In 2022, we wrote about a Google patent application on "information gain," a theory that the search engine would reward content for being different. Add new information to the online conversation, we thought, and you’ll be rewarded.
The concept became popular among SEO practitioners, but it was still just theory. (Google files ideas all the time that never make an impact outside the patent office.) In reality, as far as we could tell, the SERPs remained full of copycat content — essentially a load of nearly identical “comprehensive guides” on any particular topic.
But now AI has made the information gain theory impossible to ignore. ChatGPT has read the internet. Claude has read the internet. They can synthesize, reiterate, and repackage everything that's already been published.
The question becomes: Why would you publish anything that's not additive?
From Displacement to Differentiation
Under the traditional SEO model, you optimized to beat the top-ranking article. You analyzed what ranked, identified gaps, and built something more comprehensive. The goal was displacement: knock the other article off or page one, or better yet, out of first position.
Under the new, “information gain” model, you optimize to complement existing sources. When Google (or any AI) synthesizes its answers from multiple sources, your job is to offer what those sources don't. The goal now is differentiation, to contribute something they can't get elsewhere.
In practice, this means competing to get cited, not competing to rank first. And that creates an opening for smaller brands. Domain authority still matters, of course — pages ranking in the top 10 appear in AI Overviews more than half the time. But you don't need to outrank giants if your content contains information theirs doesn't. You need to out-differentiate them.
If your content repeats what 10 other articles already say, AI makes it redundant before you hit publish.
When Google synthesizes an answer, it cites an average of five different sources. The content that gets cited is the content that contributes something new. The rest gets absorbed into the synthesis without attribution.
We can't prove Google uses the "information gain" algorithm from that 2020 patent — research points to "sufficient context," "complementarity," and "multi-source validation" as the mechanisms — but the principle we predicted is playing out: differentiation matters more than comprehensiveness.
4 Strategies for Information Gain in the AI Era
Differentiation — whether through 'information gain' or mechanisms like 'sufficient context' and 'complementarity' — is answer engine optimization (AEO) in practice.
Here's how to build it into your content strategy:
1. Build a Moat With Original Research
As we’ve written before, “primary research is the ultimate form of information gain.” Proprietary data is the surest way to add new data into the discussion. By definition, information you create can’t be found anywhere else.
Original research doesn't require expensive market research or complex studies. It can be:
Customer surveys or user data. Use data you already have and collect. Product usage statistics, customer behavior patterns, or aggregated feedback from your user base gives you proprietary information to cite.
Personal perspectives and company experiences. Your direct experience implementing a strategy, the specific results you achieved, and the obstacles you encountered can't be replicated by other sources. When AI synthesizes information about a topic, first-hand accounts add context that generic advice lacks.
Expert interviews and quotes. Conversations with practitioners in your network create original content even on well-covered topics. When you include perspectives from people actually doing the work, you bring voices to the discussion that other articles don't have access to.
An information moat also compounds over time. As you publish original research, you become the primary source for that data. Other articles cite you, AI Overviews reference you, and your authority in the space grows.
2. Create Content That Builds On (or Challenges) Its Predecessors
Instead of trying to create a more comprehensive version of existing top-ranking articles, assume AI has already synthesized the core information. Your job is to offer what those articles don't.
Analyze what the current top results cover. Where do they stop? What questions do they leave unanswered? What's the logical next step they don't address? Then create content that fills those gaps:
Share a practical next step. If existing articles explain a concept, show how to implement it. If they cover theory, provide a tactical walkthrough. When AI synthesizes "what is X" from five sources, being the one source that answers "how to actually use X" makes you distinct.
Elaborate on a key idea. Pick one concept that existing articles mention briefly and go deeper. When top-ranking content covers 10 strategies at surface level, your detailed analysis of one strategy offers information the others lack.
Write the 102 version of their 101. Existing articles handle the basics well. Create content for readers who already understand fundamentals and need the next level of depth, nuance, or complexity. AI can pull introductory information from established sources and cite you for advanced insights.
Complementing existing content is one path. Contradicting it is another.
3. Experiment With Risky Framings and Angles
Information gain rewards differentiation. The safest content strategy — matching what already ranks — becomes toothless when the goal is to stand out.
Search results often converge around a single interpretation or approach. That convergence creates opportunity. AI Overviews pull from multiple sources specifically to provide a complete picture, which means there's value in being the contrarian voice or the alternative perspective.
Challenge outdated beliefs. Search results often reinforce practices that are no longer effective. When top-ranking articles recommend tactics that stopped working two years ago, your updated perspective adds new information to the discussion.
Take a strong stance. Generic advice that tries to please everyone says nothing distinctive. "It depends" might be accurate, but "here's exactly what works and why" gives AI something specific to cite when synthesizing different viewpoints.
The content that looks riskiest — the article that contradicts conventional wisdom or focuses on an underserved angle — can become the most valuable in AI Overviews.
4. Write for a Specific Cohort
Audience segmentation is differentiation in practice. A 2025 study of 300 B2B SaaS websites found that companies segmenting by industry increased Top 10 Google rankings by 43.4% on average. Companies without segmentation saw rankings decline by 37.6%. The segmented sites achieved 15.7X higher organic traffic growth.
Instead of writing "The Ultimate Guide to Customer Retention" for everyone, you write "Customer Retention for Fintech Startups" or "Retention Strategies for Healthcare Platforms." The more niche, the better. Narrow content creates information gain because industry-specific advice can't be replicated by generic articles.
You can narrow by:
Company size. Retention strategies for enterprise companies differ from strategies for startups. The budget constraints, team structures, and decision-making processes are completely different.
Experience level. Advanced practitioners need different content than beginners. Writing "Email Marketing for Teams Already Doing A/B Testing" targets readers who don't need Email Marketing 101.
Use case. Instead of "How to Use Our Product," write "How E-Commerce Brands Use Our Product for Abandoned Cart Recovery" or "How SaaS Companies Use Our Product for Onboarding."
Long-tail, industry-specific queries will be more helpful. You don't need to abandon broad topics entirely. But when you do cover them, bring a narrow lens. The audience you exclude becomes the differentiation that makes you citable.
The End of “Comprehensive”
For years, "comprehensive" was the goal. Cover everything. Address every angle. Build the Single Definitive Resource that consolidates all available information in one place.
Back in 2022, we predicted this consolidation arms race would end, and now it has with AI.
Now that AI can compile and synthesize comprehensive coverage from ten articles in seconds, "comprehensive" is no longer the differentiator — it's the baseline. The information gain theory has evolved from patent filing to practical necessity, from nice-to-have to required.
Which means every piece of content now demands a more honest question: "Does this need to exist?" If AI can already answer it by synthesizing existing sources, you're probably better off not publishing. That admission saves time you can spend creating something that actually contributes — and something that actually gets cited.