There. I said it. It’s possibly my biggest pet peeve in marketing. What I see, over and over, is otherwise smart companies getting sucked into massive redesigns of their websites for no better reason than “it looks old.” Or “we haven’t redesigned it in three years.”
It isn’t that I am a complete idiot about the value of good design. For Pete’s sake, I managed the crack IBM web design team not that many years ago. (OK, it was ten years ago. And, no, I don’t even know who Pete is.) I like good design and I do think that it pays. But the way we do it makes it really dangerous.
We like the “all-in,” big bang approach. Tear down the entire site and build a new one where it stands. Completely change the design, the navigation, and as much of the content as we can afford. Spend as much as we can possibly budget. Burn the candle at both ends for the team to make the deadline.
It’s certainly exciting.
But what happens next is way too exciting. The team falls over the finish line and delivers the new site. And then they go to sleep for a month because they are exhausted. The problem is that this is the month that really counts because it is the first time we find out if the site is any good. This is the month where the project actually starts, because now we can get feedback from customers (by their testimony yes, but more by counting the actions of what they do) as to whether the site has driven more sales, more conversions, or more of whatever you are expecting your website to do.
That’s why I favor a much more incremental approach–implemented with agile marketing techniques.
I’d rather see us do thousands of small tests leading to hundreds of permanent (but tiny) changes over a long period of time. Each test tells us whether the change is actually worth making, because we collect the feedback as we go. It doesn’t always have to be conversions that you are testing for each change–if you change the offers being made, then you test to see whether more people clicked the offer–but overall conversions should be going up as you make all of these changes.
Now, I understand that it might feel uncomfortable to slowly change the look and feel (the design) of your site. And if you’ve decided on a whole new color scheme, doing that on only some pages as a test might feel hard. But maybe you could change just one part of the site–a small country, perhaps, or the service area of the site–to see what happens. Even if you change the whole site at once, you might be able to find new ways of testing if you really put your mind to it. Anything that takes some of the risk out is great.
But most of what we do does not require all of these changes to happen at once. It is perfectly possible to change navigation without changing anything else. To change the masthead of the site without a complete redesign. You get the idea. These smaller, incremental changes can more easily be rolled back if they don’t work. They cause less disruption than one huge change that you are committed to, good or bad.
So try not to redesign your website–except in that you are redesigning it in small ways every day.
