Conversion Rate Optimization in eCommerce

If you ever sit at a meeting of e-tailers, you will hear the words “conversion rate” sooner than later, followed by percentages, all invariably with two decimal positions. E-tailers brag about their conversion rates like fishermen about their catches: “My conversion rate is 2.89%” or “My category [sic] of pet food converts at a 4.23%” followed by appreciative nods.

eCommerce is transparent in real time

Imagine that you are in charge of the e-commerce division of a retailer with thousands of euros of total revenue. A new version of your website is ready to launch in a few hours after many months of work.

Admittedly the current online turnover is nothing to write home about, but boy, the monthly growth of the online business is by far the best of all the areas of the retailer.

The investment of the new version is significant: “It is our future“, your CEO reminded you earlier today. One of the beauties of e-commerce is that the data of sales are available in real time, so you are a few days away from knowing whether all this endeavour will pay off or otherwise.

Nobody can blame you for feeling a bit twitchy just before pressing the button “Enter” that releases the new version.

The curse of the new versions

When McLaren joined forces with Honda and Fernando Alonso in 20176, many expected that the best builder of cars, the -potentially- best engine manufacturer and -arguably- the best F1 pilot would create an instant winning combination. The resulting epic fail is history… and the halo of the picture of the 2017 prototype added insult to injury.

Have you resisted to try the new version of your online bank or, say, the more recent Google Calendar? And when you have used the new one, have you come back to the previous version if the website allowed you to? I have, and it is only natural. Changes of a familiar interface increase the cognitive load.

Many online services are cautious about introducing new versions of their interfaces. In this example Typeform makes it easy for users to try the new version and how and when to revert to the old one.

New versions of an online shop that significantly change the front or back end of the website change, more often than not, the conversion rate. The reasons for this change are a mix of:

  • User Experience.
  • Search-engine optimization.
  • Technical performance.

It is easy and cheap to get hard, actionable data about SEO and performance metrics like page speed. Fresh data and insights lead to adjustments and improvements of the new version with measurable impact.

User experience, however, is a matter of perceptions that are harder to measure meaningfully. Behavioural science is a fascinating topic that, for an e-tailer, is summarised into a single metric, again, the conversion rate.

Web analytics

The first step in your CRO process consists of asking your data analysts as many questions as you can imagine: how are visitors behaving before and after the new version? Which are the flow of navigation that presents the most remarkable changes in bounce rate, depth of the visit and conversion rates? Are the conversion of branded searches a good control group before and after the launch?

A rule of thumb with CRO is to invest 3 times as much in research than in coming up with ideas to try to improve the conversion rate. Know yourself before you decide to change.

Conversion Rate Optimization CRO

How can we drive the user experience to improve our conversion rate?

The expertise and sensibility of UX experts are invaluable. UX practitioners are particularly familiar with the mental models that people have about how websites work based on their past experiences visiting other sites, so consulting with them is a second step in the process of optimising your conversion rate.

For instance, the burger nav element of responsive websites is detested by many designers and UX consultants. They claim that many users do not understand the icon and what it is for.

Designers hate these three parallel lines. They are the icon known as side menu, navigation drawer, or a hamburger, that hides the elements of a desktop menu. Such designers are insensitive to the fact that the character on which the nav burger is based, Unicode U+2630 is called the Trigram for Heaven.

A dilemma: product pages vs category pages

Category pages usually capture less organic traffic than product pages… and they tend to convert worse.

The table below contains real figures of an e-commerce website capturing more than one million visits monthly. Product pages capture more than twice the number of visits of organic traffic of category pages. Product pages convert 50% better than category pages. In this example, Product pages are 3 times more valuable than category pages.

Landing pages Capture organic % over total % Conversion Rate orders / visits
Category pages 23.72% 0.63%
Product pages 49.65% 0.93%

Category pages are great to present the products in the ranking that maximises sales. The most common rankers of category pages are price, relevance and the score that the store decides. The latter is usually an algorithm weighting operative margin per reference and factors pertaining the business like users’ preference, providers rank, etc.

We know that some users have a search intent in organic for which category pages fit better than product ones. Examples of this can be “Lego” which is so generic that it probably is about browsing the catalogue of tens of sets of the brand. A search containing the keywords “First Order Heavy Assault Walker” expects a product page.

A user searching “First Order Heavy Assault Walker” expects a result of a product page of the reference, not a category one about other sets of Lego.

Category pages are thus relevant in some instances and they do eventually convert nevertheless. In fact, marketers present recommendations on product pages because the behaviour flow of users includes

The search engine

Search engine technology is usually very important in e-commerce. We found that sessions with search involved at some point were 80% of the total visits to a large e-commerce website. The searches converted seven times better than sessions without them. Logged-in customers did also use search as intensively despite being familiar with the navigation and the inventory.

An e-tailer wants to: improve the performance of the search engine and make the search functionality accessible, clear and easy to use.

Photo of Rubén Martínez

MBA with more than fourteen years of experience in Paris and London, today he is a part of Paradigma in the areas of digital marketing and analytics. Ruben has the good fortune to head the team organising Big Data Spain conference, one of the most important analytics events in Europe, Cloud, and Artificial Intelligence

See all Rubén Martínez activity