FREE CRO TEARDOWN ]

Find the 3 biggest revenue leaks on your store.

Every day a conversion leak goes unfixed, you're paying for traffic that doesn't buy. Get a 5-minute Loom through your PDP, cart, and checkout, with mockups of the fixes. No pitch.

Get My Teardown
or
Book a Call

The cart drawer change that moved the number was not a trust badge or a timer. It was a short box telling shoppers what they already get with their order.

We rebuilt the drawer for a DTC intimate wellness brand. After 14 days, the version with that recap box was ahead: +4.7% revenue per visitor, with an 87% chance to beat control. That is a strong lean, not proof. Below is what we built, how we read it, and why it is not live for everyone yet.

How we measured it: the numbers count only visitors who opened the cart drawer, tracked with a custom event, not every visitor to the store. The test only changed the drawer, so that is the group it could affect.

The Problem: A Drawer That Answered Nothing

This brand sells products people care about buying quietly. Before they pay, shoppers want three answers: will the package be discreet, when will it arrive, and is it covered if something breaks.

The old drawer answered none of that. It showed payment icons, a countdown timer on a free gift, a few checkboxes for order protection, and a savings line sitting apart from the subtotal. Most people who added to cart never reached checkout.

Our bet: if the drawer answers those worries and shows what the shopper is already getting, more carts reach checkout.

One Drawer, Built in Four Layers

We did not test five unrelated drawers. Each variation kept everything from the one before it and added one more layer. That way, if a later step pulled ahead, we could see which layer did it.

Version What changed in the cart drawer
Control The drawer as it was
Variation 1 A trust row (privacy, delivery, warranty) instead of the payment icons
Variation 2 Variation 1, plus the free gift countdown timer removed
Variation 3 Variation 2, plus a "With your order" box under the products, a single "Shipping protection" switch in the footer, and the savings merged into the subtotal line
Variation 4 Variation 3 without the box, plus a two-step bar (free gift, then free shipping) and the brand's own upsell card

Setup: control + 4 variations in Intelligems, sitewide (the drawer opens on every page), desktop and mobile. Main metric: revenue per visitor.

What the "With your order" box said

Three short lines under the cart items: what ships with the order, how fast it arrives, and how much the shopper saves. No new offer. Just the things that were already true, said in the one place people look right before they pay.

Why the savings moved into the subtotal

Two separate lines (savings, then subtotal) make the shopper do math. One line with the old price crossed out, the saving, and the final total gives the same facts in one glance.

The Read: Variation 3 Ahead From Day One

At the 14-day read we sent the client, Variation 3 led for visitors who opened the cart drawer:

Metric Variation 3 vs control
Revenue per visitor +4.7%
Chance to beat control 87%
Test length 14 days

The lead was not a late spike. The trend was positive from the start and stayed that way the whole test. That matters more to us than a single snapshot.

Variation 3 was the only one we called ahead. The trust row and the timer removal sit inside it, but on their own steps they did not stand out. So the safe reading is this: the leak in this drawer was not about badges or urgency. It was about clarity on what the shopper gets.

What 87% Means, Honestly

87% likely to beat control means the data leans clearly one way. It does not mean the lift is locked in at exactly 4.7%. The real number could be smaller, and there is still a real chance the change does nothing.

We call it a leader, not a proven winner. For a change this low-risk (no new offer, no price change, nothing removed that buyers need), a strong lean is a fair reason to move forward. For a risky change, like a new price, we would want more.

Why It Is Not Live for Everyone Yet

Mid-test, the product behind the brand's upsell card went out of stock. Variation 4 lost its card with no warning, so we paused it for new visitors and let the other versions finish.

When the product came back, the brand had a real question: ship Variation 3 without their upsell card, or keep the card? Instead of guessing, they chose a re-test. Variation 3 is now the new control, running against Variation 4 with the card back in stock. Whichever wins that test goes live.

Two lessons we took from this:

  • If a variation depends on a product being in stock, get a stock promise before launch
  • Read a paused variation only up to the day it was paused, or the late, returning visitors skew it

Steal This for Your Own Drawer

  • List the 2-3 worries buyers have at checkout for this product, and answer them inside the cart drawer
  • Recap what is already in the order before you add anything new
  • Merge savings and subtotal into one line
  • Stack variations so each one adds a single layer, then you know which layer did the work
  • Test trust badges and timers before you assume they help, here they did not stand out

For the wider list of cart drawer tests worth running, see our Shopify cart drawer optimization guide.

Questions About Rebuilding a Cart Drawer

What lifted revenue per visitor in this cart drawer test?

The version with a short "With your order" recap box, a single shipping protection switch and one merged savings and subtotal line. It led with +4.7% revenue per visitor for visitors who opened the cart drawer, 87% likely to beat control, over 14 days.

Is 87% probability to beat control enough to call a winner?

It is a clear lean, not proof. The real lift could be smaller than 4.7%. For a low-risk change with no new offer or price change, it is a fair reason to move forward. For a risky change, we would want more data.

Do trust badges in the cart drawer increase conversions?

Not in this test. Swapping payment icons for a privacy, delivery and warranty row did not stand out on its own. The lift came when the drawer also recapped what the shopper already gets.

Why test the changes stacked instead of separately?

Each variation kept the one before it and added one layer. If a later step pulls ahead, you can see which layer made the difference, without running five separate tests.

Why was the winning cart drawer not shipped right away?

The brand's upsell card was out of stock during the test, so it never got a fair run. With the product back, the brand chose to re-test the leading version against the version with the card. The winner of that test goes live.

[ SAY HI AND LET'S MAKE YOU SOME MONEY ]