Big results from small efforts: using discount usability for valuable UX insights
This post originally appeared on the FreeAgent blog in October 2017
How do you provide the most value for your buck when researching a project which is tight on time and resource?
As a UX researcher at FreeAgent, a large part of my role is ensuring that FreeAgent customers have an exceptional user experience. As a Design team, we were were aware that our process for new users to set up their accounts needed improving. We wanted to take an informed approach to change this, however due to other commitments and priorities within the business we had a narrow scope on this project.
The challenge was on — we had to ensure that we got the most out of our efforts in the most efficient way possible!
What needed improving in the setup journey?
Within the setup journey we could see higher than optimal drop off rates and support tickets indicated to us that certain areas of the process were definitely problematic. Help and tips are included throughout the process but they seemed to be…well, not really all that helpful, simply because we could see through the analytics that a good part of our users had a tendency to ignore them.
On top of making usability changes, we also wanted to improve the look and feel to provide a great first impression, make it more responsive for mobile and bring a better brand integrity. We knew we would have to use a ‘quick and dirty’ approach, also known as ‘discount usability’. Discount usability aims to simplify data collection by concentrating on one feature or flow at a time and going as light as possible when it comes to prototypes. According to Nielsen Norman Group, the long standing gurus of all things UX:
“Discount usability often gives better results than deluxe usability because its methods drive an emphasis on early and rapid iteration with frequent usability input.”
Paper power
First of all we asked ourselves this question: ‘What is the minimum amount of work we can do to test our assumptions and get meaningful insights?’, which is good practice borrowed from the Lean UX methodology.
In their article on discount usability mentioned above, Nielsen Norman talks about the value of paper prototypes:
‘[One of the 3 main components of discount usability is] Narrowed-down prototypes — usually paper prototypes — that support a single path through the user interface. It’s much faster to design paper prototypes than something that embodies the full user experience. You can thus test very early and iterate through many rounds of design.’
Bearing this in mind, we decided to take the easiest route and printed out screenshots of our current setup process. We also had insights from our support tickets that some of the labels on the sign up fields were misleading, so we prepared a second version where all we did was change some wording. This meant we literally spent 20 minutes on designing and preparing the prototypes — other than stapling some pages together. How’s that for efficient work?
Guerrilla testing
We also wanted to save time on the effort of recruiting testers, so we decided to turn to our friends at Creative Circles who kindly let us do some guerilla research during one of their Edinburgh meetups. The audience, made up of freelancers and small creative businesses, (i.e. typical FreeAgent customers), suited us perfectly.
We simply spoke with eight people (four for each version of the prototype) and asked them to go through the printouts page by page and tell us what they saw, what their impressions were and how they would go about filling in various fields.
Although we didn’t expect to get in-depth results from quick and dirty research, we were pleasantly surprised at the great insights that we got.
Making a difference doesn’t have to cost a fortune
One of the things that came up almost immediately was the realisation that there were quite a few unnecessary distractions in the flow of our sign up process.
Our call to start the setup process was a modal designed as a flight boarding pass. We had intended it to signify the user taking off in their FreeAgent journey and getting a flying start. We thought that it would give a friendly initial impression, but as it turned out it distracted our users from what they were really meant to achieve.
The second key finding from our guerilla testing was the confirmation of some suspected issues in regards to our labelling within the form. Not only were the labels confusing to our users but we also found out that we used language which was mismatched to our audience. For example, originally we asked our users to provide their company name, however quite a few of our research participants who were freelancers remarked that they didn’t see themselves as a company.
Another label that users found problematic in the setup form was ‘FreeAgent start date’ — a date from where the users wanted us to keep their books for them.
Although explanatory text was given below the input field, quite a big percentage of our users were getting it wrong, thinking that they should input the current date. Clearly no one reads the small print! This research gave us a good opportunity to test some alternatives and in our second prototype we swapped the text from ‘FreeAgent Start Date’ to ‘When would you like us to start keeping your books from?’. The result was a 100% pass amongst our testers at Creative Circles. We didn’t even need to include any description below the field.
There is a one simple conclusion from this — if your fields need an explanation, then there’s something wrong with your labels.
Minimum effort, maximum impact
We can’t emphasise enough just how little design time, cost and resource was involved in this research project, though of course we had our UX expertise. With no real work needed in designing and building of the prototypes and virtually no recruitment time for the research participants, we attained pretty outstanding results. Along with analytics, this research helped to inform the redesign work of our setup process (which you can read about here) and helped to bring about two key improvements:
- Firstly we managed to reduce the support tickets sent to us about the ‘FreeAgent start date’ issue. Year on year, the number of tickets fell 46% for our free trialists and 20% for the users who had already signed up. That’s a pretty impressive win!
- We also saw an decrease in the overall drop off rate for the setup process,from 24.5% down to 20.5%. Those changes also affected our three-month rolling conversion rate with an increase of 4% after all the changes were implemented.
Of course other factors might also have affected this, however having a better understanding of where we were missing the mark meant we were able to act from an informed position and make better decisions.
This whole project demonstrated to us that there can be great gains in doing even the most simple research and talking with users.
So as a final word we’d say there really is no excuse for skipping on user testing and research. Grab some paper, get out of the office and just do it!
