There's No Such Thing as the Perfect Design
Your client may not know it, but you certainly do: your designs could be better.
Maybe you're creating a new homepage, or redesigning a checkout flow, or creating a new flash game. Maybe you're even helping your customer rework their business model.
And, hopefully, you've done your research. You've synthesized your client's goals and their customers' goals into a coherent business strategy. Hopefully, you've conducted a competitive review. Maybe you've even done some observational customer research.
But that's all preparation for design. And no matter how much preparation you've done, at some point, somebody's got to sit down and actually draw out what the new experience should look like. Call it a design, a wireframe, a mockup. Whatever you call it, somebody's got to do it.
Here's what goes through my head as I work on a new design:
"Wow, I love this new design. I think it will make such a large improvement in our client's business. This new strategy rocks. We should have charged more."
But I'm also thinking:
"Hmm… Is this really the best design I can come up with? There's definitely more than one way to design this while adhering to the new business strategy. I wonder which one is better?"
So I start designing multiple versions of the same experience:
• Two checkout flows
• Three homepages
• Four banner advertisements
• Three subscription pricing models
Since there's more than one way to translate a business strategy into something your customers will understand, how do you chose the best version?
In the old days, we would work with our client to make the best decision we could, based on the business's goals, and what we knew about the customer.
Those days are over.
So, how do you choose which version of the design to go with? You don't.
Let your customers decide.
Thanks to the technologies* of A/B and Multivariate Optimization, you can enter different customers into different online experiences. For example, you can show some customers one version of a homepage, and some customers the other version. Then, tally up your metrics to see which homepage performs better in terms of click-through, conversion, or even RPV.
Why aren't you testing multiple versions of your designs on real customers?
*Note: Technology is no substitute for marketing chops. Someone still needs to design multiple versions of the online experience, and they should all be good designs. The technology will allow you to figure out which one is the best of the group, and why.
Learn more about A/B and Multivariate Optimization here:
- Scientific Web Site Optimization using A/B Split Testing, Multivariable Testing, and The Taguchi Method
- Five Tips For Online Testing
- Running Multiple A/B Tests on the Same Landing Page
Posted by Michael Krypel
