One of the common questions Make it Clear gets asked by our clients is, ‘How do we improve the findability of our product or service?’ A common starting point for the conversation is that they know where it should sit on their website, but their customers tell us they can’t find it. Another common scenario is that, since they have made changes to their services or product range, their conversion rate has dropped significantly. Analytics might show plenty of traffic to your site, but few people are reaching the pages that actually convert. That’s a strong indicator of a structural problem. Whether the stats indicate something is wrong or the evidence is anecdotal, it is often very difficult for the organisation to see where the problem lies due to familiarity bias. User insight and data can help overcome this bias, allowing you to design a new structure that can materially improve findability and sitewide conversion by helping users choose the right category and recover from getting lost.
Navigation
Navigation is incredibly important to any sales or conversion metric you are looking to achieve on your site. You might think that a decent structure is only applicable to big sites with hundreds or thousands of pages, or that, because your site has grown haphazardly over the years, you can’t improve the structure you have. While it is true that if you have a larger navigation, the reward can be considerable, it is also true that even small sites can bury the most important content behind confusing menus, unclear categories or a menu that reflects how the business thinks the structure should be, rather than what the customer actually thinks. A site can look slick and modern and still send its users round in circles. Frustration doesn’t just cost you clicks; research shows that 32% of customers would abandon a brand they loved because of one bad experience (PwC – https://www.pwc.com/ng/en/assets/pdf/experience-is-everything.pdf).
If people can’t find their way to what they came for, it doesn’t matter how good your offer is. A better structure, one that reflects how your customers think and search rather than how your organisation is arranged, will directly improve how easily people can act, buy or get in touch.
What is Information Architecture?
Let’s start with the basics. When we refer to your site’s structure, we’re talking about Information Architecture (or IA for short): how content is grouped, connected and signposted. Improving the IA involves understanding your users’ behaviour and their mental models of how your products or services are organised within your site. Our approach to uncovering that insight can vary from organisation to organisation, but we always use a combination of available data and bespoke research, tailored to your business.
As with many of our clients, you might have identified that you have issues with your current structure, or you might be designing your site from scratch; either way, the structure should never be an afterthought. It is the foundation from which to design, or redesign, a great site. Your site structure should be assessed properly against how your customers interact with you, not shoehorned into how the pages happen to be designed.
Good visual design is a huge part of the story, and the importance of how your brand is presented digitally can’t be overstated, but a good structure is what allows users to do the job they came to do.
Forrester research reported that a site designed to remove friction could potentially raise conversion rates by up to 400%. (https://www.forrester.com/report/The-Six-Steps-For-Justifying-Better-UX/RES117708)
How to approach an IA project
Before any design work happens, the needs of the business need to be understood. This part is often overlooked. What does the business want to achieve? What are the goals, and what KPIs/OKRs are you looking to achieve? We want to help your business succeed; therefore, we need to understand everything about that.
The business objectives are important, but without your customers, it’s unlikely that you’ll be able to achieve them. We look at how the current site is performing to understand this in more detail. Analytics tell us how people use your site, which pages get abandoned, where journeys drop off and which routes through the site people are getting lost on. It’s not all negative, though (I promise); we can also see where you are successful. Why do we do this? It means we don’t assume what the problems are. Heuristics and best practice are great, but it’s about seeing where they actually show up in real behaviour.
Speaking of heuristics and best practice, what are they? Heuristics are principles based on expected behaviour. Humans use them subconsciously to aid the decision-making process, based on expectation and experience. NNG have 10 established usability heuristics (https://www.nngroup.com/articles/ten-usability-heuristics/). We use these and audit your site against them. Speaking of expected behaviour, we look at key competitors, or comparators, and see whether there are common patterns that form the basis of user behaviour. You might be the most innovative of organisations, but you probably don’t want to start innovating and trying to change established patterns; that’s why we’re here in the first place, right?
This is all desk research. Alongside this, however, we start working on the fun stuff. We run what’s called a benchmark tree test. This is a navigation path analysis. Sounds a little technical, but the principle is simple. Tree testing is a method used to understand how people naturally navigate information and the mental models behind those choices. It allows you to see how they interact with current information structures and also allows you to validate and refine the information architecture. It allows us to assess whether users can successfully navigate a site structure to find information relevant to their needs.
At the start of the project, we recommend undertaking a benchmark test to establish a baseline against the current structure. Once we have a suggested new IA, we recommend a validation test to measure whether the proposed IA is successful.
Participants are presented with a text-only “tree” representing the proposed navigation. They are asked to complete realistic tasks, and their journey is recorded, showing where they succeed, hesitate or take incorrect paths.
Tree testing will highlight where the IA is working well and where adjustments are needed.
Understand how people actually think
Analytics and tree testing tell you what’s happening. To understand why, we need to hear directly from the people using the site. This usually takes the form of a card sort, paired with a short interview.
For a card sort, we would typically suggest an open card sort; however, if there is particularly strict nomenclature, we might look at a closed card sort. Participants are given a set of cards, each representing a piece of content or a service, and asked to group them in whatever way makes sense to them, then name those groups themselves. In a closed card sort, they have to use a predetermined set of labels. It’s a really valuable exercise. It removes a lot of the influence and reveals how visitors categorise things in their own heads. It’s an eye-opening exercise, as the groupings people generate sometimes look nothing like the existing structure. It’s a great way to provide evidence for such a change to senior management.
We pair the card sort with a short interview to understand context: what people were trying to achieve when they last visited, what tripped them up and what they expected to find but couldn’t.
With the analytics, tree test results, card sort data and competitor review all pulled together, patterns tend to emerge. It becomes apparent which labels or paths consistently confuse people. From here, we design a new information architecture. This takes the form of a sitemap and taxonomy that reflects how users actually group and describe things, rather than how the business happens to be structured internally. The IA doesn’t necessarily stop with the taxonomy of the site, and depending on the scope of the project, structural updates can be translated into wireframes (greyscale ‘sketches’ of key webpages). The design journey is a whole other post for the future, but essentially, your structural updates don’t necessarily stop at the navigation.
Validation
It’s easy to skip this step, but we recommend you don’t. This is the validation tree test we alluded to before. Before any of the updates go near development, we run a second tree test, this time on the proposed structure, using the same style of tasks as the original benchmark.
Comparing the two results directly answers the question we have set out to explore: is the new structure measurably easier to use? If task success rates go up and completion times go down, that is evidence the redesign works. That’s something you can hang your hat on. And if certain areas are still tripping people up, it’s far more efficient to adjust and retest a structure or wireframe than to rebuild a live page.
Why invest in your structure?
While it might seem intuitive to move a few menu items around based on internal opinion or anecdotal evidence, or to A/B test structural changes on the live site, both rely on assumptions. At worst, they risk making an already poor experience worse for some users, and they still cost design, development and QA time.
By giving you evidence to back up your decisions, a detailed IA project replaces guesswork with a structure you can prove works. It is benchmarked before, validated after and measured against the goals that really matter to your business.
Final thoughts
If your analytics show healthy traffic but few people are reaching the pages that convert, or your customers keep telling you they can’t find what they need, your structure is the place to look. Get in touch with us and we’ll give you an honest, no-strings-attached assessment of where the problem lies and how we can work together to fix it.
