Ecommerce Case Studies: 7 Blueprints With Real Data

Every result below started as a growth problem and a specific fix. This guide is built as a set of blueprints: what each brand changed, the number it changed, and the steps to apply the same idea on your store.

Each example below is a brand we have full campaign data for, so you get the number and the mechanics behind it, from AI recommendations to a routing quiz for a services business.

Choose your goal

Pick what you are trying to fix, then jump to the brand that did it.

Pierre Hardy: AI recommendations that drove 22% of revenue

Pierre Hardy sells luxury footwear and accessories, a catalog where a sitewide discount does more damage to the brand than it earns back. So the growth question was narrow: how do you get more from the same traffic without touching price?

The team used AI product recommendations, but matched each one to what they already knew about the shopper. A first-time visitor has no history, so they saw the pieces most people buy. A shopper with something in the cart saw items bought alongside it, while they were still deciding. A returning visitor saw what they had looked at last time.

None of it relied on a discount. Over a year, those recommendations influenced 22% of online revenue, and the feed carrying them held an 11.6% click rate, most of it on mobile.

How to match recommendations to intent:

• Show best-sellers to first-time visitors, since you have no history to personalize on yet.

• Trigger frequently-bought-together the moment there is an item in the cart.

• Bring recently-viewed products back for returning visitors to pick up where they left off.

Emma Sleep: a quiz that raised email signups by 50%

Emma Sleep captures email and SMS in more than 20 markets, so a small change to the signup form is multiplied across a lot of traffic. The weakness of a standard form is that it asks a stranger for an email before giving them any reason to hand it over.

So the first screen asks a question instead: what are you shopping for, mattresses, pillows, or beds? It costs the visitor nothing to answer, and once they have made that first small choice, a micro-commitment, they are far likelier to finish. Of the people who pick an option, 68.6% go on to give an email.

That one change raised email signups by 50% against the old single-step form. At the cart the team runs the reverse move, a reminder with no form at all that leads with the 200-night trial and free returns, which cut abandonment 5.54% against a control group and raised average order value 5.5%.

How to build a one-question signup:

• Open with a one-tap question the visitor can answer, then ask for the email next.

• Use the answer to segment the contact and tailor the first message they get.

• At the cart, swap the form for a single reminder of your strongest reason to trust you.

Nutrimuscle: educational tools that drove €3M+ in revenue

Nutrimuscle sells sports nutrition, where first-time buyers stall because they cannot tell which protein or dose is right for them. A discount does not answer that question, so it rarely unblocks the sale.

Instead of a coupon, the brand built a comparison tool that walks a shopper through the differences between its proteins and lands on the one that fits. It removes the exact doubt that was keeping someone from checking out, and it handles email capture as part of giving the answer.

That single tool ran a 7.1% click rate and brought in €278K. With popups and a feed handling everyday capture alongside it, onsite campaigns attributed more than €3M across the year.

How to turn a buying question into a tool:

• Find the one question that stops first-time buyers from checking out.

• Build a short tool or guide that answers it, on the pages where it comes up.

• Ask for the email as part of the result, not as a gate in front of it.

Ziggy Family: lead magnets that captured 9,300 subscribers

Ziggy Family sells pet food on subscription, and its buyers show up with questions: how much to feed, when to change formula, how to manage a plan. Those questions are normally a support cost. Ziggy turned them into its main source of new subscribers.

On each article answering one of those questions, it offers a matching download, a feeding guide on a feeding article, in exchange for an email. Because the offer is the thing the reader already wants, the strongest of these reached a 74.9% click rate.

The approach added 9,300 subscribers in five months at a 5.7% signup rate, close to double the usual benchmark. The same click-to-open format on account pages lets customers pause or reschedule a delivery on their own, before they open a ticket.

How to turn questions into lead magnets:

• List the questions buyers ask most before and after they subscribe.

• Answer each with a matching download on the page where it is asked.

• Reuse the format on account pages so it deflects support tickets too.

Ideal of Sweden: mobile-first capture that collected 698K emails

About 90% of Ideal of Sweden's visitors are on a phone. A campaign that looks fine in a desktop mockup but sits cramped on mobile is quietly losing most of the audience, so the team designs for the phone first and treats desktop as the edge case.

Then they tested the offer. A flat 15% off beat a gift-card prize draw by three to one on signups, because a shopper on a phone will trade an email for a certain reward far sooner than for a chance at a bigger one.

Across 12 storefronts the program collected 698,000 emails at an 18.8% click rate. A one-tap question after checkout, asking how the buyer first heard of them, fills in attribution the analytics miss, with response rates as high as 80% in some markets.

How to run a mobile-first test:

• Build and check every campaign on a phone before you look at desktop.

• Test a certain reward against a chance-based one and keep the winner.

• Add a post-purchase source question to see where buyers actually came from.

Maison Lejaby: non-intrusive product discovery at a 15% click rate

Maison Lejaby sells premium lingerie, where a popup pushing a discount into a shopper's face cuts against the brand. The team still needed a way to help people find matching pieces and best-sellers, without the interruption.

Their answer was an onsite feed the shopper opens by choice, a small marker they tap when they want it, rather than something that blocks the page. Inside, new visitors see best-sellers, and a returning visitor sees the piece that completes what they viewed, a bra matched to its set.

Because it never interrupts, it suits the brand and still performs: a 15% click rate over four months, higher on mobile than desktop. It has run since 2021, which is the clearest sign it is not wearing customers down.

How to recommend without interrupting:

• Put recommendations in a feed the visitor opens, not a popup that blocks them.

• Show best-sellers to new visitors and a matching piece to returning ones.

• Keep the wording about discovery, not discounts.

Foncia: a routing quiz for a services business

Foncia manages property across 600 agencies, so the people landing on its site want very different things: some are renters, some are landlords, some want to buy or sell. One homepage cannot lead with all four at once, and a generic message fits none of them.

So the first thing a visitor meets is a quiz: which of these are you here to do? Each answer sends them to the matching page and team, so nobody has to dig through a menu built for a different visitor.

On top of that, campaigns written for a visitor's own city raised main call-to-action clicks by 6.5%. It is the same principle twice: show people the version of the site that already fits them.

How to route visitors by intent:

• Ask visitors what they came to do, then route each answer to its own page.

• Target campaigns by city when location changes what you offer.

• Version campaigns for mobile and desktop, since the context is different.

What the best ecommerce case studies have in common

These seven brands sell different products, but the same few principles drive their results, and those hold up across the 1.8 million A/B-tested sessions we measured.

Five stand out:

• Ask before you sell. Emma's one-question start raised signups 50%, and the pattern holds at scale: multi-step and micro-commitment formats drove up to 8 to 9 times more interactions and doubled signups, with bounce unchanged.

• Trigger on intent, not on arrival. Foncia, Nutrimuscle, and Emma all wait for a signal before they ask. In the aggregate tests, a 20 to 50 second delay cut bounce by up to 45% and raised email capture 20 to 43%, while campaigns shown in under five seconds pushed bounce up to five times higher.

• Lead with concrete value, not emotion. The plain, numeric offers behind Emma's cart recovery and Ideal of Sweden's 15% test match the data: quantified value copy beat emotional messaging in 68% of tests and grew revenue 8 to 15%, even when click rate stayed flat.

• Reach the visitors a one-time campaign cannot. Maison Lejaby and Ziggy both added an opt-in feed for returning shoppers, the audience a single popup can never show again.

• Judge on revenue, not clicks. Value-led variants often won on revenue while their click rate held flat, which is why every result here is measured against a control group.