In the past 3 years, we have barely been doing any marketing for ourselves. Why?

Because we were busy working behind the scenes for brands like KICKZ to deliver mind-blowing results.

The results that we delivered for KICKZ are just one of many success stories we wrote in the past 3 years…

We started working for KICKZ in 2022. They had a horrible conversion rate (0.59%), and they were struggling to turn a profit. It was a “fix this, or we die” situation.

I remember the first meetings we had with the team from KICKZ. Everything was hectic and chaotic, which is normal if you are in this situation.

However, from the first meetings on, we knew exactly what to do and how to systematically fix their low conversion problem because we have done it hundreds of times. We thrive on challenges.

This was also the exact reason the KICKZ team chose us—they knew we had a structured way of fixing low conversions, and they also knew that we have a track record with a 95% success rate.

Within 3 years of working, we achieved something that most companies dream of…..

• In 2023, we increased their conversion rate to 2.9% (+222%)

• In 2024, we got the conversion rate all the way up to 3.7% (+42.1%)

KICKZ was finally turning a profit again. But not only that—these numbers also impressed investors.

In 2023, KICKZ was acquired by 11 Teamsports, one of Europe’s biggest sports retail conglomerates.

In 2022, the KICKZ conversion rate was 0.59%. One year later, we increased it to 2.9%. Then, in 2024, we increased it from 2.9% to 3.7%.

 

How did we do this if we didn't spend any money on ads?

We simply optimized the online shop and tried to squeeze as much money from the existing user base as possible. Here are some tests that we ran:

These 3 tests alone generated KICKZ €221,000.00 while they were running, which basically covered our entire agency retainer for 2+ years…

In total, we have run 84 tests for KICKZ with a win rate of 30%. This means that every third test had numbers like the ones you see in the screenshots above.

The best part about all of this?

The only necessary cost was our agency fee to develop the tests—which is marginal compared to the ad spend brands like KICKZ have.

And just running the A/B tests paid for this. In fact, just running the tests already had an ROI of 6+, not even counting the compounding effect of having additional revenue coming in every month from the changes we made.

This small investment in CRO helped KICKZ to heavily increase its profit margin and get acquired with a good multiple.

For us, KICKZ is not an outlier; it is the hygiene standard and expectation we set when brands work with us—we have done it countless times.

Context & important market trends

The restructuring and the economy then led to the following things. (And you might have felt or still feel similarly due to these factors):

• Was hitting a revenue plateau.

• Their new customer acquisition rate was decreasing.

• Their conversion rate was at an all-time low.

• Their average order value was stagnant.

• And most painfully—their profit margin was shrinking by the week.

But one thing that KICKZ understood was this. They can’t:

• Change the economy.

• Control consumer behavior on a macro level.

• Bring back the days of cheap money.

They knew that all of this was outside of their control. And I have to give them credit where it is due. They knew they had to do things differently in order to survive & strive again! Now, when profits are at an all-time low.

what do you usually do?

You cut “unnecessary” costs and marketing spend and focus on your most profitable channels. Usually email, WhatsApp, or your own app…

But do you know what the second- and third-order consequences of this are? You probably don’t—you probably do not even know what that is. (Which is ok—most people do not know this)

So let me explain the second- & third-order consequences to you. Your existing customers will only buy so often. Without a steady inflow of new customers, you’ll eventually run out of people to sell to.

Focusing on retention channels during a crisis isn’t solving the problem—it’s just postponing it. The real solution to the problem?

It is not cutting these channels. It simply is figuring out how to become profitable on the first order.

And a third-order consequence if you are slashing your investments in crisis mode? You are falling behind in a big way...

You stop experimenting. You stop understanding what actually improves your economics. And this eventually leads to stagnation - and in heavily saturated markets, to bankruptcies. I’m not making this up - it is exactly what Amazon did - just take Jeff Bezos’s words:

If you do not invest and optimize, your competitor will. And you will lose out to them. Here is what happens if one of your competitors is gaining 1.5% on you every month for 5 years…

And this was what KICKZ realized. They realized that instead of slashing their budgets, they had to invest to change their trajectory. 

In the next pages, I will go into detail about how we helped KICKZ succeed and almost 4X their conversion rate over 3.5 years.

If you are still struggling to understand how big of a difference CRO and A/B testing makes, use some of our free tools and take some time:

• Impact of CRO on Ad Performance

• A/B-Test Profit Calculator

Activity 1 CRO Funnel Alignment to find Leaks

When we joined KICKZ, it felt like being a firefighter. Every day we would get messages like these:

• WE NEED TO FIX OUR CR ASAP!!!

• OUR CART ABANDONMENT RATE INCREASED!!!

• OUR CR IS DOWN 10% COMPARED TO YESTERDAY?! DID YOU CHANGE STH?

It was chaos. We don't need to sugarcoat it. But it is ok. It is about the company's future. But the worst thing you can do when things are bad is to panic.

And that was the first thing we did - we outlined a clear plan and approach and explained it and emphasized it in every single meeting. So, what does this mean? Our plan consisted of three key steps:

1. Get tests up and running as quickly as possible.

2. Focus on testing low-effort changes (to stay within budget).

3. Understand where and why users are dropping off to increase the chances of testing something that will actually move the needle.

So, how did we approach the entire optimization process?

First off, we needed quick wins. It was critical to convince KICKZ to continue investing in testing, and we knew the best way to do that was to deliver results fast.

Here’s what we did first: we dug into our internal database of over 2,000 A/B tests we’ve conducted. We filtered the results by industry (fashion) and by win rate.

This gave us some ideas for the first tests, allowing us to get started as quickly as possible without needing a lot of initial data from KICKZ (while leveraging our own data).

We came up with these initial tests:

• Header bar

• Hot badge

Both were simple to develop and launch. Once we completed these, it was time to get to the root of the problem.

The conversion rate was alarmingly low, and we knew there had to be some significant underlying issues. We dug deep and thought critically about the situation:

We approached it like solving a math or physics problem. I often quote Albert Einstein because I believe he summed it up perfectly when it comes to tackling problems:

“If I had 60 minutes to solve a problem, I’d spend 50 minutes thinking about it and 10 minutes solving it.”

So, how exactly did we uncover what was truly preventing users from buying? First, we had to answer a few critical questions:

• Where are users coming from? (Which channels brought them in?)

• What framing and motivations do they have based on the ads they’ve seen?

• When are they buying the products?

• Why are they buying the products?

• And much more...

These questions might sound simple, but answering them requires extensive digging. Here’s how we figured out the answers.

Step 1: Identify Category Entry Points

You’ve probably never heard of that term before. So, what is a Category Entry Point, and how can it help?

Category Entry Points (CEPs) are the key reasons, situations, or triggers that prompt your customers to think about your category. In simple terms, they answer these six questions:

Category Entry Points are incredibly powerful for understanding why people actually buy your products. And it’s crucial to understand this if you want to figure out what’s stopping them from buying.

It’s like solving the known part of a math equation and then using that information to solve the unknown.

But where do you get the data from? It’s actually quite simple. There are several ways to gather that data:

• Option 1: You just ask people. Not the exact questions, but similar ones that subtly extract the information you need. You can ask them directly on your site using a tool like Hotjar. Or, if you already have a larger email list, you can send out questions to the customers who’ve already bought from you.
• Option 2: Use publicly available data. Review data from your Trustpilot reviews, Amazon reviews, the brands and products you sell, or forums like Reddit.

For KICKZ, we did both. Once we collected all the data, we input it into something that most brands don’t have – our proprietary research tool.

This tool is based on several Large Language Models, and we’ve spent the past five years fine-tuning it to extract the best insights from large qualitative data sets.

You might not have the luxury of having such a tool. However, if you have a lot of raw data, feel free to DM me on LinkedIn, and if I have some spare time, I might be able to run it through our tool for you (especially if I see potential for a future collaboration).

We’re not selling this tool (at least, not for now). It’s exclusively for our customers, giving them an unfair advantage in the market by helping them understand data better than their competitors.

We ran multiple analyses like this to get the best possible understanding of the audience we were dealing with and documented everything. If you want to do the same, feel free to use our internal CEP Template.

Step 2: Understand Customer Journey

Once we understood the audience and knew who we were selling to, the next step was to piece together the entire customer journey.

This required even more digging and research—and it’s one of the most critical steps in the process. It’s also one of the most time-consuming parts of optimizing a website. Most people skip this step because they either don’t have the time, lack the knowledge, or mistakenly believe it’s irrelevant. What am I talking about?

• Super in-depth funnel analyses

• Analyzing every heatmap on every page

• Watching 40+ hours of session recordings

The more data you collect on your audience and their behaviors, the better your insights—and the higher the likelihood you’ll uncover something costing you money that can be optimized.

The more data you have, the higher the Success Rate. Let’s dive into some of the things we analyzed and what we discovered. Spoiler: we analyzed a lot.

The first thing we did for KICKZ was map out the entire funnel visually so we could fully understand the user flow. Visualization is such an overlooked aspect of marketing and CRO.

Then, once we had everything visualized, we started asking questions like:

• Which pages have the highest drop-off rates?

• On the homepage, which slider image generates the most revenue?

• Which page group has the highest conversion rate as a landing page?

• What are the most frequently used filters?

• Which filters have the highest/lowest conversion rates?

• How do users interact with the wishlist, and how well do they convert from it?

• Which payment methods are used most often, and what are their respective conversion rates?

• At the time, KICKZ had a minimum spend requirement to place an order - how many users encountered that error message?

• Which products are frequently bought together?

• Do people who shop by look convert better than those who don’t?

• How many users use coupon codes when placing orders?

These are just a few of the questions we started asking ourselves. Once we had these questions, we began building custom reports to uncover the answers. Let me show you how this translated into findings and test ideas:

These insights were just the tip of the iceberg of what we discovered over the last three years.

Within a few weeks, we had uncovered 100+ insights from the data about how users were interacting with the shop - what made them more likely to convert and what was preventing them from making a purchase.

After we presented our findings to the KICKZ team, they started to relax and even get excited about fixing the issues - of course, by testing everything first. At this point, we had two critical things:

1. A clear understanding of who we were selling to and why they buy from KICKZ.

2. A detailed picture of the actual behavior of users on the store.

With the initial research complete and a ton of insights into how the funnel was performing, we went back to the drawing board to figure out the best strategies to fix the drop-offs we had identified.

Activity 2 - Decrease Time to Purchase to fix leaks

When it comes to fixing revenue leaks, the first step is understanding the behavior that leads to a conversion. Again, it’s all about getting a deeper understanding of the problem and the mechanics behind it. Here’s the thing: purchasing behavior is complex.

• It’s not simple.

• It’s not linear.

The best way to visualize it? With this graphic we call Decision Trees.

Stop thinking in funnels for a second. A user doesn’t buy like that. Let me explain. A user will primarily buy your product for one of two reasons:

1. To solve a problem, or

2. To reach a specific desire.

For a KICKZ customer, the decision to buy could stem from solving a problem, like finding a shoe that fits well. Or it could be about fulfilling a specific desire, such as gaining status in their peer group by purchasing the latest Jordans.

However it’s not a simple yes-or-no decision. The decision to buy is made up of many small decisions. A user needs to complete every single one of these decisions. From start to finish to make a purchase.

But here’s the good news - This gives us tons of opportunities to optimize and get more people to buy.

You need to understand why the user isn’t completing a specific decision. To figure out what’s missing, we use a psychological behavior model that explains how decisions are made. This model is easiest to understand when visualized as a graph:

For a decision to happen three things need to align at the same time:

1. Enough Motivation

2. The Ability to Perform the Action

3. A Trigger

There’s a direct relationship between motivation and ability. If motivation is low but the action is hard to perform, the behavior is unlikely to happen. This leads to abandonment, stopping the customer journey, and ending the session without a conversion.

If you’re struggling to understand this model, take five minutes to look at the graph above. Spend some time thinking and try to understand the connections. I can guarantee you that this is one of the highest-leverage activities you can do.

Now, what exactly are motivation, ability, and trigger?

Here’s a quick but detailed explanation:

So, if we want to fix conversions, we need to focus on improving the outcomes of small decisions like:

• Searching for a product

• Going to a shop the look page

• Using the filter more

• Adding a product to the cart

• Continuing to checkout

To fix abandonments at these decision points, we need to figure out which part of the equation is out of balance:

• Do users lack enough motivation?

• Is it too hard for users to complete the action?

• Is the trigger simply missing?

Now, let’s look at some of the tests we ran and how we used this framework to improve results.

We focused on fixing issues and tapping into opportunities (based on the framework) to drive more sales.

Step 3: Identify Revenue Leaks & Opportunities

When it comes to identifying revenue leaks and opportunities, ideas are rarely the problem. The real challenge is testing the most impactful ones.

So, here’s what we did first:

After completing all the research, we put our ideas into our prioritization engine.

This is essentially a formula that ensures we test the ideas that:

• Have the highest potential revenue uplift,

• Are most likely to succeed, and

• Are the least difficult and costly to build.

This way, the best ideas are always tested first. For KICKZ, we used an extremely sophisticated prioritization engine that factored in everything we needed to move quickly and leverage all the data we had. Here’s what we prioritized:

• Where the test will run (revenue exposure)

• How many people will see the test (scroll depth = revenue exposure)

• What research indicators we have (success rate)

• Ease of implementation (cost)

• Our database of 2.5K experiments (success rate)

For every category, we used several data points. In total, our prioritization score was calculated from approximately 25 different data points to identify the highest-priority ideas - the ones most likely to drive more revenue.

Next, we asked ourselves. How can we use the Fogg Behavior Model to best address the issues or tap into the highest-priority opportunities?

Let’s walk through some of the tests we ran and the thought process behind them. One of the first things we discovered was that new users had a much lower conversion rate compared to returning users.

For us, this was a clear signal that people were comparing prices—especially on Product Listing Pages (PLPs).

• Offer a price match guarantee

• Show stock levels of the products

• Provide incentives like free shipping or discounts on future purchases

We chose an approach that required minimal coding. We simply added a badge to certain products, indicating that they were in demand and scarce.

We called it the “hot badge.” When addressing an issue, you should always frame your idea as a hypothesis. Why? Because you need to test whether it actually works or not.

A hypothesis forces you to think carefully about what you want to test and challenges your assumptions. Here’s the structure we use:

• IF we change XYZ

• THEN we expect metric XYZ to improve

• BECAUSE this change should trigger behavior Z and help more people complete their decision

This was our main hypothesis:

• IF we mark hype or bestseller products on the listing pages with hot badges

• THEN the average revenue per visitor will increase

• BECAUSE users are visually drawn to these products (nudging), and they are perceived as particularly popular (social proof), creating a sense of urgency to buy the product now before it sells out (urgency/scarcity).

See what we did there?

We focused on creating a trigger to tap into motivation. Our assumption was that people weren’t buying from KICKZ right away because a strong trigger was missing.

We accomplished this by leveraging cognitive biases such as Social Proof, Urgency, and Scarcity.

It’s an overview of all cognitive biases (with links to Wikipedia), so you can quickly find which ones are most relevant to your situation.

Once we had the hypothesis and the behavioral design, our design team created this version.

Eventually, he agreed to let us run the test with the design. And I’m glad we pushed this hard because this test ended up becoming one of the best A/B tests we’ve ever run. This single test increased:

• The conversion rate by a staggering 8.00%

• The average order value by 6.57%

It’s not typical for a test to produce such a significant uplift. But this result shows that if you truly understand what drives conversions and identify what’s missing (in this case a trigger) you can achieve incredible results.

This badge became one of the most important tools for KICKZ to shift sales toward products they needed to clear out of their warehouse.

If there were items that needed to go, they’d add the hot badge and quickly sell them off to make room for new stock.

Let’s move on to another test. Remember the Shop the Look statistics?

People who viewed looks converted better and spent more money, but not enough users were seeing these pages. So we asked ourselves: How can we make more people aware of the looks?

At the time, the looks were only placed on the homepage. But most users were focused on collection pages and product pages.

This led to a simple idea: why not show the looks on collection pages while people are browsing? Our idea was to simply insert looks into the collection pages and link to them. What was our hypothesis on a behavioral level?

• IF we offer users the opportunity to shop complete looks on the Product Listing Pages (PLPs)

• THEN the average revenue per visitor will increase

• BECAUSE the visual presentation of a complete look encourages users to buy multiple products at once, rather than just individual items. Additionally, showcasing complete looks encourages users to spend more time browsing and discovering products.

The result?

The results were positive for some collection pages and negative for others. This was great because it gave us valuable insights into which collection pages are visited by buyers who are highly transactional versus those looking for inspiration.

Now, we are working on a follow-up test with a very high chance of driving significantly more revenue.

Now let’s look at one final test we conducted. Once again, this test focused on the new vs. returning users revenue split. This was one of our biggest priorities to address.

Why? Because if we are able to convert more new users into paying customers right away, the conversion rate problem and the profitability issues would slowly start to go away.

How can we get more users to commit to their carts and take one step further? Our simple goal was to move users from the cart page to the checkout page and focus exclusively on optimizing this step. However, the cart page was extremely cluttered: