While you could just finish this series with going live with a website, in this final installment I’ll go into the mindset of iterative design and the art of never considering a website as complete. Part of the beauty of building, maintaining and commissioning websites is that we can always tinker, experiment and improve, and ultimately we should embrace this in order to realise the website’s potential. Think of a suit tailor creating a new bespoke suit, you might on first thought consider the go live as the customer walking out of the shop with his finished clothes, but actually the go live is more accurately the first fitting where the customer tries it on and we get the classic scene of the tailor sizing it up, marking in chalk for improvements and overall assessing his/her work.
So how do we approach this and where do we start? We don’t want to start ripping all of our hard work to shreds for sure, and really we just want to take baby steps to start this process for improvement. Before we get into this it’s worth noting that while I touched upon this briefly in the previous article, we’ll now be going into the practical detail of steps you can take in the weeks, months and years after website launch to maintain and improve the health of your website.
In the weeks that follow…
After the initial launch you’ll obviously be promoting links and pages, and returning customers or visitors will be getting their bearings a bit, so analysis of this data may be a bit flawed, but after a couple of weeks, you should be able to go in to the analytics of the website to start to review how things are working and coming along. Really in this period of time you’re looking for quick wins and obvious improvements rather than more sweeping changes. Look at things like page load time and popular search keywords so you can better cater the experience, and try and resist any knee jerk reactions such as deleting a page/product if it doesn’t get any views.
Try and set aside some time every couple of weeks to repeat this task and see how the improvements you made are working, plus identify new or knock on areas of work. As with most other types of data analysis, make sure you’re using logical thinking to make decisions and assess the previous weeks, taking into account things like Christmas, heatwaves, season sale periods and if your content has been featured anywhere else, as this will skew results.
Overall treat this process scientifically and you won’t go wrong. Assess a statistic, make an improvement, assess the improvement taking external factors into account, and if it worked employ the rule elsewhere or in another way. An example of this could be moving the add to basket button above the description on a product page so it’s more visible. If it makes more people add to basket (and hopefully more sales) you could try putting the button in a standout colour to see if that improves it further.
Part of this work, such as the example above, may very well need code and design changes, so you may need to refer back to your previous web designers (or maybe even find new ones), and try and work out some form of retainer for small improvements like this, otherwise you may well be paying a premium charge for every little change. Utilising web professionals as part of your long term website improvement process will also help them gain a better understanding of user behaviour and your objectives, meaning they can start to quickly and easily identify suggested improvements of their own.
If the above paragraphs make your head a bit fuzzy, why not start your first weekly analysis with asking some of the questions below, either to yourself or your web designers along with if there is anything technical you or they can do to improve this.
Weekly diligence checklist:
- What’s the average load time of pages?
- How well does the bounce rate compare to averages for that type of website?
- Are there any popular keywords or phrases new visitors are finding you by? Is it expected? Are you serving them the content they are looking for?
- How many pages per visit are there? Is that enough to complete a transaction? Are people getting lost?
- What’s the average time spent on pages? Is content too long? Are the important points near the top?
In the months that follow…
After a couple of months, you’ll be surprised how natural this process starts to feel, and data analysis and iteration should start showing some significant improvements to your website in theory. By this point we should now open this out slightly and start to consider ‘users’ as people.
While statistical analysis is great and can teach us a lot, users are people, and our customers are people, so really we want to start getting more qualitative data from them to see the website and our services through the eyes of our customers.
Again, with the internet being wonderful, it allows us to now reach out to our audience quickly and easily. If you’ve got a social media following, try sending out a user survey. This could be as simple as asking a couple of questions in a post, to creating an online poll (lots of platforms now have this built in) or using something more comprehensive like Survey Monkey.
You might also have a mailing list of customers and people who have acknowledged an interest in your business, so don’t be afraid to utilise this list to ask for feedback. This can be a super easy way to gain quick feedback, especially if it’s as simple as clicking a button in an email, and being taken to a survey with a couple of questions. Even easier if there’s a simple star rating within the email they can just click. The main thing to remember about asking for feedback (or sending email marketing in general) is not to over bombard. Nothing can make a person unsubscribe more than a barrage of emails they aren’t interested in.
If you’re not on social media for whatever reason, you could always enlist the help of someone that is (who is also fairly fluent in them) or a social media agency to create some targeted adverts directed to your audience.
When drafting these questions, try and get a mix of yes/no or multiple choice, and open ended ones. Try not to ask loaded questions such as “how much do you love us: a) loads, b) lots, or c) times a million”, but rather things like “which of these do you find we are best at: a) response time, b) accurate product descriptions, or c) packaging quality”. Some other useful questions you could ask could be…
- Where do you think we should look at next for improvement?
- What is it about our company which you think is most important for your needs?
- Do you feel you can find the information you need quickly?
- What would encourage you to spend more time on our website?
- Is there anything you think we do particularly well?
- If you feel you need to speak to us, do you find our contact details accessible?
You could even go as far as hosting focus groups, or if you have somewhere nearby that’s public access you could conduct some ‘hallway tests’ which are simply just stopping and asking people about your website, or your business, or just your industry in general.
To make all of this worth while for potential responders, try offering a discount code to your website as an incentive, or even free cake/beer/pizza for focus group sessions. For the sake of the £50 it will cost you, some of the results and insights will earn you a lot more.
A final point on gathering user information in this way is GDPR and general privacy. You’ve likely come across these terms, and you’ll need to be wary of the data you’re collecting. First, you’ll need to be clear with surveys and focus groups what exactly the data is being used for e.g. the improvement of your website only, and if you’re getting their personal details such as name and email address, that you’re storing that information securely. You must also be sure that you only contact them afterwards with information directly relating to the survey or group they were part of, and not general marketing (unless they explicitly say you can do so) as its against GDPR, and also generally annoying.
To help you along the way here are some other tasks you can do on a monthly or semi-monthly basis…
Monthly diligence checklist:
- Have you touched base with your web designers to see if they’ve had any thoughts or suggested improvements?
- Are all areas of the website still working properly? Have you tested forms recently or any automation? As browsers and operating systems update, things that once worked may suddenly not, so it’s good to test
- In your analytics are there any areas of the site that just haven’t been viewed at all in months? Can you remove them?
- Does your website data show returning visitors as well as new? If so, how are they behaving differently?
- How does your analytics compare with what you studied in the months before? Have things stayed the same or continued to rise?
- Are the goals you set for the website at the start of the design process coming to fruition or at least on the way to being met?
In the years that follow…
In reality, even the most solid of websites won’t last forever, and even with the improvements and tweaking, the time will come to recreate your site. Browsers and technology change and improve, as does our expectation of what a website will do. Think of it like a car. We recently scrapped our trusty Ford Cmax which was around 15 years old. When that came off the production line it was the latest, cleanest and smoothest to drive car, but as time went on, it simply degraded, and our expectation of what a car should have such as heated seats and screens, changed so it didn’t seem as good as it once did. It’s sad, but we learned what we liked about it (large boot, separate back seats for the kids) and applied that thinking to buying a Ford Galaxy.
It seems a deflating prospect that all the hard work will at one point be designated for the code scrap heap, but that isn’t to say all of the information with regard to process and your users is lost too. Starting again you should have an eagle eye for a design process, be at the ready with your content, and have a head fully loaded of user insight and behaviour. Just as the go live isn’t the end of the story, neither is this version of your website.
With over a decade of industry experience, Jon Dodd has worked with a range of clients, from government bodies to luxury brands, in the UK, United States and Australia, to create user-centred, accessible and end-to-end web experiences. Formerly Creative Director at the award winning digital agency Un.titled, he is Design Partner at the design studio Alloneword, specialising in creative direction, visual identity and web experience.Read full profile