Anya Geimanson

@O_oPalych Exactly! Important note: the principle makes sense, but outcomes will always vary by brand and audience. So never assume that what worked for another brand will work for yours 👉 Always A/B test.

@AnyaGeimanson @Shopify Interesting observation from a marketing perspective — content visualization and user retention. Plus, expanding the lineup and immediately offering other options provides a significant portion of the benefit.

A @Shopify brand grew revenue 6% by moving ONE thing HIGHER on their product page They A/B tested moving product recommendations to the top of the PDP   (for new shoppers landing on PDPs from ads). That’s it. IT WORKS FOR 2 REASONS: 1️⃣ The “low-price click” effect Some shoppers land from ads because the low price pulled them in. But if you show them better / slightly pricier alternatives immediately → they often upgrade. RESULT: higher AOV 2️⃣ The first seconds are the decision window. More options early → more chances to win the click. RESULT: higher conversion rate — I added a quick video in the comments showing how to set up this A/B test on @Visuallyio so it targets only new visitors who land on a PDP.

@Shopify Here's how you can setup the correct audience for this one

A @Shopify brand grew conversion rate 5.6% by making ONE thing 5% smaller. They narrowed the cart drawer. That’s it. WHY IT WORKED: • The old layout had a “Continue Shopping” button • People kept clicking it → and not checking out Shrinking the drawer made it obvious: You can keep browsing without being “invited” to leave checkout mode. Less distraction → more checkout focus → higher conversion. IMPORTANT NOTE: If your average cart is ~1 item, you may want to test removing “Continue Shopping” cues entirely. That means keeping the cart drawer at full width, and pushing a single path: Checkout. Don’t assume. A/B test it.

A @Shopify brand lifted conversion rate 6.2% by removing ONE thing from their collection page. They removed the banner. That’s it. WHY IT WORKS: Collection pages are for finding products, not branding. A big static hero pushes the grid down (especially on mobile) → fewer products seen → fewer product page views. Sometimes: no banner → more products above the fold → more PDP clicks → higher conversion. IMPORTANT TO NOTE: If your banner adds real decision help (quick filters, promo/shipping/returns info), it may be worth keeping. This is mainly for the static “brand-y” banner. Don’t assume. A/B test it.

A @Shopify brand grew revenue 5.8% by removing ONE thing from the sticky Add to Cart. They removed the price. That’s it. WHY IT WORKS: • When shoppers are deep-scrolling, they’re building confidence. • A sticky bar flashing $265 can snap them back into hesitation mode. Sometimes: no price → fewer second guesses → more Add to Carts. IMPORTANT TO NOTE: This often works best on higher-ticket items. (Although Amazon also hides price in sticky CTAs) A/B Test it. Don’t assume.

If your average items per order is under 1.3 - Here’s a 9.9% revenue lift you might be missing: 👉 Test adding a “Buy Now” button that skips the cart and sends shoppers straight to checkout. Sounds simple? Because it is. For single-item purchase brands, the cart step is just friction. It adds time. Adds distraction. Adds drop-off. Removing it doesn’t just smooth the experience, It directly impacts conversion. Takes minutes to test with @Visuallyio.

A @Shopify brand grew revenue 4.6% by turning ONE thing red. They changed the cart numerator. That’s it. WHY IT WORKS: Red pulls the eye to the cart, the closest step to the checkout. - No lost momentums - No re-hunting past picks More shoppers move forward → more checkouts. BONUS TIP: Pair the red numerator with a subtle animation.

A @Shopify brand grew revenue 5.2% by making ONE thing in the cart 60% bigger. They enlarged the product image. That’s it. WHY IT WORKS: Bigger visual = faster reassurance - No squinting. - No second-guessing. - No bouncing back to PDP to double-check. Less friction → more people actually hit checkout. BONUS TIP: Add a zoom button. Let them see every detail before they commit.

@codyplof @Visuallyio Sure you could, just like you could with dev. The question is at what speed/effort. Plus, if you’re building experiences that are personalized, where and how you are managing your audiences. Lmk if your want to check Visually out for a few months free. It’s a simple app install.

@AnyaGeimanson @Visuallyio That’s cool. I was able to built almost all of it in Claude code and close tin finishing the rest.

I've been determined to use claude code to be able to create pixel perfect designs and I had a major breakthrough today. I'm fairly confident we're have a human in the loop AI CRO system in the next month that can be done end to end, at a super high level, with one mid level person.

@codyplof @Visuallyio

I love this - Claude Code is getting really close to pixel-perfect CRO… but the missing piece is the last mile: knowing your Shopify context (theme, catalog, promos, shopper behavior) and actually shipping safely. That’s exactly why we built @Visuallyio Copilot: one-click install, zero code, it analyzes your store in real time, then you type what you want and it turns it into requirements → design → production-ready code → deploy (and you can A/B test it). It’s the difference between a great mock and a live a/b test in minutes - here are some real life examples.

@codyplof @Visuallyio

@codyplof @Visuallyio

@codyplof @Visuallyio

@DavidMagee @hazan0608 @Visuallyio ty for the kind words @DavidMagee !

@hazan0608 No so much for audit but for a killer platform and team - you gotta speak to @AnyaGeimanson and the @Visuallyio gang. Next level for Shopify.

Who is a CRO god for ecom? Need an audit of our site

@DylanAnder That sucks 💙 Not sure why, but I'm still surprised when these things happen.

I CAN'T STAY SILENT ANYMORE. I'M BEING PUBLICALLY DISCRIMINATED AGAINST. This is not a political post, but I need to share how I'm getting targeted IN THE ECOMMERCE COMMUNITY nonetheless. I've done countless public webinars with ZERO pitches, and THREE TIMES PEOPLE HACKED OUR ZOOM AND PUT UP SWASTIKAS. Happened again today. Today was my last straw on not speaking up. I was born into a beautiful Jewish family that I love very much. There's a lot going on in the world right now. It's complicated and not easy for anyone. This is about how we treat our fellow humans. I'm just a dude trying to help. Why are people adding swastikas to my fucking online events??? Anyone denying antisemitism is oblivious, but I also don't blame them if they don't have exposure to antisemitism. I support anyone with a positive intention, whether I agree or disagree with their perspective of the world. This is not just about antisemitism, it's ALL discrimination. If you see it but don't speak up, you're part of the problem. We need to condemn any form of discrimination. I was born to a family of immigrants who had to flee for their lives to the United States 60-70 years ago. If you see antisemitism or any descrimination, speak the fuck up, for all of us. I'd never expect this from the Ecommerce community, nonetheless... It's shaped who I am, I give back everywhere I can, and being born the way I was is no means of stopping me in supporting the community. Lastly, don't pity me. I'm just raising awareness about how much antisemitism exists in the world. Also, how is @Zoom so insecure that it can be hacked at any time, and names get randomized??? Am I the only one getting targeted out there?

This is how top @Shopify brands test for real business impact. We built Visually to make it simple. Memorial Day is coming. You ought to test smarter.

✅ Fix it: 1️⃣ Mirror onsite offers Match your site experience to each promo using UTM parameters (e.g. ?utm_content=20off and ?utm_content=20dollar). 2️⃣ Safeguard direct traffic Split non-UTM visitors 50/50 with an onsite A/B test. No more built-in bias. 3️⃣ Track what matters → CR, AOV & profit per user in Visually → CAC / CPA in your ad platform

Most promo tests are built to fail. Why? Because your test setup - not your offer - is killing profit. Here’s how to stop your A/B test from lying to you (and stealing 30% of your margin):