We all know that SEO is about marketing, getting bums on seats and eyeballs on pages, while Information Architecture is about structuring content in a way that makes sense to users and helps them find what they are looking for.
But it turns out that seeing both of these as two parts of the same customer experience might not only help SEO deliver better traffic, but also produce a more user-friendly site structure.
Judd Garratt is @juddgarratt
Justin Tauber is @brtrx
1. What SEO brings toInformation Architecture Justin Tauber Judd GarrattOctober 2010
2. What about SEO? You say: If Google can't find what it's looking for in this site,then there's something wrong with Google. The site is correct. If those ham-fisted SEOs touch my labels, I'll personally organise a denial of service attack for launch. That's a marketing problem.
4. To relinquish all control of SEO amounts to: Giving up control of how user expectations of the site are set Ignoring the linguistic context of the labelling of the site
12. To relinquish all control of SEO amounts to: Giving up control of how user expectations of the site are set Ignoring the linguistic context of the labelling of the site
13. Your Site Journey Journey Google Other Site Journey Journey Journey Other Site Other Site
20. 1. Brief SEO with your taskanalysis and personas
21.
22. 1. Brief SEO with your taskanalysis and personas 2. Test the topicality of yourlabels in Google
23. 3. Get action labels from SEO data Some labels don't just represent buckets of content, they are calls to action e.g. join, sign up, purchase, etc What terms are used by users who complete a task How do people actually describe these actions?
24. 4. Be part of the decision to create new contentfor SEO purposes 5. Ensure new contentfits within the users’ mental model cards sorts and/or tree analysis
Can’t SEO and IA ever be complementary?Isn’t there some way IAs could make search engine optimisation into something more user-centred than keyword stuffing? Couldn’t keyword analysis actually contribute to producing a better IA?
We’ll explain these in more detail below.Let’s start with user expectations…
There’s the perception that SEO is just marketing creates a division of labour. SEO is for marketing, IA for better user experience.This division of labour sets up an implicit barrier between what happens in search engines (SEO), from what happens on the site (IA), but…
Aren’t these just two aspects of the same user journey?
The first question a user asks on landing on your site is: Am I in the right place?The first task of an IA should be to ensure users answer that question in the affirmative.How a user answers this questions depends on their expectations.
If the Search Engine Results Page (or SERP) sets the wrong expectations for visitors, these can’t be fulfilled by the site, no matter how good the IA might be.
Is there really a site offering a roadside psychological counselling service, for just $55 per annum?The expectations set by this context, can never be fulfilled by this site.
For all we know, Shoalhaven Community Credit Union a) provides financial services, and b) collects personal data. Does the privacy policy really set the best expectations for all the tasks that the site fulfills?
Search engine results pages and the site are two aspects of the same journey.If we, as IAs, dismiss SEO as just marketing, we run the risk of being unable to set expectations and provide a coherent journey for a great many of our users.
Your site is not the internet. Users are likely to be working with content from multiple sites simultaneously (some of them with have multiple sites open in different tabs).
You do not have a captive audience, so you can’t dictate the appropriate language to be used when labelling content. Competitor analysis can help to determine labelling conventions, but we’d argue keyword analysis is an even better source. After all, while your competitors might all get it wrong, it’s hard to argue that your users are searching for the wrong term.
You could, of course, get your users to apply labels to the content groupings they produce in a card sort. However, one of the most implausible assumptions in IA/ UX is that users are reliable reporters of their use of language. It’s worth at least complementing a labelling exercise in a card sort with data on which terms users actually use when looking for a particular set of content. Keyword analysis can provide this.
A different division of labour. Concepts are the glue that unifies items in a collection. Card sorts are a great way of picking out the appropriate concepts (the mental map of your users). However, keyword analysis is often a more reliable source for determining how users express that concept in language.
It can be worthwhile to turn to search data to empirically validate your labelling
Search engine optimisation of content can actually be a way of diminishing the risk of picking the wrong label, by increasing the likelihood that users do find the term they’re looking for…
While Service & Support (which seems to have been optimized for search) may be a key area of the business, is this the right starting point for users? Making sure SEO is working to the same objectives is important.Also, while the page description (the text below the link on the results page)doesn’t affect page ranking, it clearly does affect user expectations. As IAs, we should probably pay more attention to these descriptions.
If you do nothing else, do these two things, and we suspect you’ll significantly improve the user experience for users coming from search
This is where IAs can legitimately demand a seat at the SEO table.
It’s just a matter of seeing the search queries that list your site content as part of the user journey: that is, as both indicating and setting user expectations.