The Other Kind of Testing

Martin Greif
by Martin Greif 29 Sep, 2017
Those responsible for website conversion tend to engage in either split or multivariate testing. These two types of tests work well for a particular set of tasks, from improving conversion rates for pages that have a steady stream of traffic (or conversions) to finding the optimal combination of page elements to get people to act.

These two approaches to testing are actually quite limited, however. They are unable, for example, to point out the most crippling usability issues on a site, provide firsthand knowledge about user behaviors or offer insights into why those users behave in a certain way.

This kind of data generally comes from another kind of test, one that not many sites leverage: the usability test. While most sites can afford to do some version of usability testing, and would benefit greatly from it, many never actually get around to doing so.

The good news for those companies that have not done so is that getting regular usability tests running is extremely beneficial and not that much of a hassle.

Defining Usability Testing

Usability testing is essentially watching visitors use an interface while they talk aloud. Enterprises create tasks for them to do, watch them work and get them to talk aloud. Sounds simple enough, so why don't more engage in the practice?

The simple answer is that most digital teams are overworked and budget-constrained. That is not a great environment to suggest "let's do this extra step that takes time and money to make things more usable." Like with many strategies or tools, time and money are the core reasons usability testing never becomes part of the conversion toolbox.

Here is the thing, though - it is not that much time, and it is not that much money. Most organizations simply do not know, however, that usability testing can be both lean and efficient.

The Goals of Usability Testing

To start off the right way, marketers need clear expectations on what goals should be established for the test. Usability testing is not about volume, so teams should not expect to gain statistical significance from the test.

Be clear about this from the start: usability tests can help improve websites with a small volume of users because that is all it takes to spot the worst usability problems on a website.

The goal should be about finding usability issues so marketers can make improvements to specific sections of their website, either those that are already live or those they are about to launch.

Leave statistical significance to traffic, survey and split testing tools.

Participants in Usability Testing Unless a company is in an extremely complicated field or industry with rigorous specifications and processes required to make a decision, it can get away with using general audiences alone to test a site. The truth is that even an average person using a browser can find serious usability issues with a site.

The other thing to think about is the number of participants in the usability test. Enterprises can usually get 12-15 general audience participants and find the most glaring mistakes on their site. That said, marketers will not get all 12-15 to test the same thing:

+ Get 4-5 people to test a section, and they will find over 80 percent of the issues
+ Make the appropriate changes, and then test with an-other group of 4-5
+ Edit the section again, then run it with the third group

Running a test with 4-5 people three times will catch more issues than running it with a large group once. This sample group is usually just enough to find the worst issues, and using this sequence will help companies find more of the worst issues.

Involvement in User Testing

Usability testing tends to end internal debates about whether something works or not. Seeing users fail to use something - whether in person or on a video - is typically a powerful enough motivator to get people to agree on issues.

As a result, the findings (i.e., a video of the user) of a test should be circulated to project managers, designers, analysts and anyone in the marketing team who can spare a few hours a month to see what works and does not work on a site. Marketers will ultimately get more from usability tests, essentially, if they can get more people to watch the tests.

When to Start Testing

Another general rule is do not wait for the perfect moment to run the tests; just run the tests.

If a company's competitor has a feature they are also developing, for example, have a few people test the competitor's site to check what works and does not work on their site. Or, if all a team has is a mock-up, it can run a prototype test. If certain sections are not finished on the prototype, teams can simply apologize to participants when they reach those sections and have them test the sections that do exist.

In other words, digital teams should not let the current state of their site stop them from engaging in usability testing.


Frequency of Usability Testing

There is really no one accepted cadence for running usability tests. The idea, though, is that marketers run them on scheduled intervals. If they have a small team, they can start with once every two months, and adjust from there. Running tests regularly - rather than every time there is a major project - has several benefits. Schedules of major projects, for example, cannot derail when tests are conducted, and once key stakeholders start to associate dates with usability tests, those responsible will get more people to attend.

Reporting Usability Test Findings

This is one of the most important things to remember: a large usability PowerPoint review is not needed to make testing effective.

Include This, Not That
Access a sample usability report at

What is needed is a recording of the video (so it can be referenced later), bullet points about the three most serious usability issues (after a batch of tests has been completed) and a recommendation on the simplest possible fix for the worst of the issues (those that do not need a website redesign or a new content management system, for example).

Marketers should not try and formalize report findings by using phrases like "40 percent of the users could not perform their intended tasks" when a statement like "2 out of 5" will read better.

Using the techniques above will be more effective than a 50-page slide deck that no one will read, and will keep things actionable instead.

Putting it all together

Usability tests are powerful tools that should be utilized by just about every enterprise - large and small. It happens too few times because most organizations think it is a formal, costly, time-consuming activity geared toward finding statistical significance. It does not have to be any of those things. What usability testing needs to be is a regularly used tool for improving the site experience and a company's bottom line.

About the Author
Martin Greif brings 25-plus years of sales and marketing experience to SiteTuners (host of Digital Growth Unleashed) where he is responsible for driving revenue growth, establishing and nurturing partner relationships and creating value for its broad customer base.