Skip to main content

Improving navigation on GOV.UK

Posted by: , Posted on: - Categories: Finding things, Product changes

Featured image

One of the biggest frustrations users sometimes have with GOV.UK is the difficulty of finding what they need - and discovering what they don’t yet know they need - through site search and navigation.

We’ve made solving this problem one of our top priorities this year, and have a multidisciplinary team (the biggest single team on GOV.UK) dedicated to this work, appropriately called the Finding Things team.

Our end goal is to make GOV.UK joined up and really easy to understand. We want it to be easy for users to find the content that is most relevant to their needs irrespective of how they entered the site or their current understanding of the subject they are interested in.

This is a huge task that will require lots of departments and organisations across government to work together to implement appropriate content strategies and accurately categorise content. The prize for getting this right is equally huge as it makes good on the promise and potential of the single government domain where users no longer need to understand the structure of government to engage with it.

The team has been working since February to understand the main issues which prevent users finding what they need, and how to address each one. We’ve been blogging as we’ve gone along, see:

Background to the problem

GOV.UK grew and changed rapidly during transition in 2013 and 2014 as the web content of more than 350 government organisations came together on one site. It’s now a large website (180,000 pages, up from 115,000 this time last year).

Now that the whole of government’s content is in one place, we need to consolidate the navigation systems we built and review all the content again to join it up more effectively.

Defining the problem

There are 4 main categories of problem.

  1. Our navigation is hard to understand
    Currently, users have too many navigation choices to make and so they often don’t feel confident that they are in the right place for them. We know from feedback and user research that people are getting stuck, or going round in circles, and some leave thinking GOV.UK doesn’t have the information they need.
  1. Internal search does not work as well as we need it to
    GOV.UK's current search system only works well for the most common needs.  More specialist searches tend to produce less relevant results, mixing content from different areas of the site in an unhelpful way.  Using structure such as subject areas would allow a focus on more relevant results.
  1. Users don’t know where they are in the site
    There are many frequent users of GOV.UK, but the majority will use it very occasionally. They will often come from search engines and enter deep within the site without context - so we need to make it clear at a glance whether they’re in the right place for their needs and if not how they get to where they need to be.
  1. Our tagging systems aren’t well understood by publishers
    This is also a problem for the people who publish content onto GOV.UK. For example, currently a publisher has to intuitively know what a collection is, when to use it and how it differs from a topic. We need to make this easier to understand to ensure consistency in the site’s navigation.

The plan to solve the problem

Put simply, there are three main steps:

  1. Build a topic hierarchy (taxonomy) that covers all content on GOV.UK
  2. Tag all content to this hierarchy
  3. Use this hierarchy to support navigation, orientation and search

The complexity, of course, is in the details so I’ll expand a little on each component part below with more detail to come soon in future blog posts.

Building a topic hierarchy for GOV.UK

We’re planning to have one subject-based taxonomy for all content on GOV.UK - in other words, a single list of topics arranged into a tree-like hierarchical structure. All content, regardless of content type, will be tagged to at least one topic. The topics a piece of content is tagged to indicate what that content is about - and in the hierarchy related topics are grouped together.

We already have taxonomies on GOV.UK such as mainstream, policy areas and the topic hierarchy, introduced towards the end of the transition project. However none are comprehensive and they don’t support each other in a constructive way.

The topic hierarchy, our chosen single taxonomy, currently only has a few hundred topic entries, meaning many areas of the site aren't yet represented. To represent all the content on GOV.UK we will need many thousands of topic tags. We plan to work very closely with content creators across government, subject matter experts and end users of content to expand the available list of topic tags.

Tagging content to the topic hierarchy

Most of the content on GOV.UK can already be tagged to topics, but only around 4% of it is. This is partly because we have only a limited range of topics right now, but also because the workflow and tools for tagging content to topics aren’t as good as they need to be.

A common observation from our research with users of the publishing tools is that we aren’t helping them understand what effect tagging a piece of content has on how it’s shown on the site and in navigation structures.

To be done well, tagging needs to be well understood by the entire content community. This is both an application design challenge (for the publishing systems) and a communication and training challenge.

Use the hierarchy to support navigation, orientation and search

Multiple browse structures (mainstream browse, topics, policy areas, collections and previously detailed guide categories) are confusing to GOV.UK end users and publishers alike. A more consistent and joined up grouping will reduce the number of decisions GOV.UK users have to make when browsing the site and therefore make the site easier to understand.

The above outlines a large programme of work which is already well underway. Over the next few weeks we will be sharing more detail about each component part of this work as well as some of the lessons we’ve learnt along the way.

Follow Ben on Twitter.

Sharing and comments

Share this page


  1. Comment by Sonja posted on

    Have you thought about letting users solve the problem and using recommendation engines? Why not let users tag or group the content the way they use it and then see if that can be of use to similar users. As a café owner I could set up my own navigation widget of links to useful government content ...other café owners might choose to follow me and others with similar needs or interests. A user could tell the widget a few things they are interested in/need to know and get shown relevant content which they can also add to their nav widget, tag and rate.

    • Replies to Sonja>

      Comment by Ben Andrews posted on

      Thanks for your suggestions, Sonja.

      This is a great idea which has lots going for it. We've thought about a folksonomy approach but felt this wouldn't help us a, consolidate our navigation in the short term and b, provide the foundation we want quickly enough for future improvements to the sites navigation. It may be something we return to when time allows.

      Thanks again.


  2. Comment by Simon Howard posted on

    I agree that this is probably your biggest challenge at the moment. I am frequently on the receiving end of comments telling me that the site is impossible to navigate. It strikes me that there are two related issues that you haven't captured in this post that I think it would be worth considering.

    Firstly, you mention that people "will often come from search engines and enter deep within the site without context". I agree, but one problem that I regularly hear from users is that PHE content is not labelled as such in search engine results. When I've discussed this, I've been told that this is because is intended to be the trusted "superbrand". I think this is misguided. Users I interact with are often looking specifically for PHE content for clinical purposes, yet often hosts much broader content on the subject - think Ebola as the biggest example. When searching on Google for Ebola treatment information, many clinical users want to find clinical content under the trusted PHE brand, not the plethora of FCO and DfID content which also falls under the brand.

    Secondly, blogs appear not to be included in the search or taxonomy at the moment. Given that blogs are being used (perhaps wrongly, I don't know) for unique and relevant content, this is problematic. Users cannot be reasonably told to look at and then search through related blogs to find the content they want on a single topic - that surely breaks the whole principle of as a single unified repository.

    But, overall, it's great to hear that you are tacking the navigation problem, and I wish you every success with it!

  3. Comment by Ben Andrews posted on

    Hi Simon - thanks for your comment and feedback.

    I think you/your contacts have misunderstood the proposition of GOV.UK. The site is deliberately organisation agnostic. By that I mean we are designing the site so that users don't have to understand the structure of government to interact with it. That said, there are use cases where such a mental model is used and we need to support that too. There is work to be done on both sides.

    Some, but not all, content returned on the GOV.UK search results page show which organisation created the content. We need to make this approach clearer and more consistent.

    Regarding the integration of blogs into navigation and search results, I agree this to needs to be more consistent.



  4. Comment by Sebas posted on

    Will this change the way you currently display "Mainstream" navigation (for example Childcare and parenting) VS the "Government" navigation (for example departments, publications). It's like it's two different sites on the same site now.

    • Replies to Sebas>

      Comment by Ben Andrews posted on

      Hi Sebas - thanks for your comment.

      I'm not able to answer your question with certainty yet as we're working through the taxononomy part of the problem right now. The taxonomy and the presentation of the navigation / information architecture are not the same thing.

      I agree there is currently a friction between the mainstream and departments and policy areas of the site. I want to make this work more effectively than it does at the moment but whether that means removing or combining them somehow, we don't know yet.

      Lots to do.

      Thanks again,


  5. Comment by JM posted on

    What will happen with "topical events" will they continue to exist in this new model?

    I like this new approach with topics and subtopics, but if "Economy" is a topic and "Budget" is a subtopic, how can you present everything that is connected to "Summer budget 2015".

    Perhaps topical events can exist inside a subtopic?

    • Replies to JM>

      Comment by Ben Andrews posted on

      Hi Joel.

      Thanks for your comment.

      We are reviewing all the 'content types' used on GOV.UK with a plan to reduce these down if we can. I don't expect Topical Events as a content types to go away as it's frequently used. Like any content type, I would expect it to be taggable to a topic.

      I hope that helps.



  6. Comment by Shirley Dick posted on

    I know you guys don't like over engineered interfaces (neither do I) but some way of allowing users to favourite or otherwise group together the things they do most on GOV.UK might be useful. Indeed, as a (contract) researcher/designer on Gov projects myself, there are bits of GOV.UK I refer to regularly, such as pages related to the department I'm working for at the time. Perhaps presenting users with a wee panel that links them to the last n services/pages they visited?

  7. Comment by JM posted on

    Do you have any progress on this matter, will you go live with anything regarding this new navigation soon? Or share insights, concepts from testing.