We help clients develop the 'best' solution - not just 'any' solution.
When and Why
When developing a new site or improving an existing one, we usually find clients and their agencies develop and test one design solution. If you test one option and it doesn’t work well you know you’ve got to develop another version. But – if you test one version and it seems OK you don’t know if there is a better approach that would work better. Testing multiple versions shortcuts the process of finding the ‘best’ solution, means there is less chance of implementing a mediocre version, and saves money in development costs.
Most of the research we do on websites is now done during the development stage: thank goodness the days of usability testing just before, or even after, launch is now in the past for the vast bulk of our clients (see User test early – Quicker, better, cheaper outcomes)
However, what is notable, is that most of the time we are still only testing one option. Clients and/or their agencies go through the discovery phase, make sure they understand the business requirements, often get us to do some research to understand user goals and user journeys, and then start developing the wireframes or early prototypes. Very quickly they home in on one version which we then test with users. But there are usually lots of different ways a web page can be configured so, if you only test one version, how do you know what would have happened if you had tested an alternative approach? Or 2 or 3 or 6 alternatives? You don’t.
Web Usability adopts a truly user focused, systematic and evidence-based approach to looking at the alternatives. Because we have a user focused philosophy we do not come at problems with a particular design, marketing, content or organisational perspective.
We systematically research the market, look at all the different ways an issue can be addressed, assess the different options and identify the key solutions. We then test with users and use the user evidence to guide us and the client as to which would be the best solution. All this is done before any expensive coding is carried out. This approach:
- Identifies the best solution
- Gets all stakeholders and agencies aligned
- Saves money
Also, we find when testing multiple options that different bits of each approach seem to work best – often not the bits you expected. This allows us to identify the most effective elements and combine these into the final version.
How We Do It
We adopt a five step approach to identifying, evaluating and testing possible solutions.
Initially we will discuss with the client a high-level plan of the issues they wish to address over the coming months or year: this may be improvements to the existing site or the development of a new website.
Then for each issue we:
- Identify requirements – We understand the business requirements and any technical or political constraints that will affect potential solutions so we can flag any likely UX issues at an early stage (e.g. if the business requirements are likely to be difficult to achieve, say because of the complexity of the offer).
- Review & evaluate options – Before developing the test assets we identify and assess potential approaches. We look at the ways the issue could be addressed (e.g. by looking at how other sites achieve similar requirements). We evaluate and document the options. We then discuss this with the client and agree 2 or 3 options for testing.
- Develop prototypes – We then develop low fidelity wireframes or visuals so we can test the concepts with users
- Usability test – We then test the prototypes with users, typically 6 users. The testing is conducted either remotely, by sharing the tester’s desktop, or in our studios in Chippenham and is facilitated by one of our UX consultants; the testing session is recorded. We summarise the issues that surface from the testing and provide access to the test videos. We then have a telephone conference with the client to discuss the outcomes and agree any changes.
This process allows us to identify the ‘best’ solution to an issue, rather than just ‘a’ solution and can be done quickly and economically.
The outputs are low fidelity prototypes of the optimum solution which can then be taken and implemented by the designers and developers.