Website redesigns: Top 5 Tips

We recently partnered with a client whose website had undergone a redesign promising a better-looking layout, and ultimately, higher conversions.

The results did not match expectations.

Flightdeck Consulting, which offers online recruitment and coaching services for pilots and flight attendants, went from a consistently well-performing e-commerce site to one losing a significant proportion of sales from its books and courses section.

homepage

Flightdeck Consulting’s website after recent redesign (homepage)

With reduced sales post-launch, Flightdeck partnered with us to (quickly) identify ways to move conversions rates back to the volume they once knew.

From this project and other similar experiences, here are some key rules of the road we would have given to Flightdeck and any other clients considering a redesign:

1. Get a second opinion

There are different ways to make changes to a site: small changes made over iteration, total overhauls and something in between. Before committing to one (potentially costly) strategy from which you cannot revert back, ask around and see if other experts in your field would recommend a different tack.

Flightdeck engaged an agency who proposed a fairly radical change. With 20/20 hindsight, a third-party opinion may have been worth considering, especially if it were to favour optimisation over total redesign.

2. Don’t throw out what’s working

For Flightdeck, a great thing about their old site was that it welcomed users to identify themselves first and foremost as either flight attendant or pilot. This feature was gone from the new site, and we believe it may have been a key entry point to the sales funnel.

The philosophy is to first recognise and preserve the revenue drivers already in play.

3. Establish a baseline of revenue drivers and detractors

Review your analytics, commission a third-party review and/or conduct a super-lean round of user testing to identify what customers like best and least about using your site. After that, design with a goal of eliminating the customer’s pain points while ensuring your keep the magic formula in tact.

That’s step 1 before even picking up the phone to your design agency or team.

Ensure design changes reflect an alignment between your business objectives and what actual customers want and do on your site.

For example, Flightdeck’s new design buried the “Shop” within a (long) list of top navigation items. A quick look at analytics shows that customers weren’t buying because they were not beelining to the “Shop” section. Also, some of those navigation items simply did not get trafficked by purchasing customers (if at all). With that evidence, we made a recommendation to cut down the number of navigation items to only those critical the customer’s purchase journey.

design

We recommended streamlining the top navigation to fewer options and returning to an informational architecture that helped the customer self-identify before moving on. However, we’d also recommend testing designs with real customers before implementing them.

4. Pretty can hurt (the bottom line)

Redesigns often focus on making things look better. But its #1 goal should be conversion, not aethestics.

Upon first glance, the new Flightdeck looked great. Consistent use of font and colours, contemporary look and feel. In reality, the better-looking, more responsive site produced worse results than its predecessor.

Because the old site was ‘replaced’ without sufficient investigation of other revenue drivers in customer experience, Flightdeck realised a risk in changing too much – or not changing the right things.

5. Test as you go, don’t launch with baited breath

Imagine the opposite scenario to the classic ‘big reveal’ that typically accompanies redesigns.

Suppose your site could change one thing and then test the results—either with A/B testing (running two designs of a page concurrently and comparing results) or by a very lean round of user testing – before changing the next thing… then you’d get even more clarity about what changes to make, and what elements to keep.

That’s not practical for all sites, but many companies are moving more in that direction.

Amazon, for example — who remembers the last overhaul of their website? Likely no one because Amazon make incremental changes and test every day throughout the year. For them, optimisation doesn’t look like a revolution — it’s a series of changes that represent continual improvement.

Iterations of testing followed by design keeps you close to your customer’s needs and helps ensure your investments of time and money are well spent.

With the right tools to first run diagnostics and then test design changes iteratively, you’ll be able to get rid of pain points without harming the main drivers of conversion.

Jolly Zhou

CX Consultant
Objective Experience
linkedin.com/in/jollyzhou

A time and place for each device

Image

We advise our clients to design for context first, not device.

Part of that context is the time of day.

Take a look at the peaks and troughs for each device’s usage throughout the day:

image

What are the implications?

  • A user’s context: Understanding the time of day gives you a sense of potentially where the user is and where their headspace is as well. In simplistic terms, PCs are for work, tablets are for home and mobile is for all the time (don’t tell the boss).
  • Myth busted: This goes counter to the myth that mobile is for people who are “on the go” and are for transactions and way-finding only. The phone is on all the time.
  • Planning content: Whatever content users can see on their PC, they might be expecting it on their mobile or tablet as well – whether they are at work in front of their computer, on the bus back home or in bed right before they turn out the light.
  • Another little insight: see that little peak in PC usage right after midnight? We’re not surprised. We’ve seen this behaviour when reviewing analytics of e-commerce sites: a lot of users do their online shopping late at night.

What this data doesn’t show is how relevant it is for your own site or app. You cannot rely on assumptions or aggregate data to optimise for contexts.

You must do the work – reviewing analytics, talking to users, observing their (unreported) behaviour and see what is really going on your website or app – by persona/user, by context, and indeed even by time of day.

Of course, we’re here to help you do the work.