What is ecommerce conversion rate optimization?
Based on over 15 years of experience, I will tell you what you should focus on regarding ecommerce conversion rates!
What is ecommerce conversion rate optimization?
My adventure with ecommerce
What is the ecommerce conversion rate?
How does the conversion rate affect revenue?
The conversion rate optimization process
How can changes be checked to see if they increase the conversion rate?
What are A/B tests?
How should the process look?
What’s my recommendation?
Summary
Ecommerce conversion rate optimization is about changing your online store to increase the percentage of users who complete a transaction on your website.
That is clear if we want a simple answer to this business question. However, if you want to delve deeper into the topic, I encourage you to keep reading.
I’ll tell you a very short but fascinating story. In 2007 or 2008, I started an online store. It was a sales channel for my family’s business and did not have any competitive advantage.
We mainly sold consumer electronics, household appliances, and so-called “small goods,” such as watch batteries or headphones. However, the big problem was that this store did not have any particular competitive advantage—neither the assortment was extraordinary, nor the prices we offered were the lowest on the market. Occasionally, we hit a sales hit (like the Zelmer Elf vacuum cleaner), but that wasn’t often.
This was a time when we imagined we were competing with big online stores. “Imagined” because we were a much smaller store and posed no threat to these (at the time) giants.
We decided that we needed to build a competitive advantage. And we did it by developing an ecommerce conversion rate optimization process. At that time, there wasn’t much material on this process in Poland, so we naturally based our work on materials from the USA.
Based on this, we built a conversion rate optimization process, data analysis processes, and web analytics. This gave us an incredible competitive edge, allowing us to develop efficiently over the next few years.
Later on, this online store also led to a B2B ecommerce solution, so although it’s not a store you would probably recognize today, I think it was quite a significant success for those times—especially since I was still a college student then.
Having established the conversion rate optimization process we developed in this online store, we decided to package it into a regular service and build a company that would implement and carry out such a process for other online stores (this company is Conversion).
We implemented and carried out the optimization process in other stores, such as Reserved, Answer, Conrad, or Budapester. These were, and are, some of the largest ecommerce stores in Europe, and perhaps we contributed a small brick to their development and success.
Everything you read here about conversion rate optimization is all the knowledge I’ve gathered over the years in optimizing ecommerce sites, and it’s the state of knowledge today because I also have the humility to keep up with how quickly market trends change, as well as technology solutions and new ideas for processes.
Let’s start with what the ecommerce conversion rate is. Simply put, it’s a ratio that tells us what percentage of users visiting our website make a purchase, i.e., complete a transaction.
In other words, if we say we have an ecommerce conversion rate of 1%, it just means that one in 100 people who visit our website makes a purchase.
Of course, the ecommerce conversion rate can be given by segment, for example, separate for desktop and mobile. However, we can talk about an ecommerce conversion rate for the entire site and all traffic for simplification.
It’s very simple. Here, I will use a simple formula for ecommerce revenue.
Ecommerce revenue can be calculated as average order value x number of users x conversion rate.
A simple formula for ecommerce revenue.
This means that if we, for example, improve the conversion rate by 10%, then revenue also improves by 10%. That’s how math works. Similarly, if we improve the conversion rate by 100%, we double revenues.
And to put my money where my mouth is, here’s a simple example. On one hand, we have 100 people making transactions with an average value of 75 EUR/USD. And we have a conversion rate from one side of 1%. If we multiply 1% by 100 people, we have one transaction with an average value of 75 EUR/USD, so we have a 75 EUR/USD revenue in the end.
On the other hand, we have a conversion rate twice as high, higher by 100%, so we have an ecommerce conversion rate of 2%. Out of 100 people, two make transactions. With an average order value of 75 EUR/USD, we have a 150 EUR/USD revenue. So, on the one hand, we had a conversion rate of 1% and revenue of 75 EUR/USD.
On the other hand, we have an ecommerce conversion rate of 2% and 150 EUR/USD in revenue.
Example of conversion rate.
Simple conclusion. If we increase the conversion rate by 100%, the revenue from our website also increases by 100%. What’s important is that we can improve the site to generate a higher ecommerce conversion rate.
Of course, it’s not just the look of the site that affects the level of the ecommerce conversion rate.
Here’s a simple case. It’s often the case that we launch a campaign, and it seems to us that it will be great and bring us a lot of revenue. It turns out that it brings us a lot of traffic but not so much revenue because we say we brought non-converting traffic (low-calorie traffic).
Converting traffic, for example, comes from price comparison sites, but such traffic from some banner campaign somewhere on large sites doesn’t convert very well. On the one hand, we can improve our campaigns, the users we bring in, how we arrange proportions between different sources of traffic, and how we arrange user paths, which will also affect the conversion rate.
However, we’re not referring to the conversion rate optimization process. By improving a campaign, we are talking about campaign optimization and traffic source optimization. That’s how it’s understood in the market.
So this optimization is one thing; it affects the ecommerce conversion rate; however, when talking about the ecommerce conversion rate optimization process, we usually mean what happens after users enter our website.
That is, we should change the website so that no matter what kind of traffic enters, the site always converts slightly better than before.
In simplified terms, the process of such conversion rate optimization can be reduced to four points:
For example, we noticed on the competitor’s site that it has very interestingly communicated discounts. Namely, a previous price is shown, but it is crossed out; next to it is written “-20%” and a new price.
Our website doesn’t have this “-20%” notice. In our opinion, it would be nice to put such information. And this is precisely the whole analysis of the competition.
The big question is always how to check if the conversion rate has increased. Of course, we can do a simple thing: implement a change and see if the conversion rate increases, decreases, or stays at the same level.
However, the conversion rate is constantly changing. If, for instance, we don’t implement anything in October, the conversion rate will still grow by December. Because the closer to the holidays, the usually higher the conversion rates are. Users have a higher purchase intent, and we bring in traffic that (due to the proximity of the holidays) is determined to buy.
So, how do we assess whether a change has an impact on the conversion rate? Most importantly, what is that impact? There are things called A/B tests.
In a nutshell, an A/B test means that if a user enters our website, the user lands on one of several versions of our website. This version converts users with a specific efficiency. After some time, for example, after a month, it is visible how versions A, B, C, and D (etc.) convert users into buyers.
Given that this is a simple statistical random draw, traffic distribution between these versions should be very, very, very similar. All versions have the same traffic and offer; they operate at the same time. The only difference is what the website looks like.
What does this mean? As a result, we get information at the end about the conversion rate for a given version of the page, considering that all these versions worked under identical conditions. Brilliant. Thanks to this, we can find out that, for example, the information “Product available. Shipping in 24 hours” is worse than “Product available. We will send it today, 9/04/2024, and the product will be with you on Wednesday, 10/04/2024”.
This is a trivial example, but I’m only trying to convey the main idea. Here, I could develop a lot of myths about A/B tests. For example, does a 20% increase in the conversion rate measured in an A/B test mean that, in reality, after implementing this change, we will observe precisely a 20% increase in the conversion rate? I think it’s not worth losing time on that now; we could make a separate post about it.
What’s important is that this is a more accurate measure than simply implementing changes and watching whether the conversion rate changes (we call this observing the trend).
However, it is not always worth conducting A/B tests. The bigger the site, the more revenue it generates, the more returning users it has, and the riskier the changes are, the more important it is to conduct an A/B test to get a clear answer about how these changes translate into later revenue.
However, if we have a developing site with few users—really too few to even conduct an A/B test—we still want to optimize the conversion rate. We can observe the trend by implementing changes and checking how they translate to the conversion rate.
I would have lynched myself for such statements just 5 years ago, but business is not just statistics. The fact that we have too little traffic to conduct A/B tests and cannot apply the ideal methodology does not mean we should not optimize the conversion rate.
A/B tests have two disadvantages: we need a lot of traffic, and they cost a lot. So, if I had an online store selling cycling clothing and generating revenue of 75,000 euros per month, I wouldn’t do A/B tests; I would just implement changes and observe roughly whether the conversion rate is increasing in the trend, and that would be completely sufficient for me.
If I worked as an ecommerce director in a huge online store, I would conduct A/B tests because implementing functionalities without testing them and checking them thoroughly could lead to the service losing millions. In such cases, it often turns out that good changes for one user segment are bad ideas for another segment that we haven’t considered.
An extreme case is the optimization of very large services, such as those that generate several billion euros in revenue per year. We have also dealt with such in our careers, and there it turns out that conducting A/B tests is simply quantum physics. One must accurately measure both the statistical significance level and the test’s power, and the data must be prepared for a long time before any analysis.
Regardless of the size of your service, start a constant process of optimizing the conversion rate. It can work wonders and improve revenues.
Of course, it’s not that you’ll make a minor change, and suddenly you have 60% more revenue. Instead, it’s a struggle for a few percent with each change. One change won’t give you anything, but the next one may. It’s about such ant-like work, which, step by step, will improve the ecommerce conversion rate.
If we’re talking about this having to be a process, then the process must meet several conditions.
Someone in the organization must be responsible for this process. It will become a nobody’s process if everyone is responsible for it.
For example, you should establish some cadence to improve the conversion rate in a monthly cadence. This means that in one month, you do analyses, implement minor changes, and check if this has affected the increase in conversion rate.
A budget must be allocated for this. Both to conduct simple analyses or to outsource these simple analyses to be implemented into the store. Because if you have, for example, Magento or PrestaShop, you will need developers to implement changes.
Another point is to write down your ideas for improving the conversion rate in a regular Excel sheet. It’s enough information about what change you want to implement, what’s the growth potential (how you rate it on a scale of 1 to 5, how much it will translate into an increase in the conversion rate), and whether it’s an expensive change (also on a scale of 1-5). Thanks to this, it will be easy to set priorities for individual recommendations.
Just multiply these two values, and you will receive, for example, that one change is costly (1) but has a high potential for conversion rate growth, so 1×5=5, and another change has two 5s – it’s cheap and has a high chance of growth – so 5×5=25, then you know that you should tackle the second one first.
Another thing I would like to point out is to analyze changes in the conversion rate separately for mobile and desktop because these two resolutions often behave differently compared to the exact change theoretically.
If you’re optimizing the conversion rate at all, you’ll probably find the biggest growth at the beginning in changing and improving the order process, then the product card, and only then other parts of the service. If you think that, for example, changing something on the homepage will dramatically change the entire conversion rate, that’s not true.
Conversion rate and average order value
An important thing is to always check the changes in the average order value in addition to the conversion rate at the end. It is easy to make changes that improve the conversion rate but reduce the average order value. As a result, we implement changes, incur costs for them, and revenue stands still. This is very important, especially in large businesses.
And remember, if someone says there’s no point in working on conversion optimization because you can’t run A/B tests, it’s nonsensical. Lack of budget or sufficient traffic on the website doesn’t mean you shouldn’t optimize conversion.
Adapting the optimization process to your own needs is worthwhile. Don’t give up on optimization, even if you are a small player.
Start a simple process as soon as possible and work on it consistently. Over time, you’ll see a steady increase in the conversion rate. Don’t limit yourself by thinking that conversion optimization is reserved for the big players.
Smaller players can be more agile, react more quickly to changes, and build competitive advantages through a higher conversion rate because the ecommerce conversion rate has an incredible impact on the profitability of the entire ecommerce business.
E-commerce conversion rate optimization is the process of making changes to a website to improve its conversion of visitors into buyers.
The ecommerce conversion rate is the percentage of users who purchase on the website. For example, if the conversion rate is 2%, 2 out of 100 people who visited the online store made a purchase.
Start working on optimizing the conversion rate as soon as possible. Over time, you will see an increase in the conversion rate, which will help you develop your business profitably.
