SEO | March 18, 2019
What You Know About User Intent is Probably Wrong
The way we talk about and consider search intent hasn’t changed much since 2002 and Google’s ranking factors have matured immensely since then.
Do you classify search intent as Navigational, Informational, or Transactional?
- Navigational – Is the intent to reach somewhere particular.
- Informational – Is the intent to learn and acquire information.
- Transactional – Is the intent to purchase.
Well, you’ll be surprised to know that this methodology dates back to 2002 – almost two decades ago, even before the first major Google algorithm update ‘Florida’!
Fast forward to the early 2010s and Google begins using their own terminology to identify search intent, Know, Go, Do, and Buy. Which, we agree, covers the same categories of search as above, but as is typical of Google are a little more user friendly.
#1 Too Broad.
Whilst the search intent classifications are great for hypothetical examples, say for SEO beginners or clients, they simply don’t work in practice.
Start applying the classifications to keyword research and the system falls apart. Trying to understand how [cat videos] and [cat café] are both labelled ‘navigational’ really disrupts the system and makes it difficult for content creators to run with. All of a sudden, the seemingly simple classifications aren’t so simple anymore.
#2 Doesn’t account for overlapping intent.
What happens when a search query falls into multiple classifications? Is [cat mat] ‘navigational’, to watch Fat Cat Mat on YouTube? Informational, for information on cat litter mats? Or, transactional, to purchase a cat feeding mat?
(…we like cats)
#3 Labels are often guessed by the query, not the results.
Did you stop to think “let’s refer to the search results” for the problems above? Of course not! Because we instinctively guess classifications based upon the query. And this is common among SEO strategists too, who manually label a long list of keywords with what we expect is its specific intent.
What needs to happen is for SEO’s to move away from understanding the true intent of users, and instead concentrate on the type of content Google serves to users based upon what Google already knows about a user’s intent.
What we know about Google’s publishing factors.
We can gather a lot of information on how Google is looking to develop their understanding of search and intent through their many patents. Whilst there’s no guarantee that any patent’s mechanism is used, it’s a good indicator as to what Google are researching and looking into.
Some of these are Google’s changing crawl behaviours, and it’s greater understanding of context and categorisation. We recommend reading our “best site structure” blog for a greater insight into context and categories.
It seems that Google may no longer simply crawl a URL to discover new pages or to see if the page has changed, it can also classify URLs by intent and topic. Inherently refining the probability that a URL is a good match for a given search.
Top tip no.1: Monitor crawl behaviour for correlations to search rankings.
Since the crawl function is a constant loop, always adjusting the classification and probability of a page, it could be an indication as soon as a page stops being crawled that it is no longer considered useful by Google.
Top tip no.2: Include semantic analysis of web pages.
Google classifies URLs by intent and topic and creates relevant ‘clusters’ i.e. topic groups by the similarity between words that form the concepts on a page. As these ‘clusters’ are formed on related content, a semantic analysis will indicate the probability of clustering on a website and help predict where content is lacking.
We’re proud to share that our Senior Technical SEO Specialist has built Liberty a bespoke tool to analyse and help visualise semantic analysis.
Top tip no.3: Develop a content strategy around topic hubs.
In other words, write more than one interesting piece on any given item of content. Creating pages on the same topic will mature a ‘cluster’ and help promote the pages in SERPs, simply because it increases the probability that Google’s classification is correct.
By switching our analysis of search intent away from the user (i.e. the query), we are able to better serve Google’s publishing factors, resulting in what we were trying to achieve all along… To appear in the SERPs for a given query.