Why should we test websites?
Imagine buying a car which hadn’t been tested or gone through quality control checks. Not just whether it starts and runs, speeds up when you press the accelerator and slows down when you press the brake pedal.
What about the radio, the windows, the heating, air con, the seat adjustors, lights, speedometer, the remote control locking, and...
Well you expect them all to work. You also expect that when you go to a restaurant, your meal doesn’t tastes how it should taste, without tasting too salty, too sweet, too spicy, etc.
You expect that what you buy has been tested or checked, and periodically rechecked.
As someone who has spent over 27 years building and managing websites, it STILL amazes me how many websites suffer from similar problems which cause those websites to underperform, frustrate customers, and in some cases completely fail.
Often these website are crucial to the success of the business they’re representing, yet the owners are unaware. Or maybe they’re aware that something is wrong, but they don’t understand why and hope that it will just go away.
I don't understand why it suddenly stopped working!
After all, many website problems just seem to happen all by themselves... don't they?
Website problems don't happen all by themselves!
Some of the most common causes of frustration with websites are due to how they have been built, and the way that they're managed. In some cases they're not managed at all, which is often the case with with many web designers who built websites quickly and cheaply to compete on price.
There are, of course, large corporations who have dedicated development and Quality Assurance (QA or testing) teams. Their websites, content management systems and supporting back-end systems change constantly, yet despite a wealth of resources to build, develop, maintain and test their systems, things still go wrong.
So what’s the problem?
I have covered the subject of cheap and problematic web design elsewhere, so I won’t repeat those same points here. But it’s also fair to say that regardless of the platform a website has been built on... whether it’s small website for a local business or a large e-commerce website for a global corporation, most websites have problems.
These problems come in many different flavours, including:
- A link hits the wrong page or doesn’t work at all.
- Website doesn’t work
- Pages don’t display correctly.
- Pages display fine on desktop, but not on mobile.
- Pages some pages don’t display at all.
- People cannot buy, enroll or log in.
- Contact form doesn’t submit.
- Shopping basket calculations don’t add up correctly.
These are all
functional problems which... if they’ve been identified and diagnosed... can be fixed.
Not all website faults are obvious
But there may be other problems which might not be causing errors and are less easy for people to spot.
- The business requirement was that it should do this, but instead it’s doing that.
- A particular feature should only display for under specific conditions, such as when a loyalty customer has logged in, or when a customer has made a specific selection or a product, product variant, service, delivery option, etc.
- A customer might be based in a location or country which needs to have specific content, options or restrictions applied.
There can be so many variations of a customer journey for a single website for one company that only knowledge of those customer journey variants, often based on customer personas, can determine whether or not that website is working as intended.
In such cases failure to work as designed might not result in failure of the website or cause error messages to be displayed, but a different kind of testing (often referred to as
User Acceptance Testing is needed.
Doing the right thing for your customers
Even away from the functionality of a website, your customer types might have requirements or preferences based on age, gender, nationality, etc.
Again, this is another reason to test.
However, this kind of testing might be considered to be subjective. Whilst some testing can be achieved through marking off steps as
No some testing might consideration from the perspective of the target customer. In the case of
Usability Testing we might need to allow a representative portion of our target customer type, and then we listen to their feedback.
What if we didn’t test thoroughly?
This can be answered very easily:
- The website breaks.
- Customers get frustrated and leave.
- The business loses revenue.
- The website needs to be fixed, which also means that other projects get delayed.
- People lose their jobs.
Personal experience with website testing
I spent many years in corporate e-commerce dealing with these kinds of problems.
As a business analyst, capturing and documenting business requirements for a new or amended website, section, feature, service, product or promotion. Communicating those requirements to teams of web designers, software architects, etc., and making sure that they understood what was required.
All variants would be discussed, as well as impact analysis on how the new or updated features may impact other things.
Then during development and at various sign off stages, there needed to be a way to verifying that the updates are working as intended, and other things which may have become impacted due to the new updates continue to work as they had been previously.
In other words
WE TEST IT!
Communication is key
My role within that company was important enough for me to be on daily conference calls with teams and company representatives in UK, US, Europe, Ireland and the far east. My days were therefore long, so I could accommodate multiple time zones. I frequently flew from the UK out to the US... to New Jersey, Florida and Austin... to work with teams on multi-billion dollar projects.
It was successful because I was able to combine an understanding of the business with a technical background, and have an meticulous attention to detail which enables me to find problems and explore how else and where else those and similar problems might exist.
Lack of business understanding, lack of communication, lack of clarity and meticulous attention to detail, and lack of what to test, how to test, and actually testing, reporting bugs and re-testing after bugs have been fixed would spell disaster.
I believed it then and I believe it now, with websites of all shapes and sizes. And that’s why I’m still amazed at how many broken and underperforming websites are out there, and disappointed that they’re allowed to go live in such a poor state.
Have I made my case?
Surely those cases are bad enough to make a good enough case for testing. f you don’t agree now and are prepared to run the risk, when things go wrong, come back and re-read the above to see if you’ve changed your mind!
But joking aside, obviously testing needs to be proportionate, and the costs need to be weighed up against the risks.
You might, of course, expect that the design team or agency checks and thoroughly tests their work before it goes live, and periodically test it to make sure that everything is still good. But you, as a business owner, need to be sure. So ask for their commitment and for evidence of what was tested.
Download a free website checklist
If you would like to download a free website audit checklist, please go ahead. It contains over 130 check points which you can audit your own website against.
On the other hand if you feel you could benefit from a thorough website audit which, depending on the size and nature of your website, could involved over 400 tests and a video walkthrough explaining what was found and in some cases ways to fix or improve what the audit revealed, please get in touch.
Thoughts, suggestions or comments?
Comments or Feedback?
If you have any comments, thoughts or suggestions about this article, please let us know.
Use the social media buttons below to share this article.
Fields marked with * are mandatory
About Daron Harvey: I'm the founder of Targa Web Solutions, specialising in website management, testing, auditing, troubleshooting & consultancy. I'm now in my 27th year of professional website production, testing and eCommerce best practices, including management of large multi-lingual multi $Billion global websites. No AI was needed or used to write the content on this page!