Ecommerce replatforming: don't redesign for the sake of it

Have a strategy for design — or risk throwing the baby out with the bathwater

Pretty much every replatforming brief we get at Onstate includes a requirement to ‘redesign’ or ‘refresh’ the site as part of the process. With the exception of Brown Thomas — who chose to subtly evolve their UI when moving from Venda to Demandware — it’s been part of every replatform project we have undertaken.

It’s part of a recognisable process where design and technology cycles are in lock step. Often, it seems like an afterthought — “while we’re at it, we may as well redesign the site”. The current site is thought to look ‘tired’ or ‘dated’ (often by those who look at it every day).

Usually the design is old and has outgrown itself as new features have been grafted on — it needs some TLC. New platforms often bring additional features that need new UI. But to radically rehash the front end for the sake of it is adding risk to an already complicated process.

Why redesign?

Delivering a solid user experience is essential to maximising ROI on technology projects. You can have the greatest systems in the world, but the battle for the customer’s attention is fought largely in the browser. A misplaced call-to-action here, some essential content that the user screens out there — seemingly trivial factors can have a big impact.

When drawing up the list of replatform requirements, the situation you should aspire to is one where you ask yourself “Why on earth would I want to change the front end?”. It should be sacrosanct: living thing that has evolved to be precisely optimised to the needs of your users.

There are two reasons this doesn’t happen:

1. It hasn’t evolved in this manner: no-one really understands what works and what doesn’t on the site, nor what users really think of it
2. Someone somewhere wants the design to change — often for unspecified and arbitrary reasons

Taking these two points in turn:

1. No-one really knows what works and what doesn’t

When you redesign your site, one of three things is guaranteed to happen:

1. Performance will improve
2. Performance will deteriorate
3. Performance will stay more or less the same (this is the norm)

Wholesale changes to the site design obscure the changes that have worked with those that haven’t. In point 3 above, what if half of the changes drove improvements but were cancelled out by the other half? You may have thrown away something of value, you may have done something amazing. It would be impossible to unpick this.

People hate change

What’s perceived by internal teams and stakeholders as ‘tired’ and ‘dated’ is often valued by loyal customers as ‘familiar’. Messing with this can unsettle your most valuable users — and could be a reason why lots of rebuild/replatform projects see a temporary fall in conversion following launch.

I started my career building websites for football clubs, and I’ve not come across a more vocal and opinionated bunch of users since. Football club website redesigns — even those which improved the experience by every available measure — were met with a cacophony of feedback, most of it negative. The volume decreased as the user base grew accustomed to the new order, but this feedback was no bad thing and often resulted in useful improvements to the site.

Football clubs are a special case — most of the retailers and brands we work with would kill for this level of user engagement. But this experience taught me the importance on involving end users in the design process.

The importance of feedback

For me, this lack of feedback is one of the reasons why UX has played second fiddle to issues of technology. Technical issues have a very clear feedback loop: something stops working, a business process fails as a result so everyone jumps on it and gets it fixed.

In contrast, UX issues can be likened to the tree falling in the forest. There are issues on every online store that adversely affect performance day in, day out. Over time the aggregate impact of these can be every bit as damaging as a failing system, if not far worse. But with no feedback system, they are rarely noticed. Shoppers aren’t like football fans — they’ll switch allegiances at the drop of a hat and you’ll never know.

2. The political question

As well as the lack of certainty over UX performance, there is often a significant human factor at play. Large technology projects come with large budgets. Someone somewhere has to sign this spend off. These stakeholders often:

• Aren’t close to the day-to-day running of the site
• Aren’t from a design/UX background
• Have strong opinions — especially on creative
• Want to see where their money has gone

For these stakeholders a radically-overhauled design is a good sign that their investment has gone somewhere. But their (well-meaning) input can often have damaging effects on the project. Managing their expectations is critical. Quite often these stakeholders are vital to the business — their gut instinct could be the reason you are there in the first place — and their opinion needs to be respected.

Warning symptoms of this include:

• Arbitrary creative requirements: the last word on this is here.
• Spending a disproportionate amount of time on the design of the desktop homepage, and not being that bothered about the mobile version — even though the latter gets 2x or 3x the traffic.

Solution: Developing a Design Strategy

If UI and UX design are to be treated with the respect they deserve then we have to adopt a serious, strategic approach to design. Adopting a structured and rational approach to design can help highlight its importance and overcome subjective challenges. Initiatives you could undertake as part of a Design Strategy include:

• Develop feedback loops: user polls, surveys, session recordings, prototyping, A/B testing and usability testing are all great ways to hear the customer’s voice. Do it as part of a project and keep doing it once live. With a mature retail site you’re unlikely to get many big wins through A/B testing, but you should try to measure the benefit of each change. Have someone from customer services sit on the project board — you’d be amazed at the ideas they bring to the table. Once you have this data, design with it. Remove subjectivity from the equation.
• Adopt design principles: take inspiration from Gov.uk and draw up a shortlist of guiding principles. Refer to these with each new evolution of the design. One principle that we always advocate is simplicity: if you have multiple ways of achieving something, opt for the simplest.
• Develop a design system: as the ecommerce space matures we’re likely to see a lengthening of the re-platform cycle. There’s no reason why the design should keep the same timetable as the platform. Develop a flexible design system that can accommodate change whilst preserving the integrity of the design and keep building value into it. Consider new interfaces such as voice. When a replatform does come around, the presentation layer should be dropped onto the new stack. Review some notable design systems here.
  
  By developing a Design Strategy, you can take control of UI and UX design to the benefit of the overall business.

Considering re-platforming your online store, or wanting to take a more structured approach to ecommerce design? We’d love to talk and help you put the ideas above into practice.