The Winner Doesn’t Take It All: ‘Information Gain’ and the New Future of SEO

Content marketing today is a story of underdogs and goliaths, of upstart brands trying to compete with industry giants — or else fight over the scraps they leave behind.

You know what this feels like: when every topic, every keyword, every conceivable avenue of growth is already dominated by an oligopoly of big, well-known brands. Their content is long and comprehensive. It’s immaculately search optimized, and if we’re honest … it’s pretty darn similar to the content we had in mind.

There was a time when it would have been wise to chalk this scenario up as a loss and focus on another keyword that lacked so much competition. But in a world where dozens of brands compete over the same lucrative keywords, you can’t afford to do that. White space is becoming rarer and rarer. Ignore crowded search results, and you’ll soon have no keywords left to target.

But a patent from Google illustrates why this “winner takes all mentality” isn’t necessary. It all boils down to a concept called information gain, and it’s the most important idea in content marketing today.

Winners Through the History of Search

Content marketing used to be simpler than it is today. There was a time when a Google search would yield a page of only vaguely relevant search results; finding an article that addressed your specific question was rare, and incredibly welcome. These articles were the winners, earning the lion’s share of the traffic and engagement on offer.

As more companies entered the content marketing fray, it became easier for searchers to find relevant information: you could probably pluck out useful snippets from most of the top-ranking search results. The criteria for SEO domination shifted to aggregation: any article that consolidated the SERP’s fragmented information into one convenient place became the de facto winner (hence the skyscraper model and all of its problems).

Today, we have a different problem. Instead of hunting for signal amongst the noise, we’re searching for signal amongst … signal. In the most mature of industries, every search engine results page (SERP) feels like it’s contested by half a dozen huge brands, with long, ultra-comprehensive articles built on mountains of backlinks and backed up by rigorous on-page SEO. These articles generally look and sound the same, containing minor variations of the same basic information.

Through the old mindset of “winner takes all,” these SERPs are lost causes. Huge, established brands can trade first-page rankings back and forth indefinitely; smaller, less established blogs will never overcome the moat of domain authority and brand awareness. But in the same way that content marketing has evolved over time, so too has Google.

Information Gain

In April 2020, Google filed a patent that set out to solve a familiar problem:

“When a set of documents is identified that share a topic, many of the documents may include similar information … a user may submit a search related to resolving a computer issue … and may subsequently be provided with multiple documents that include a similar listing of solutions, remedial steps, resources, etc.”

This is Google recognizing the problem we run into every day: copycat content. Most search results contain the same information. Once a reader has read one article, they’ve effectively read them all.

For Google, this is a problem of user engagement: copycat content curtails the user journey, removing any incentive to continue searching and discovering additional helpful information. For content marketers, this is a problem of differentiation: unless you’re able to outrank those huge skyscraper articles, there’s little incentive to bother contesting the keyword.

Google’s solution is straightforward: reward articles that bring new information to the table.

They suggest doing this through something they call information gain, a measurement of the new information provided by a given article, over and above the information present in other articles on the same topic.

Google's patent maps out a possible workflow for adding information gain to search results.

The simplest way to measure information gain is to compare the content of a given article to those already ranking in the SERP, but Google suggests going a step further and taking into account the user’s search journey to work out what constitutes “new information.”

In other words, information gain will be relative to the competing content but also to the articles each searcher has read.

Google suggests a few ways to use this information:

  • Showing responsive, personalized search results to users that change according to the content they’ve previously read.
  • Facilitating smarter conversations with voice search and chatbots.
  • Influencing how summary information is extracted from articles and shared with searchers.

Rewarding Content for Being Different, Not Just Better

The simplest interpretation of this patent is that an article that adds something new to the discussion may rank more highly than an article that repeats the same information as others. Inversely, if your article offers the same core information as one the searcher has already read, it might be ranked lower.

But there are broader implications of the information gain concept:

  • This opens the door to a huge fragmentation of every SERP. The idea of search results as something static and monolithic — a single set of “best” articles serving a single primary search intent — is becoming more and more dated. The content marketing industry is slowly clocking on to the idea of a single SERP serving multiple, fragmented intents, but this patent suggests a future where every searcher has their own customized set of results.
  • Being “different” is a viable search strategy. Ultimately, this is Google rewarding content for being different and not just better. In this new world, risk-taking content — anything that differs from the existing search results in terms of focus or opinion — has a mechanism for being rewarded. Underdogs have a new tool in their struggle to outcompete skyscraper content.
  • There’s incentive to collaborate with the existing content instead of competing. Against this backdrop, it makes sense to think of our articles as being active participants in a collaborative ecosystem of content. SEO is no longer “winner take all”: instead of a single article dominating the SERP for a given topic, there are now possibilities for many more articles to “win,” albeit on a smaller scale.
Information gain gets rid of the idea of a monolithic, "winner takes all" approach to SEO.

Now, the linked-to patent application is just that: a patent application. Google’s ideas always precede actual changes to the search algorithms, and there’s no guarantee that they will come to fruition.

But regardless of the precise mechanics of if and how information gain is implemented, the biggest takeaway from this patent is simply that Google intends to reward articles that contain information other articles don’t. That means every content marketer will be well served by keeping this question front of mind for everything they publish:

"What new information am I bringing to the discussion?"

Here are a few practical ways to factor this question into your content:

1. Create Content That Builds on Its Predecessors

If you acknowledge the idea that your content is part of a collaborative ecosystem, your objective shifts from beating rival articles to complementing them. Instead of trying to blithely outrank that gargantuan skyscraper article, work on the assumption that the reader did read the competing post. If that’s true, how can you add value beyond what they’ve already read?

  • Share a practical “next step,” something that acts as a continuation of the competing content. 
  • Elaborate on a key idea contained within the competing article.
  • Write the 102 version of their 101, going into more depth, detail, and nuance.

2. Experiment With Risky Framings and Angles

Information gain rewards articles for risk-taking and deviating from the status quo. Instead of looking at search results with a view of emulating the top-ranking content, it encourages marketers to differentiate. There are plenty of ways to achieve that goal (while still meeting search intent):

  • Address unserved intent (“My specific use case isn’t represented here.”)
  • Fill in missing information (“It’s weird that no one has mentioned X here.”)
  • Challenge a differing or erroneous opinion (“That’s an outdated belief.”)
  • Correct mistakes in Google’s comprehension (“That’s not what I meant by this keyword.”)

Read more: The ‘Google Knows Best’ Fallacy

3. Build an Information Moat With Original Research

As we’ve written before, “primary research is the ultimate form of information gain.” Anything new and proprietary is the safest and surest way to add new data into the discussion — by definition, anything that you create can’t be found anywhere else. While “original research” sounds intimidating, there are plenty of ways to incorporate it into your content:

  • Include personal perspectives and company experiences
  • Survey your customers, users, or network
  • Add quotes from SMEs

Read more: Copycat Content: SEO Tools Got Us Here, Humans Will Get Us Out

The Underdog Search Engine

As a general rule, most content marketers don’t like acknowledging the competition posed by existing search results (or the reality that sometimes, we just can’t compete with them).

When we do acknowledge the competition, it’s usually to pick and choose elements from their structure in an attempt to create something “better,” leading to the continued arms race of skyscraper articles, each containing the same tired information as the articles that came before.

“Information gain” is Google’s attempt to curtail this problem and create a reward mechanism for articles that add new information to the discussion — and aspire to be different instead of just better. The future it presents is good, creating a path to more helpful, interesting search results but also unlocking new ways for smaller, underdog companies to benefit from search simply by answering the single question:

What new information am I bringing to the discussion?”

Thank you to Camden Gaspar and the late Bill Slawski for introducing me to the concept of "information gain." You can read Camden's article and Bill's patent analysis here and here.