For the past two years, Google’s AI Overviews have been changing the economics of informational content.
The logic was simple: if Google can answer a user’s question directly in the search results, why would that user click through to the article?
That led to a strategic shift in SEO.
Instead of investing heavily in educational blog posts, many SEO teams started looking for content that AI couldn’t easily replicate—interactive tools, templates, calculators, proprietary data, and other “value-add” assets.
But there’s a new problem with that strategy.
Google has now started bringing interactive experiences and tools directly into its AI search experience too.
So if Google can summarize your article and surface a tool without sending the user to your website, what kind of content can still earn visibility and traffic?
We wanted to find out.
So, we analyzed Google Search Console and AI visibility data across 7 B2B SaaS clients to understand:
- Which page types drive the most clicks (survive AIO)
- Which page types Google AI and LLMs cite the most
- What characteristics separate cited pages vs non-cited pages
Here’s what we found.
Key Findings
- There’s no universal “best content format” for AI. Each AI platform has their own preferences. AI Overviews and AI Mode mostly cite informational content, ChatGPT prefers product pages, while Gemini and Perplexity lean on listicles.
- Informational content still works. It’s the most visible content type in AI Overviews and AI Mode, and it’s the second biggest click driver behind homepages.
- Data-backed content is key to winning visibility in Google AI. The top-cited informational articles in AI Overviews and AI Mode have 36% more statistics than least-cited pages.
- Organic performance correlates with higher Google AI visibility. Pages that are invisible in AI Overviews and AI Mode have zero keyword rankings, zero traffic, and zero backlinks.
- Listicles are the most-cited content type in LLMs. Listicles dominate citation share in Gemini and Perplexity. In ChatGPT, however, it’s the second most cited content type behind product pages.
Let’s dive deeper into the full study.
1. Informational Content Still Drive Clicks
The first assumption we wanted to test was whether AI search had already made informational content obsolete.
The data says no.
We analyzed Google Search Console data of our B2B SaaS clients between March and August 2026.
Despite AI Overviews taking away some of the clicks, informational content is still the second biggest click driver among B2B SaaS companies, behind only homepages.
| Page Type | Share of Clicks |
| Homepage | 40.3% |
| Informational article | 23.4% |
| Templates & tools | 13.8% |
| Account / login | 12.9% |
| Docs / help | 3.2% |
| Listicle | 1.0% |
| Careers | 0.8% |
| Announcement / product update | 0.8% |
| About / company | 0.7% |
| Pricing | 0.6% |
| Product / feature page | 0.5% |
| Comparison article | 0.5% |
| Industry / use-case page | 0.3% |
| Blog hub / index | 0.3% |
| Integrations | 0.2% |
| Legal | 0.2% |
| Gated content / resources | 0.2% |
| Case study | 0.1% |
| Contact / demo | 0.1% |
| Expert interview | 0.1% |
Templates and tools came in third place with a 13.8% share. But the number might go down as Google shows more tools inside AI Overviews.
What this tells us
The strategic response to AI search shouldn’t be to abandon informational content as a whole.
Instead, the better question should be:
How do we make informational content valuable enough that Google wants to cite it—and users still want to visit it?
2. Informational Content is the Most Cited Page Type in Google AI
There’s another reason not to abandon informational content: it’s by far the most cited page type in AI Overviews and AI Mode with a 63.8% share.
Far ahead of homepage (9.9%), templates & tools (9.1%), and listicles (8.9%).
| Page Type | Share of Citations |
| Informational article | 63.8% |
| Homepage | 9.9% |
| Templates & tools | 9.1% |
| Listicle | 8.9% |
| Comparison article | 2.4% |
| Announcement / product update | 1.9% |
| Docs / help | 1.5% |
| Industry / use-case page | 0.8% |
| Product / feature page | 0.7% |
| Account / login | 0.2% |
| Integrations | 0.2% |
| About / company | 0.1% |
| Case study | 0.1% |
| Pricing | 0.1% |
| Gated content / resources | 0.1% |
| Q&A page | 0.1% |
What this tells us
AI visibility doesn’t appear to require a completely new type of content.
The same articles that were built to answer users’ questions in traditional search can also become sources for Google’s AI-generated answers.
The optimization challenge has changed—but the underlying content asset remains valuable.
3. Top Cited Content Has 36% More Data Points
For each client, we compared their informational blog posts with the highest AI impressions vs those with minimal visibility.
Here’s what we found:
Top-cited informational pages include 36% more numerical data points and back their claims with roughly 2x more external sources, compared to pages with low impressions.
| Metric | High-impression (median) | Low-impression (median) |
| Numeric data points | 17.0 | 12.5 |
| External links | 5.0 | 2.5 |
What this tells us
The pattern suggests that specificity and evidence may be important characteristics of AI-citable content.
Consider the difference between:
“Most companies struggle with employee retention.”
And:
“42% of mid-market SaaS companies reported employee retention as a top HR challenge in 2025.”
The second statement gives an AI system something much more concrete to work with.
Numerical claims, benchmarks, statistics, and specific examples create discrete pieces of information that can be extracted, synthesized, and attributed.
4. The Least Cited Pages Have Zero Organic Visibility
We also found that pages with low AI Overview and AI Mode impressions have zero visibility in organic search.
They have zero keywords ranking in the top 10, zero traffic, and zero backlinks pointing to them.
What this tells us
This suggests that AI search isn’t a completely separate system from traditional SEO.
Google’s AI experiences still operate on top of Google’s web index and ranking ecosystem. A page that has never demonstrated relevance or authority in traditional search may have little chance of becoming a source for an AI answer.
5. But LLMs Tell a Different Story: Listicles Dominate Citations
So far, the story has been about Google.
But Google isn’t the only place people are discovering B2B SaaS products and information.
We also looked at citations from ChatGPT, Gemini, and Perplexity.
Across 7,258 citations, listicles were the most frequently cited page type, accounting for 23.1% of all citations.
| Page Type | Share of Citations |
| Listicle | 23.1% |
| Informational article | 15.4% |
| Product / feature page | 15.0% |
| Documentation / help | 8.3% |
| Homepage | 8.3% |
| Gov / regulator page | 5.1% |
| Other | 4.2% |
| How-to guide | 3.8% |
| Glossary / definition | 3.1% |
| Social / blog platform | 2.3% |
| Review / directory | 2.1% |
| Video | 2.1% |
| Template / free tool | 1.4% |
| Forum / UGC thread | 1.1% |
| Comparison article | 1.0% |
| Research paper | 0.9% |
| News article | 0.7% |
| Careers | 0.6% |
| Legal / trust | 0.6% |
| About / company | 0.5% |
| Pricing page | 0.4% |
| Case study | 0.1% |
But when we broke it down by each AI platform, an interesting pattern emerged:
ChatGPT favors product/feature pages slightly more (18.9%) than listicles (17.7%). For Gemini and Perplexity, listicles remain the top performers.
| Page Type | ChatGPT | Gemini | Perplexity |
| Listicle | 17.7% | 24.5% | 31.9% |
| Informational article | 12.0% | 16.7% | 20.2% |
| Product / feature page | 18.9% | 13.3% | 12.8% |
| Documentation / help | 14.4% | 3.4% | 3.3% |
| Homepage | 7.6% | 9.8% | 3.2% |
| Gov / regulator page | 8.0% | 3.3% | 2.6% |
| How-to guide | 2.3% | 4.6% | 5.4% |
| Video | 0.1% | 5.7% | 0.0% |
| Social / blog platform | 1.0% | 3.7% | 3.3% |
| Review / directory | 2.8% | 0.9% | 2.6% |
| Comparison article | 0.7% | 1.2% | 1.6% |
What this tells us
LLMs love citing listicles because they have a highly extractable structure.
“10 best project management tools” gives an AI system a clearly defined set of entities to evaluate and potentially recommend.
Each H2 can represent a discrete answer.
Each item can contain a description, use case, pricing information, pros and cons, or other facts.
In other words, listicles package information in a format that’s easy to decompose.
That makes them particularly well suited to recommendation and comparison-style prompts—the kinds of queries users increasingly ask AI systems.
6. What Makes LLMs Cite a Listicle?
We compared the best performers (listicles with multiple citations) with worst performers (listicles with only one citation).
Here are the top factors that separate the two:
- Adding a number in the title. “Top 10 CRM Tools” performs better than “Best CRM Tools.”
- Number of items covered. We found that the top-cited listicles have 13 headings on average, compared to 9 for low-performing ones.
- Writing headings as questions. Optimize for real questions people are asking, such as “what’s the best X software for Y?”
| Signal | Top-cited Listicles | One-Off Listicles |
| Title contains a number ("Top 10 X") | 59.1% | 39.6% |
| H2 headings (items covered) | 13.0 median | 9.0 median |
| Question-form headings | 3.5 | 2.0 |
What this tells us
The more clearly a listicle is structured around discrete questions and entities, the more citation opportunities it appears to create.
This doesn’t mean every listicle needs to be stuffed with headings. The goal is to make each section independently useful and easy for an AI system to understand, extract, and cite.
What You Need to Do
Let’s turn this study into an action plan:
1. Don't neglect traditional SEO
AI search isn’t replacing traditional SEO. Our data suggests that better rankings, organic traffic and backlink profiles correlate with greater visibility in Google AI Overviews and AI Mode.
So before creating an entirely new AI search strategy, make sure to build strong SEO fundamentals.
If you’re not sure where to begin, consider working with a SaaS SEO agency like Position Digital.
2. Don't stop publishing informational content
Informational content remains one of your most valuable content assets.
It was the second biggest driver of organic clicks in our dataset and accounted for 63.8% of Google AI impressions.
AI search hasn’t made informational content obsolete. It has raised the bar for what makes an informational article worth citing and clicking.
3. Make your content more trustworthy and extractable with data
The informational articles with the highest AI visibility in our study contained 36% more numerical data points and roughly twice as many external sources as low-visibility pages.
Data, evidence and specific examples can turn generic advice into citable information.
Where possible, incorporate:
- Original research
- Industry benchmarks
- First-party data
- Survey results
- Statistics and numerical evidence
- Specific real-world examples
- Citations to authoritative external sources
Our content marketing services can help you create data-driven, authoritative content that earns visibility in both traditional and AI search.
4. Structure listicles around entities and questions
Listicles accounted for 23.1% of all LLM citations in our dataset, making them the most frequently cited page type overall.
But simply publishing a numbered list isn’t enough.
Structure each entry around a clearly defined entity and the questions a user is likely to ask about it.
For example:
H2: 1. Monday.com
H3: What is Monday.com best for?
[Answer]
(H3) How much does Monday.com cost?
[Answer]
(H3) What are the pros and cons?
[Answer]
(H3) Who should use Monday.com?
[Answer]
This creates clearly defined information blocks that an AI system can understand, extract and potentially cite.
5. Get featured in third-party listicles
Don’t just publish your own listicles. It’s equally important to get featured in third-party roundups that LLMs already cite.
This can put your brand directly in the sources these models use to build recommendations.
Use ListBrew to identify the listicles that rank for your target categories and appear frequently in AI responses. Then prioritize the pages where a relevant inclusion, update or link is realistically achievable.
If you prefer having an expert to do it for you, consider Position Digital’s listicle link building services.
Want More AI SEO Insights?
Explore more AI SEO studies that can help you increase AI visibility in 2026:
Or simply talk to our GEO specialist. Tell us what you need, and let’s build a plan to get your software recommended by AI engines!
