The pricing pages that we find on the Internet can be very diverse and original, but they share a common objective: to sell products and services in the most effective way possible.
Today we will talk about the pricing pages, and their pricing tables, in particular. These are the blocks that show the different products or services for sale, along with their price and detail. I am sure you have seen them of all kinds: more concise, more detailed, with more or less design. And you probably agree that its importance is vital for online businesses.
As it is such a relevant element, it is really decisive to monitor its degree of effectiveness. We need to understand how well they are working. And of course, constantly testing changes in the way we display our pages is the best we can do to see if we can improve the number of sales we get.
In this article we are going to see in detail the case of our pricing page for Nelio A/B Testing. And specifically, we are going to explore a very easy change to apply in the pricing table that may make us improve (or not) our sales.
Changing The Order of The Products
Usually, a pricing table on a web page shows the products in several columns. In each of these we see the detail of the product along with its most relevant characteristics and, of course, its price.
The pricing tables are arranged so that the cheapest product appears in the first column and the following products are ordered so that their price increases. But this may not always be like that.
In the case that we are going to discuss today our table of products follows this order, but we want to test if by changing the order of the products, those more expensive receive more attention. Below you have a comparison showing the two versions to test:
If you slide the divider above left or right you will see the changes in the order of the products. Originally, as we have already said, these go from the cheapest to the most expensive, but we are going to turn them around to see if the most expensive products receive more clicks.
We could directly measure sales, but we preferred to focus on the micro-conversion of the click on the subscribe button. When your website does not have huge traffic, focus on improving micro-conversions that can give you faster knowledge and a possible improvement that is easier to understand and apply.
Since our website is fully developed in WordPress, what we are going to do is an A/B test using Nelio A/B Testing to analyze the two variants of the order of products in the pricing table.
Our entire website is made using Gutenberg blocks. Most are common blocks that already come by default in the editor, but we also created some additional ones to cover special needs of our website.
However, on the Nelio A/B Testing pricing page the first block with the pricing table is not a block. Being quite complex, and because we do not change it frequently, we have kept it as it was before Gutenberg came into our lives: it is defined in a special page template within our theme. Legacy code, baby!
This template has this first block defined in PHP + HTML + JS + CSS, and then takes the page content (which is defined using Gutenberg blocks) and renders it below.
In this case, to define the variant in the A/B test we must first duplicate the page template in our theme and change the order of the products directly in the PHP file. In other words, we create a new template, a copy of the previous one, and we change the order of the products. It seems somewhat complex, but in the end it is only a simple change in the order of two blocks of PHP/HTML code.
In the screenshot above you can see the definition of the A/B test that we have created. The original template, named “[Compat] Testing Pricing 2020” has the default product order, that is, the cheapest plan first, then the intermediate plan, and finally the most expensive plan.
As a variant we selected the template that we just created. We named this new template “[Compat] Testing Pricing 2020 (Alternative Order)” and it shows the most expensive product first, then the intermediate, and finally the cheapest.
Nelio A/B Testing allows you to do A/B testing of templates. It detects the templates that are defined in your theme and when you create the test you can select the control version and its variants. If a page uses a template that is being tested, Nelio A/B Testing shows visitors that template or the variants from the test to see which one works best according to the defined conversion goals.
In our case, as you can see at the bottom of the previous screenshot, we have 4 defined conversion goals. This will allow us to study the test results from 4 different perspectives:
- What click-through rate the basic package receives (the cheapest one).
- What click-through rate the professional package receives (intermediate price).
- What click-through rate the enterprise package receives (the most expensive one).
- What click-through rate does any subscription button receives.
Our hypothesis to test is that, by changing the order of the products, more expensive products will get more clicks when they appear first (we are used to reading from left to right). If this is so, we may have more sales of those products, and therefore increase our revenue.
Analysis of The Results
We have had the A/B test running for almost two months. In that time, the behavior of more than 1,600 visitors has been analyzed. You can see the results below:
As you can see, for the goal that measured the clicks on any plan in the pricing table the results are not conclusive. Even though the template with the reverse order (most expensive prices shown first) has a somewhat higher conversion rate, the differences are so minimal that it is not possible to obtain results with enough statistical confidence to choose one of the versions as the winner.
Therefore, we are facing a technical draw from a statistical point of view. Neither version manages to establish itself as the one that attracts more clicks.
Let’s now look at the results of clicks for the cheapest product, Nelio A/B Testing’s basic plan:
In the screenshot above you see that in this case, showing the cheapest product first causes more clicks on this product than if we move it to the final column (the one on the right) of the pricing table.
The reverse order variant achieves a click through rate that is more than 26% worse than the control version. However, the statistical confidence, although high (more than 80%), does not allow us to be sure enough that we are in front of a clearly winning option.
Let’s now look at the case from the perspective of the click to the professional plan (the one with the intermediate price):
In this case, the plan does not change its order of appearance in the table since in both variants it remains at the center. What changes is the order of the plans that accompany it. Here we discovered that when the plan that appears first in the table is the most expensive, then the intermediate plan receives more than a 50% improvement in the click-through rate.
It is a clear case of how our brain works. We see the first plan and it seems too expensive, but of course, the cheapest one may not seem as good for us. In this case we tend to prefer the middle plan. Basic human psychology.
However, again the statistical component of the test tells us that confidence is not decisive enough either: just over 80%. We wish we had found a higher value, above 90% at least. But it’s not the case.
Let’s now look at the case from the point of view of the most expensive package, the enterprise plan:
It is clear that the results make sense. If we put it first, the most expensive plan attracts more attention. Specifically, it has a nearly 22% improvement in click-through rate. But in this case, we are facing conversion rates so similar between variants that statistics cannot assure that one version is really better than the other.
It was very interesting to see how a simple change in the order in which we display products and their prices in a table can have an effect on visitor behavior.
However, the test has not been able to give us statistical confidence values high enough to venture to apply the variant in which the products are shown in descending order of price as the definitive option.
For now we are not going to make any changes in the order of the products within the table, but it is very possible that in the future we will repeat this test to validate or refute the hypothesis.
As you see, there are times when the results of an A/B test open up more doubts than those that solve you. But what is important here is the knowledge we obtain and to think about these results. It is the only way to improve and navigate in the right direction.
Seeing this information, would you have made any changes in the pricing table? Comment down below with your opinion about it. I will be happy to read your thoughts on this topic.