I think back to the major site relaunch projects I’ve worked on in the last ten years and I see a pattern.
When I started working on the web back in 2003 the projects I worked on were all second versions of sites. Version one sites were typically basic, hurried and often naive designs that had one objective, get us on the internet.
The move from version one to version two sites was a glory period. A client once said to me “You could hit our website with a stick and it would be better than it is now”.
Success was relatively easy to achieve. Crazy increases in conversion and other metrics; 500% increase in sales, 200% increase in successful applications were not uncommon. Version two sites were good, solid well executed builds. We as an industry had learnt how to do things well.
The majority of large scale projects are version three. The days of huge increases in sales are behind us. In fact I’ve come across projects where the opposite happens. Where a new site performs worse than it’s predecessor.
The common factor? Re-designing and re-platforming at the same time.
Your trusted but long in the tooth version two website has had years of tweaking from user research and MVT / A|B testing. It has evolved to be pretty damn good. There are always incremental improvements that can be made but large scale jumps in performance require a re-design.
Your tech team, similarly, have had years to iron out issues around uptime, performance and integration with your platform.
A re-design will give you a leap in performance but requires a new technological platform to deliver the improvements. Moving from an old CMS or eCommerce platform that is creaking and groaning to a more modern backend will deliver operational and business benefits like better inventory management or quicker site updates. Improvements that, visually at least won’t be apparent to your customers. Introducing a flexible architecture that allows new features to be added as and when the business needs them.
Re-designing and re-platforming are both huge undertakings. Re-designing takes time, effort and craft. Re-platforming requires changes to business processes and infrastructure. Doing both at the same time is a huge undertaking. Doing both doubles the complexity, the time and the budget but mostly it doubles the risk.
So what’s a pragmatic manager to do?
The most successful re-designs and re-platforms happen in two stages that ensure we minimise the risk.
First, we have the lift and shift. We take the existing version of the site and put it on a new platform perhaps fixing some quick UX wins while were at it. We test, we let the platform bed in by sending traffic incrementally from the old site to the new fixes technical issues as we find them. Then we launch, quietly.
A good re-platforming project involves putting a flexible architecture in place below the UI. It needs to be flexible so that an innovative UI can be put on top, rather than locking the site into the old design patterns.
Second, we look at a redesign. We understand what works and what doesn’t with the site on the new platform. We plan the design to make the most of the new platform. We MVT test new ideas because on our new platform that’s much easier. We update the look and feel, again this is easier on our new platform because we have experience using it.
We’ve reduced the risk associated with making so many changes in one one go. It’s about spreading risk: don’t change too many variables at once, don’t overburden the team.
We can then look forward to launching the new design and basking in all the associated buzz it creates. All the time knowing the performance, uptime and infrastructure issues have been addressed.
A relaunch takes time and money. Splitting the approach into manageable phases makes success all the more likely.