“Lazy Registration” and the value of iterative research & design
1 March 2010
1 March 2010
One of the things we concentrate on at Madgex is the use of iterative research and design to improve conversion rates on our platforms. We run a lot of usability interviews, make design changes, and track the conversion rates via analytics.
The lazy registration system we introduced in version 3 of our platform last year is a great example of this. One of the biggest leakage points of any webapp is registration: users are forced to make a clear decision about whether they want to enter a long term relationship with your site. It takes some effort to complete a registration form and choose a password, so it’s no surprise that many users drop-out at this point – which is a big loss for the site owner.
In 2009, UIE (a leading user experience consultancy firm) found that removing compulsory registration from a large US ecommerce store delivered a staggeringly huge uplift of $300,000,000 in annual revenue– this just goes to show what a significant psychological barrier registration really is!
Analytics data from early versions of our job board platform revealed a similar pattern, though at a smaller scale, so we set out to solve it. We started our analysis by deconstructing the concept of registration. Registration is actually a cluster of things: the user typically has to enter their email address as a unique ID, accept your terms and conditions, choose a password, and, in some cases, verify their email address.
Although all these things are normally bundled together within a single registration step, there is no good reason for this. If you have the technical capabilities, it’s possible to tease the components of registration apart and allow users to do each thing separately, at a time that suits them. This approach is known as lazy registration.
Interestingly, when carrying out usability tests on our prototypes, we found that when registration was a forced barrier, many users complained that it was tedious – but when we took it out, most didn’t even notice its absence (though afterwards tended to report higher satisfaction levels). This just goes to show that good design doesn’t have to “wow” users - it simply has to help them achieve their goals quickly and easily.
After an initial iterative prototyping and research phase, we came up with the following model of lazy registration (shown below), which is now implemented in all of our newest Job Boards.
It's interesting to consider that some job board providers actually take the opposite stance on registration, and use it as an opportunity to glean as much personal information as possible from end users. In some cases they require users to fill in their entire CV and work history before allowing them to register. We've steered away from this out of respect for job-seekers, and it's turned out to be great news for site owners too.
For example, when Star Tribune recently migrated their job board to Madgex, they experienced a substantial uplift for a number of their conversion rates, including 56% uplift in email alert creation, and a 67% uplift in job applications. We can’t take credit for all of this, as there are always other factors involved in real world deployments (including a big marketing push from Star Tribune), but nevertheless it provides a clear indicator that lazy registration can bring real benefits when combined with good business practices.