If you have an online store, you want people to buy your products. But not every website visitor will make a purchase. So what do you do if you want to sell more? You can change things on your website, and hope for the best. Maybe more visitors will start buying your products. Or you can figure out where most visitors are dropping out, and make improvements on those specific pages. In this blog, I describe the steps you can take to try to sell more using funnels in analytics.

Note: This blog is a summary of my presentation How to leverage Analytics data to find out where you lose potential customers at WordCamp Europe 2024. You can here view the slides.

Your website is a funnel

You can think of your Web site as a funnel. At the top of the funnel, there are a lot of people: on your homepage, your category pages and your product pages. But the further down you go in the funnel, and the closer you get to the actual purchase, the fewer people will be there.

That means that at every funnel step, people drop out. They begin to doubt whether they want to buy something from you. They decide, for example, that the product is not for them after all, or that it is too expensive, or that they prefer another brand. You will never succeed in keeping everyone in the funnel from start to finish. But if you know in which step of the funnel most visitors drop out, then you can make improvements in those places to keep more people on board. But how do you know where people drop out?

Creating funnels in analytics tools

This is the point where Google Analytics and other analytics tools come around the corner. In these tools, you can create funnels. You can decide which steps the funnel consists of. So think carefully about what types of pages your website has, and in what order users navigate between them. For a standard eCommerce website, the funnel will look like this: home > category pages > product pages > shopping cart > checkout > purchase.

Although I am using an eCommerce site as an example in this blog, it is just as useful to create funnels if you have a blog and you want more newsletter signups, or if you are the Product Owner of the help section of a website

Making the funnel easier to understand

After you have defined your funnel steps, let analytics tool see the resulting funnel. In my opinion, the funnel in GA4 is not very clear. The same data is there in different formats (percentages and absolute numbers, and rounding ratios and dropout ratios, which are each other's opposite). That's why I like to use Google Sheets to arrive at a cleaned-up representation of the same funnel data.

Keeping more visitors in the funnel

The funnel above shows how many people are present in each step, and how many of them proceed to the next step. One of the standouts in the funnel is that only 24.5% of visitors from the shopping cart proceed to the checkout. That's not much. Why do visitors drop out? And how can we keep more people on board?

This funnel is a real example from one of our clients. So we started thinking about what we could do with the shopping cart page to make sure more people click through to checkout. We thought visitors might need some extra reassurance about their future purchase. For that reason, we added an element to the cart telling them that they were buying at the lowest possible price, that the order would be delivered quickly, and that many satisfied customers had gone before them. 

Validating the solution

We then conducted an A/B test to validate this possible solution. Because just because we thought this element was a good idea doesn't mean that it actually leads to more visitors proceeding to checkout. For several weeks, we showed 50% of visitors the original shopping cart, and 50% the version with the extra reassurance. 

After statistical analysis, it indeed turned out to be a winning idea. More people clicked through from the shopping cart to checkout in the version with reassuring information. As I showed earlier, in the original version, 24.5% of shopping cart visitors went on to the checkout. When we implemented the change, this percentage increased to 26.4%.

Of course, this is only a 2 percentage point increase, but it ultimately led to more purchases at the end of the funnel. If your site has a lot of visitors, this 2 percentage point can have a lot of impact. For some sites, it can mean an extra million in sales. But even if you have a smaller site, accumulating these kinds of small increases over time eventually leads to big improvements.

Start optimizing!

If you want to start optimizing your website yourself, use funnels to find out where people are dropping out, and experiments to verify your solutions. To do so, follow these steps:

• Create funnels and see the click-through rate between funnel steps.
• Find out where you are losing the largest group of potential customers. 
• Look critically at these pages, and think of possible solutions to the problems your visitors might encounter.
• Experiment with these solutions and implement the things that work well.
• Check the funnels after implementation.
• Keep repeating these steps.

... Just a quick note: we send out a newsletter every three weeks that includes the latest blogs, team updates and, of course, news about the offerings in our academy. Click here to subscribe.

Newsletter sign up

Some final tips

To derive as much information as possible from your funnels, and thus have a good basis for your optimizations, I have some final tips. 

First, create funnels from different starting points, because not everyone will enter your site on the home page. For example, create funnels that start on category pages or product pages if that's where a lot of people land. If multiple paths lead to adding a product to the shopping cart, create funnels for all those paths. 

Second, look at specific segments, such as device types, because you will sometimes see different patterns there. But don't use too many segments, because it's impossible to maintain a Web site that has all kinds of segment-specific parts. 

Third, always use full weeks of dates for your funnels. That way, you'll average out possible weekend and weekday effects. For example, depending on the products you sell, visitors may view your site during the week but not make the actual purchase until the weekend when they have more time. 

And finally, always use multi-week data for your funnels to limit the influence of outliers. You don't want to make decisions based on one very good or very bad week. 

Want help optimizing your website? We can help! Feel free to contact with us or take a look at our services. 

You might also find this interesting: