![\"\"](\"https:\/\/trilogi.com\/en\/blog\/wp-content\/uploads\/sites\/2\/2021\/02\/Pic1.jpg\")
<\/a><\/p>\nIn our opinion, and after participating, in the development of more than 300 eCommerce projects, we are quite reluctant towards the rationale behind these statements. Our theory is as it follows: \u201cIf a checkout process works correctly, it is usable and has been optimized, it will generate conversions no matter what kind of checkout process is\u201d. It sounds logic, isn\u2019t it?<\/p>\n
Then, does a store with a One Step Checkout sell more or not?<\/strong><\/h2>\nAfter almost 2 years gathering information and processing A\/B Tests data from different online stores belonging to different sectors, we can base our answer not only in the intuition that pure logic may dictate, but with hard data. And the answer is No. Having a checkout process in one or multiple steps does not directly affect an online store conversion ratio.<\/p>\n
Of course, we all have seen specific cases where a change in the checkout process has brought improvements in the sales conversions of an eCommerce, but the question is: Has it been due only to the checkout process or have there been other improvements in the system that may have helped?<\/p>\n
In one of our last experiences, where against the trend, we switched from a one-step checkout to a multi-step checkout, the result was a slight conversion improvement. If we only analyzed the superficial data, we could reach to the wrong conclusion. In this specific case, the conversion increase was due to other improvements involving different price rates and discounts as well as other usability and functional improvements.<\/p>\n
Sometimes, this is where the main problem lies:<\/p>\n
We have a tendency to unfairly compare processes in unequal conditions<\/strong><\/p>\n\n- If we compare two \u201csimilar\u201d online stores or belonging to the same sector, we are doing it in unequal conditions<\/li>\n
- If we compare an online store that used to have a checkout process and has moved to another, we are also doing it in unequal condition<\/li>\n<\/ul>\n
Why?<\/strong><\/p>\nThe answer is because we cannot compare different audiences, different time spans, different campaigns\u2026 If we compare them in unequal conditions, we will always find examples that may tilt the balance towards one side or another and that we may use this inequality to rise ourselves wrongly with the most absolute truth.<\/p>\n
Let\u2019s do the comparison in equal conditions: Numbers don\u2019t lie<\/strong><\/h2>\nLet\u2019s go back to logic and math. It\u2019s where I feel more comfortable as I can lean on data, not only hypotheses. Here is where can see data in equal conditions as numbers don\u2019t lie.<\/p>\n
![\"\"](\"https:\/\/trilogi.com\/en\/blog\/wp-content\/uploads\/sites\/2\/2021\/02\/Pic2.jpg\")
<\/a><\/p>\nNOTE: This data sample has been taken from different online stores belonging to different sectors. This data has been taken from eCommerces with different traffic volumes and turnovers to endow with the most reliability and data breadth. However, in any case it represents an overall online conversion data representing all the online sectors. Even more than that: this is not the aim of this study.<\/p>\n
Let\u2019s do a simple exercise for those having curiosity and time: Make your online store do an A\/B test of your visitors without doing any change for visitors from one group or another. This is to say, simply divide the users in two groups. 50% in one group using a multi-step checkout process, 50% in another with a one-page checkout process. After a few weeks we can access Analytics and take the data from this division. You will be surprised with the result.<\/p>\n
In practice, we have split in two the store\u2019s data. Here you can see the result of the exercise, with real data from an online store:<\/p>\n
![\"\"](\"https:\/\/trilogi.com\/en\/blog\/wp-content\/uploads\/sites\/2\/2021\/02\/analytics-50-50.jpg\")
<\/a><\/p>\nMathematically, this is quite easy to explain: if your target audience has a statistical behavior in its totality, if you only study half of this very same audience, under the same conditions, the statistical result of their behavior must be by logic the same. Thus, it is expected that both groups have very similar results when it comes to session duration, bounce rate, etc\u2026<\/p>\n
The truth behind one page checkout: it does the same conversions<\/strong><\/h2>\nUp to this point, I want to sum up what this study wanted to prove: if we divide the same target audience in two halves in the same time span and we provide them with different checkout processes with the same user form fields and the same payment and shipping options the conversion rates of both groups are going to be very similar. This is to say, we will not be able to stand for a checkout system or another as there is none that leads to more and better conversion rates.<\/p>\n
As said before, the data sample has analyzed data from eCommerces in different sectors from January 2016 to October 2017. All these eCommerces had a responsive web design and split their users in two groups in their checkout processes. The result is the average conversion rate of those eCommerces, divided by their sector.<\/p>\n
Let\u2019s move on to hard data:<\/p>\n
![\"\"](\"https:\/\/trilogi.com\/en\/blog\/wp-content\/uploads\/sites\/2\/2021\/02\/general.png\")
<\/a><\/p>\nAs shown in the table (and putting aside the different conversion rates between eCommerce sectors, which I insist is not the focus of this study), the conversion rates are very similar within the same business sectors. One Step Checkout has a slight advantage in the food industry but it is the other way around in all other sectors. However, this difference is not noteworthy.<\/p>\n
And, what happens if we divide this data by device?<\/p>\n
- \n
- If we compare two \u201csimilar\u201d online stores or belonging to the same sector, we are doing it in unequal conditions<\/li>\n
- If we compare an online store that used to have a checkout process and has moved to another, we are also doing it in unequal condition<\/li>\n<\/ul>\n
Why?<\/strong><\/p>\n
The answer is because we cannot compare different audiences, different time spans, different campaigns\u2026 If we compare them in unequal conditions, we will always find examples that may tilt the balance towards one side or another and that we may use this inequality to rise ourselves wrongly with the most absolute truth.<\/p>\n
Let\u2019s do the comparison in equal conditions: Numbers don\u2019t lie<\/strong><\/h2>\n
Let\u2019s go back to logic and math. It\u2019s where I feel more comfortable as I can lean on data, not only hypotheses. Here is where can see data in equal conditions as numbers don\u2019t lie.<\/p>\n
<\/a><\/p>\nNOTE: This data sample has been taken from different online stores belonging to different sectors. This data has been taken from eCommerces with different traffic volumes and turnovers to endow with the most reliability and data breadth. However, in any case it represents an overall online conversion data representing all the online sectors. Even more than that: this is not the aim of this study.<\/p>\n
Let\u2019s do a simple exercise for those having curiosity and time: Make your online store do an A\/B test of your visitors without doing any change for visitors from one group or another. This is to say, simply divide the users in two groups. 50% in one group using a multi-step checkout process, 50% in another with a one-page checkout process. After a few weeks we can access Analytics and take the data from this division. You will be surprised with the result.<\/p>\n
In practice, we have split in two the store\u2019s data. Here you can see the result of the exercise, with real data from an online store:<\/p>\n
<\/a><\/p>\nMathematically, this is quite easy to explain: if your target audience has a statistical behavior in its totality, if you only study half of this very same audience, under the same conditions, the statistical result of their behavior must be by logic the same. Thus, it is expected that both groups have very similar results when it comes to session duration, bounce rate, etc\u2026<\/p>\n
The truth behind one page checkout: it does the same conversions<\/strong><\/h2>\n
Up to this point, I want to sum up what this study wanted to prove: if we divide the same target audience in two halves in the same time span and we provide them with different checkout processes with the same user form fields and the same payment and shipping options the conversion rates of both groups are going to be very similar. This is to say, we will not be able to stand for a checkout system or another as there is none that leads to more and better conversion rates.<\/p>\n
As said before, the data sample has analyzed data from eCommerces in different sectors from January 2016 to October 2017. All these eCommerces had a responsive web design and split their users in two groups in their checkout processes. The result is the average conversion rate of those eCommerces, divided by their sector.<\/p>\n
Let\u2019s move on to hard data:<\/p>\n
<\/a><\/p>\nAs shown in the table (and putting aside the different conversion rates between eCommerce sectors, which I insist is not the focus of this study), the conversion rates are very similar within the same business sectors. One Step Checkout has a slight advantage in the food industry but it is the other way around in all other sectors. However, this difference is not noteworthy.<\/p>\n
And, what happens if we divide this data by device?<\/p>\n