You may, or may not have seen the recent articles floating around about how Google has a very obvious, but tricky problem to solve.
It lies in the way that very biased queries that take opposite views on the same subject can return some very contradicting answers.
Take these queries for example. Type in “are chow chows good pets” and “are chow chows bad pets” in two different tabs on your browser.
The answers you see are two very biased views based on the preconceived notion of both queries.
Ironically, the second result actually answers both sides of the argument. However, Google has taken the sentiment of the query and assumed that the searcher only wants the negative information on this subject.
This is because, and to quote Google “Results are algorithmically ranked by a balance of authoritativeness (how often a site is linked to by other sites) and relevancy (how well the content matches the query). When a user asks “are reptiles bad pets,” the algorithm uses the exact text to assess for content relevancy. As a result, a site that uses the words ‘reptiles,’ ‘bad,’ and ‘pets’ more often would naturally be ranked above another site that uses ‘reptiles,’ ‘good,’ ‘pets.’”
Here is another example.
We all know that we need water to survive, so why is Google not returning a balanced argument? Again, the sentiment of the query assumes you only want to see the negative implications of drinking water. In fact, Google already has an idea of the best answer, see the “People also ask” section below both of these queries.
The information already exists, so it looks to me like a slight tweak is required to ensure that both arguments are presented in these types of queries.
Google will be fixing this issue by training the search engine to understand the intent of both queries as one. Using the above sentiment, this would be “how are chow chows as pets” and the same article that matches this sentiment would be returned for both queries.
Implications and thoughts:
As an SEO agency, the content we review on a daily basis and the way that we write our content is based around sentiment, theming and where the user is in the query funnel for a given query. We try to meet user intent behind the query so as to ensure that the result deserves to be in prominent positions, tailoring the functionality of pages to best meet user needs.
This should hopefully mean that once this filter bubble is fixed, we will see articles triggering and returning for more answer boxes around the same query intent.
Ignoring the answer box, Google already allows for a very comprehensive article that covers a subject in a comprehensive manner to rank for both questions. The opportunity already exists in how we produce content and user experiences whilst taking advantage of semantic best practices, such as Schema and Rich results (now a basic suggestion in Google’s SEO Starter guide).
Matching individual queries to single pages can still have some benefits, but only when the query does not overlap very heavily with a specific theme of keywords.
Gone are the days where you would optimise for single keywords. Keyword research now starts with a “theme” or “seed” topic instead, where relevant queries and supporting queries are collected and themed into a keyword content map with all potential queries.
Potential queries are then scored and matched against current content to outline where content is of sufficient depth to also highlight gaps in content topics and the depth of the site.
This allows us to match our keyword research to a themed approach to creating content and addressing clear content gaps where we believe users to be in the intent funnel.
Google has admitted this filter bubble issue in a blog post here.
They state that:
“Showing more than one featured snippet may also eventually help in cases where you can get contradictory information when asking about the same thing but in different ways.
For instance, people who search for “are reptiles good pets” should get the same featured snippet as “are reptiles bad pets” since they are seeking the same information: how do reptiles rate as pets? However, the featured snippets we serve contradict each other.“
This fix is scheduled for a rollout in Q2 for a sample of simpler queries for testing. I think this is one that needs attention ASAP. I look forward to the rollout as I believe this will benefit agencies who think about the implications of content depth and user intent already.