-
-

A/B Testing Vital to Keep up with Evolving User Behaviour: Michael Parker, Frasers Group

Endless Gain

Online shopping was commonplace even before the pandemic struck. But the pandemic and subsequent lockdowns fast-tracked the shift to online shopping in an unprecedented way. No wonder eCommerce growth is said to have catapulted by 4 to 6 years!

Apart from the pandemic-induced boom, the rise of emerging technologies like AI/ML, rapid digitisation, and fast-changing customer needs have also put eCommerce at the forefront of the retail landscape.

This month, Michael Parker, Ecommerce Trading Manager at Frasers Group, shares his views on the changing face of this sector and his take on the importance of A/B tests and optimisation for eCommerce growth.

MICHAEL PARKER, ECOMMERCE TRADING MANAGER, FRASERS GROUP

Michael has been working in eCommerce and digital marketing since 2009, with companies in a variety of industries and sizes. This has allowed him to watch the eCommerce sector in the UK bloom and flourish to its present state. He joined Frasers Group in March 2021 and now manages eCommerce for EvansCycles.com, TRIUK.com, and Sofa.com.

What trends have you noticed in the past 18 months in terms of user behaviour?

Mobile experience in eCommerce is key now

I find the use of mobile in eCommerce quite interesting. Despite a massive focus on mobile and the growing significance of mobile-first web designs and websites, across the industry, mobile conversion rate is still behind desktop by quite a lot.

Now, though mobile website conversion rate is often poorer than that of desktop, mobile app conversion rate is often higher. And that is largely because of the way that we use our phones.

When users are browsing on mobiles, they keep between 4 and 5, 6 tabs open from different websites and switch tabs to do their research. So, they are often multitasking or multi-screening, and hence are easily distracted and not paying as much attention to product selection and purchase.

On the other hand, when customers are using the app, they are more focused on deciding a purchase. They are not there to window-shop, and therefore aren’t easily distracted. So obviously, conversion rates tend to be much higher on mobile apps.

Personalisation is the way to go for improved conversions

Customers are expecting personalisation a lot more. A simple example of this would be users viewing all the list of recommended products or all the products they have looked at recently.

And for anyone who doesn’t understand why personalisation is so important, I give them the example of Spotify or Netflix, and how their algorithms take personalisation to the next level.

Users will get recommendations based on how they search/view/play content on these platforms. So, if someone else uses my Netflix or Spotify account, that’ll change the content recommendations I get for months to come! It shows how big a deal it is to have offerings tailored to users.

These personalisation elements can be scattered around on websites in different forms—the most common one is “recently viewed”; people simply expect to see it today.

On Evans, we personalise in different ways too—for example, when users search for mountain bikes, we show them a product carousel showcasing mountain bikes on the homepage. If five out of six products they have looked at are mountain bikes, then we know what they want exactly, and we will obviously not show road bikes as recommendations.

If they are returning customers, and we know they’ve already bought a bike, then we will show them accessories related to their search.

How do you currently find and diagnose problems on your website?

One of the methods we use is simple yet highly effective—we do site-walks: just browse through the website every single day and see if we can find any issues or pain points that may cause an issue for customers.

Now, because we use the website every day, we know where to go and how to navigate the website to find a particular product. But that’s not the case with customers—if they don’t find an item straight away, they’re not going to put in extra efforts navigating through PLPs and are going to leave the website.

Pay attention to what people are searching for

One of our important processes is to look at what people are searching for as well as get an indication of what they’re not able to find.

So, we often look at abandoned searches because these will tell us whether the search tool is working or not and if users can find what they are looking for through the navigation menu. If it is indeed difficult for them to find what they’re looking for, then they must search for it, and we need to ensure that the search results match user requirements.

Looking at search behaviour also gives us an idea of trends and what products need high visibility on our website. We may even overhaul our entire trading and merchandising strategy based on what people are searching for.

We have our own branded range of bikes like Fox and Pinnacle, but we also see customers searching for some brands we do not sell. In that case, we show them an equivalent bike from our brand, rather than giving them a 404 page not found error.

If we do not have anything to show for their search results, we should be able to show at least a set of recommended products to the visitors.

What are some of the biggest challenges you think eCommerce brands are facing right now?

Recruitment is the biggest challenge

We have seen how lockdowns during the pandemic pushed small stores and local retailers to go online and survive changes. The crisis also encouraged big brands to go on a recruitment drive for eCommerce experts to sustain their business upsurges during these times.

At present, recruitment is a big challenge in this sector because there are not as many eCommerce people as there are jobs.

It is indeed a unique and much-needed discipline these days, but there is no eCommerce degree for candidates who want to learn about it and make a career in the field. This is why we at Frasers have built an internal training programme for entry-level candidates, who may have a business or marketing-related degree, but are new to eCommerce. The programme gives them an overview of eCommerce and how it is different from other business disciplines.

Overall, recruitment is a real struggle because there just aren’t enough people with the skills to do the job for eCommerce brands. And when companies do find people, the hirings often fail because the process is fast-tracked and lacks important details about training and expectations.

What do you forecast are going to be key opportunities for eCommerce brands in the next 12 months?

Plug-n-play will soon be mainstream

A Headless eCommerce architecture allows brands to customise their online stores by separating the front end from the back end. This means brands do not have to worry about which tool integrates or doesn’t integrate with their existing tech stack.

With the growth of machine learning and artificial intelligence tools, no eCommerce brand would want to stay stuck on a specific platform and miss out on these emerging technologies. From a back-end perspective, embracing headless eCommerce allows developers to create custom solutions and plug them into platforms.

For example, Trustpilot has been the standard for years, but if a brand decides to integrate reviews.co.uk with its platform, they should be able to go ahead with the integration easily. This is where headless commerce comes into the picture and makes integrations like these happen quickly and with minimal effort or expense.

Time for AR and 3D shopping experiences

On the front-end, Headless eCommerce has elevated the visual shopping experience for customers by enabling the presentation of 3D models on websites in the past few years. But one of the things I’ve observed and have been involved in is the growth of Augmented Reality (AR).

While working at a previous company that sold tree tents, I found it hard to demonstrate the scale of what these products are, how big they are, what they look like, etc. There are things you cannot get across in a picture or product description.

But customers must be able to see and imagine the product in their space and be able to feel a bit more ownership over it.

However, there are not as many technologies out there now that facilitate AR on desktop. Augmented reality is a very mobile-focused experience, which can be seen as an indicator that companies will need to completely shift to mobile if they want to adopt this technology.

I’ve worked with a few companies in the past that are developing solutions for plugins that can help integrate AR into their tech stack. Leading the way, Nike have been doing some tests with it over the years by letting people try on the shoes with augmented reality. IKEA have also had the feature on their app for a few years.

Surprisingly, what looked like a gimmicky thing a few years back is becoming integral for retailers whose products require a greater understanding of scale and placement. With transportation and delivery challenges on the rise due to Brexit and other rule changes and technology upgrades, adopting AR in eCommerce can reduce returns by customers and improve their shopping experiences.

What are Frasers doing to stay ahead of the game?

One of the things we are proud of is our A/B testing and CRO campaigns. We are so committed to it that we have at least 10 ongoing tests on a website at any given time. You’ll find many articles online exploring top 10 conversion tests to try or something similar, but you simply can’t apply learnings from one website to another; everything needs to be tested.

While this is obvious to those who have been in the optimisation field for years, people who are just dipping their toes in it need to understand this. For example, some of the tests we ran on the Flannels website brought fantastic results in lifting conversion rate, but when we tried the same on Evans Cycles, they had little impact.

The reason is that Evans Cycles and Flannels offer very different products to a very different customer base, each requiring a different a user experience. So, each brand needs to have a unique approach to conversion rate optimisation.

Another reason we run so many tests is we want to test our assumptions (hypotheses) about customer behaviour on the website. Given our years of experience in this field, we know how people shop and how a website works. But each customer shops differently on each website. I know how to navigate a site or try different ways of looking for a product because I work in eCommerce, but an average customer might just leave if they don’t find a product quickly, assuming that it is not there on the website.

Even best practices in the industry are constantly changing due to continuous technological evolution. Customer behaviours are also changing at a much faster rate, and users want to look at something quickly and make decisions based on that.

As a result, brands have to integrate approaches that don’t alienate any customer and deliver a good user experience for everyone on their websites.

That said, we know some established principles that will always work; for example, people will always look for social proof and trust messaging. As we know, people often make an emotional decision and then use logic to justify the purchase. So, brands need to give them both emotional and logical reasons to purchase—get users excited about the product and give descriptions and messaging containing reasons that they can justify the purchase with.

We can’t simply identify a behaviour or accept a trend and leave it at that. Customers’ needs fluctuate quite rapidly and user behaviour is changing all the time. So, brands have to constantly test everything and assume that user needs will evolve.