Let me start by summarizing this awesome essay from Andrew Chen. Andrew argues that things like A/B testing the landing pages which produces 10%ish increase in conversion do not really increase the bottom line of the company by 10%. This is because the changes in landing page affect the “low intent” visitors than the “high intent” ones. Most of the high intent users, who are eventually going to pay for your services are not really going to be affected by moving some buttons and form here and there.
This differentiation of visitors into high intent and low intent visitors is very useful. Now here’s the kicker, you can use your landing page to actively filter the low intent users. On www.prowidor.com we do not provide register via Facebook/Google options to the users even though many websites instantly boost registration with it. The logic is simple. ‘if someone is willing to pay 50 Dollars for the service, they are willing to type a username and password’.
Of course, if your UX is so bad that even the high intent users bounce, you need to make some changes. For prowidor.com, the same registration form was actually difficult to navigate for a semi literate blue collar worker. But you know what they find easy to do? Chat with their friends on Facebook. So, we built a facebook chat bot for their registration. In fact the workers communicate with us exclusively via the chatbot.
But spending a lot of effort creating the perfect landing experience does pays, especially in two cases.
1] When there is possibility of unconscious/irrational behavior
When there is a possibility of unconscious/irrational behavior, you don’t really have any low intent visitors. Or a better way to put this is that you can convert the low intent users into high intent ones easily. That’s why you see these design manipulations working wonders for social media apps and apps/website with high novelty factor.
2] When the target customer jump between low and high intent.
There are many cases where the user jumps between low and high intent. Take for example a food delivery website. A low intent user at 8 am in the morning, can become a high intent one in a matter of hours. A low intent user for a job portal can become high intent user in a few months.
It is important that the entire user acquisition experience is designed keeping possible jump from low to high intent in mind. It will not work if a job portal sends you reminder emails every day at lunch, and it will make no sense for a food delivery company to wait 2 weeks before sending you a 30% discount coupon code.
There are other instances of not so good UX making a good business sense. There are the obvious dark patterns used to make it impossibly hard to do things like delete account and stop subscription.
Then there are high learning curve games like DotA and terribly outdated website design like classic reddit.com. These kind difficulties in UX make sure that only the high intent user go through the acquisition process. And once they are through, I guess they have a sense of accomplishment that increases their affinity for the product.
I am definitely not implying that you should make bad design decisions on purpose. Just know that changing positions of registration forms and buttons on your website ten times may not increase your bottom line by 100%
You can find more at www.hatwalne.com.