Justifying ongoing website maintenance

Every organisation requires a programme of ongoing website maintenance, but management often doesn’t understand that. How then do you convince them?

Websites are not printed brochures. You cannot just launch them and walk away. They need to be incrementally improved over time. They need a programme of website maintenance.

You know this, I know this, but often those that we call client or boss, cannot see it. And to be fair, neither should they. Their experience lies in print where once it is sent to the printer there is no opportunity to change.

We can continually tweak and improve our sites, monitoring what works and what does not. This gives us an unprecedented opportunity that is unavailable in any other medium.

However, those we report to are not going to instinctively know this, we need to educate them.

Just asking for an ongoing investment in our websites is not going to work. It is unlikely management will listen to such pleas. From their point of view, moving to an incremental process of refinement just sounds like a cry for more money.

We make matters worse by not communicating the problem in terms they can understand. It sounds like we do not have a grasp of the business challenges an incremental approach would bring.

Continuous improvement is not something most organizations are good at, but it is essential for success on the Web.

Quoting experts that you may have heard of, but your boss almost certainly hasn’t, isn’t going to get you far. It is easy for management to dismiss these experts, because their advice does not apply to the unique circumstances of your business.

To find a way of convincing clients and senior management of the need for ongoing website maintenance, we need to ask ourselves why they are so in love with periodic redesign?

The answer is two fold.

First, when you have a site redesigned you see instant, dramatic change. You can see where the money has gone.

Second, finding large sums of periodic investment is easier than finding money to hire people and invest regularly in ongoing website maintenance.

Therefore, if we wish to have a programme of incremental change we need to:

  • Demonstrate its impact in a clearly visible way.
  • Show that investment will recover its costs.

In other words we need to show and not tell.

The problem with showing the effectiveness of incremental development (rather than just trying to convince management of its worth) is that it takes time.

To build, test and launch an improvement to the site that can be measured is not a two minute job. You could end up in the bizarre position of fighting to get managements approval to run a proof of concept designed to get managements approval!

The only way around this problem is to start small. Your site may be crying out for wide ranging improvements, but by starting small you stand a better chance of proving your point with minimal investment.

The best place to start is with an under-performing call to action. These are good for three reasons:

Once you have identified a call to action that you feel could be improved, the first step is to demonstrate its weakness.

There are some excellent tools available for easily demonstrating problems with your site.

Show management videos of users struggling

The first is a service called usertesting.com. This service essentially offers remote user testing. You define a task (such as asking users to complete your call to action) and usertesting.com arrange for users within the demographic you defined to complete the task.

Usertesting.com offers a quick, affordable way of demonstrating problems with your site to senior management.

You receive back videos of each user trying to complete the task, while describing what they are thinking and the frustrations they encounter. Videos can be edited down into a highlights reel ready for presenting to management. As you can see from the video above, this is a compelling argument for change.

At only $49 per user, this is a cheap way of getting some hard evidence about the weaknesses of your site.

Lets be cynical for a moment. Let’s presume that management are resistant to change. If so, there are two likely criticisms they may level at this approach. First, you did not test enough users and second, they are not truly representative of your actual users.

There is extensive research to say that testing more than 5 users is often a waste of time and that exact demographics are not as important as one may think. However, there is actually a way to test with large numbers of your actual user base.

Backup your findings with larger numbers

Crazy Egg is an excellent analytics tool that can be easily installed on most websites. It provides heat maps and various other forms of visualisation to clearly demonstrate whether users have seen your call to action and gone on to click.

Crazzy Egg heat map of the boagworld.com homepage Crazy Egg can clearly demonstrate if users are actually clicking a call to action.

The combination of heat maps (that show where users click), scroll maps (which show where users look) and videos from usertesting.com, creates a compelling case to demonstrate the shortcomings of a particular call to action.

You maybe tempted at this point to take your findings to management, but I would advise against it. Generally speaking people do not respond well when presented with a problem that they cannot see how to fix. They are more likely to bury their heads in the sand, rather than address it head on.

People do not respond well when presented with a problem that they cannot see how to fix.

If you want to get management on board, you need to demonstrate how small ongoing incremental changes can fix these kinds of problems easily.

You probably already have some good ideas about fixing the problem you have identified. As I said earlier, there are tried and tested approaches for fixing a failing call to action. All you need to do is prove that your approach will work.

The first step in this process is to grab a screen capture of the page that contains your call to action. Edit this screen capture to demonstrate how you would fix the problem. It doesn’t need to be pretty, it just needs to get the point across.

Next you need to prove that this has fixed the problem. For that I would suggest two tools. Verifyapp and Feng Gui.

Verify App allows you to directly compare the new approach with the original by performing two simple tests.

Verify App allows you to quickly compare the improvements you have made to the original design.

The first test is called a memory test. Users are asking to look at a design for 5 seconds before it is removed. They are then asked to recall what they saw. If your changes have worked, there should be an increase in the number of people listing your improved call to action over the original design.

The second test is a click test. In this test users are asked where they would click to complete your call to action. Again, you should see an increase between your original and new approach.

These two tests will provide some solid data to backup the fact that your improvements worked. However, although data is great, seeing is believing.

For a more visual representation of your improvements try out Feng Gui. Feng Gui is a piece of software that claims to replicate a typical eye tracking test. Based on thousands of hours of eye tracking results, it claims to be able to predict where users will look when presented with a particular design.

Fengui gaze plot of boagworld homepageFeng Gui creates great visualisations that help management understand how people look at websites.

To be honest, I am doubtful about its accuracy. However, what I love about this service is that it provides some great visualisations that prove incredibly convincing to clients and management.

Simply upload your before and after screeenshots and it will provide a great heat map showing the improvement.

You should now have everything you need to go to management. You have the data demonstrating the problem and a viable solution.

Remember, all you are attempting to do is get their approval for making this one small change and so hopefully you will have no problems. If you do, suggest that split testing the new approach. By limiting the number of users who see the new version, limits the risk if it proves less effective.

Although Google Analytics allows you to carry out split tests, you might want to try a service like Optimizely. Optimizely makes split testing incredibly easy and requires almost no technical knowledge beyond being able to cut and paste some Javascript. It also provides some great reporting tools to clearly identify whether your test has been a success or not.

Optimizely makes trailing a new approach on your site incredibly easy through simple split testing.

Once you have approval, implement your change. However, make careful note of how much time and money implementing the change has cost.

After the change has been implemented track the improvement and then report back to management the results, including the cost of the change.

Being able to say something along the lines of ‘we improved conversion by 10% for 2 hours work’ is exactly the kind of thing that will get managements attention. If you want to really get them to sit up and take notice, calculate how much that 10% increase in conversion is worth compared to the cost of 2 hours work.

Remember to always measure both the cost of making a change and the return it generates.

When you present this success story, make sure you have some other issues up your sleeve to present at the same time. As before, provide evidence of these problems and suggest some possible solutions. However, this time don’t feel the need to prove as solidly that the solution works. Just move straight to split testing it.

You will find that with a proven track record of success and a clear process for resolving problems, management will be much more willing to allow you time to fix other problems. Repeat this process a few times and they will quickly see the value of investing in incremental change. As long as each success brings tangible benefit to the business, they will happily pay for more staff or whatever else you need. Just remember to always measure both the cost of making a change and the return it generates.

Leave a Reply