A definition of UAT
User Acceptance Testing (UAT) tests whether users will accept a piece of software. It does exactly what it says on the tin.
UAT tests two things: does the software actually enable users to do their jobs? And is the software designed well enough that users can do their jobs well?
The first question is mission-critical. If something has changed, which means users can’t do their jobs, the software can’t be launched. The second, though, is still important. For instance, if a button has changed position, size and color in a program, then the user might still be able to do their job – but it takes them longer because they take an age to find the right button. Fixing those sorts of issues (which typically aren’t top of mind for a developer) can protect or even boost productivity and reduce the errors people make when working.
When does UAT happen?
UAT is normally the very last testing phase. This is mainly because it’s one of the most expensive testing phases, involving a number of business users. With that in mind, it makes sense to do as much as possible to minimize the time UAT takes. Leaving it until the end gives an organization the best chance of weeding out as many bugs and issues as possible.
How long does UAT take?
Unfortunately, that’s a bit like asking how long a piece of string is. The length of time UAT will take depends on how many tests need doing. If the software change in question touches multiple systems and processes, more tests will be needed to ensure that everything still works as it should. The quality of your test design also matters, because the better you design your tests, the less time your business users spend testing. Good test design also takes into account regression testing, and how it can be automated.
The availability of business users can also lengthen the process. Getting business users to sit down and do the testing around their other responsibilities can often be a challenge – especially if the testing process is painful.
Additionally, if issues are discovered during testing, they need to be documented, passed to developers, fixed and retested, which is often laced with opportunities for errors. This can drag out the process needlessly. Feedback can be so vague that developers can’t make sense of it, and they have to ask the business user to replicate the issue. Developers may fix an issue, but the retest gets lost in the schedule, dragging out the process. And on it goes.
Possibly some of the most important factors are the ease of taking tests and management. If your testing software is difficult to use, slow or your users haven’t been trained on it, then tests will take longer. Similarly, If the managers in charge of UAT can’t easily work out what needs to happen, then everything will slow down. It’s integral to make things as easy for the user as possible.
Why is UAT important?
UAT is the only phase that focuses specifically on whether the software will actually do what the business needs it to do. It’s a subtle but important shift of focus, as opposed to other software testing phases that are focused on functional testing.
A piece of software might function perfectly well but behave in a way that means work takes twice as long as it used to – or work can’t get done.
The consequences of that scenario? Businesses grind to a halt; customers can’t buy from you (or access their money, if you’re a bank), your share price plummets, and customers might even launch a lawsuit against you. It’s real doom-and-gloom stuff.
If you want to avoid that, you do UAT. And, if you really want to avoid that, you do UAT well.
Why is UAT difficult?
The two main factors that complicate UAT are the updates and the unequipped processes. Updates are coming thicker and faster than before; partially due to the rise in cloud applications and partly due to the increasing levels of integration between applications. Therefore, the testing burden on IT teams has never been higher.
The processes that most organizations use to run UAT are not equipped to deal with this increased volume. Test managers are making do with laborious and manual processes; as a result, cycles are taking longer and the risk of errors getting through the process grows.
Those challenges lead to one of two outcomes. Either the organization delays releasing software updates to users, which eventually will start to hold the business back as users don’t have access to the latest features and functionality to do their jobs. Or, the software is released, but the risk of it containing bugs that cause serious harm to the business grows.
Don’t worry; it all gets more cheerful from here.
How do I do UAT well?
Fundamentally, the answer to this question is that you use software to manage your software testing.
Rather than managing the process through emails and spreadsheets, gathering feedback through screenshots, you use an application that shows you all the tests you need to do, monitors their progress, enables business users to take tests quickly and easily, makes it easy to create and update automated test scripts and collates feedback from tests and makes it available to test managers for triage.
Professional test management keeps UAT on track
Some UAT solutions show you all the tests you need to do, gives you information on where they’re up to and shows who is assigned to each test in case you need to go chasing people down.
You can even drill down into each test to see feedback, including screenshots, giving you complete confidence that everything is where you need it to be. They can also interface with popular dev tools like Jira so issues can be passed to your developers quickly and effortlessly.
Testing tools your users will love
Keeping your business users on side during test cycles is vital. So a good UAT solution includes testing tools that are at once light-touch and extremely powerful. While a user is taking a test, the software sits in the background watching everything. If something goes wrong, its tools step forward to help the user give feedback by annotating screenshots. All that information is available to test managers and developers, eliminating those tedious requests from developers to replicate the problem.
Powerful test automation
Extra elements in your UAT that can also be automated, with some solutions letting you create and manage automated tests with ease.
Running UAT this way is more scalable than manual processes – because everything is central, it doesn’t really matter how many tests you have on because it’s quicker to find out where each test is and take whatever steps are needed to keep things moving. It ensures that no tests or feedback are forgotten. Any issues are flagged, triaged, and actioned – not lurking in your inbox somewhere. While the results of UAT can optimize how your business runs, don’t forget to optimize UAT itself to save on time and costly mistakes. When everything works like a well-oiled machine, it hopefully keeps everyone happier while they’re doing UAT as a result.
At the end of the day, that translates to better software for your company, delivered faster, with less risk – and, actually, far cheaper.
