Imagine yourself in a conference room with the stakeholders involved in your website redesign. There are representatives from Marketing, Sales, Product, and Engineering. Your designer presents the concepts to the group. Do the majority of people around the table hide their feedback until the HiPPO has given an opinion? If so, your organization has fallen to the HiPPO method of decision-making.
You might be asking yourself, “What is a hippo doing in the boardroom?” No, not the savannah, mud-rolling type of hippo. I’m talking about the Highest Paid Person’s Opinion. Even though the person with the highest salary in the room might have good advice, there’s a good chance that they will lead you down the wrong path if you don’t have the right tools.
So what do we do about the HiPPO in the room if we have a different point of view? It can be tough to suggest an alternative approach. Luckily we can use a proven method to provide real numbers that back our theories. This method of decision-making uses the scientific method of hypothesis and testing to find optimized solutions. This is known as the customer tested method. You formulate a hypothesis about which page layouts will encourage the highest conversion rate, then use controlled A/B split testing to support or reject that hypothesis with statistical confidence. In this case, when presented with design choices, you can say: “These look like interesting design options. We should test that!”
Well then, that’s great and all, but it’s not like we’re using beakers and test tubes to get a reaction. So how do we test a design? Do you drop a page layout into a machine that spits out a paper slip that says: “PASS”? Well, not exactly. What we can do is make variations to the page and see how they affect performance. Performance is measured with conversion rates. In a typical A/B test, there is one change to a specific portion of the page that is tested against the control page. This is done by directing half of the users who come to the website to variation A, and the other half to variation B. The rates of conversion for each are then measured with analytics tools. It’s like putting your designs into the ring and seeing who comes out the winner!
Another testing method that can be used is multivariate testing. It uses the same principle as A/B testing, but there is more than one variation or difference between the two pages. This is useful when you would like to test more than one change at a time. Each test has a degree of statistical confidence represented in a percentage that you can show your HiPPO, or anyone else for that matter. So, test often so you can ensure that your design is optimal for your users. You can test virtually anything, but here are a few examples of things you can test that can make quite a difference:
- Headlines. Language and tone is very important when enticing your users. The wrong language can also be an instant turn-off.
- Imagery. It might might seem like a good idea to put generic smiling faces everywhere on your page, but you might find that your audience will tell you otherwise.
- Call-to-action buttons. Color, style, and language really make a difference.
- Lead Generation forms. There is an art in gathering information. Don’t exhaust your users by making them fill out endless forms. They will give up and go elsewhere. A social login might be an option. 77% of users prefer this approach. It also improves conversion rates by up to 50%.
There are many tools in the market for doing A/B and multivariate testing. Here are a few that are worth a look:
And if you really want to dive in, here are a few more resources:
- You Should Test That! By Chris Goward
- Experiment!: Website conversion rate optimization with A/B and multivariate testing By Colin McFarland