.svg)
Get a Demo
Interested in trying Goodie? fill out this form and we'll be in touch with you.
Oops! Something went wrong while submitting the form.
.svg)
.svg)
Buyers researching SaaS tools are increasingly turning to LLMs and AI engines as a means to do product research. Between February 1 and June 30, 2025, Goodie captured 5.7 million citations that reveal which domains these models trust and surface most often for B2B SaaS prompts.
So, which domains do LLMs lean on when it comes to B2B SaaS? And what does that mean for vendors trying to get discovered? Let’s dig in.
To give context to the data, let’s set the stage and describe how we got to our results and conclusions.
Between February 1 and June 30, 2025, Goodie captured 5.7 million citation links from ChatGPT (GPT-4o), Gemini 1.5, Claude 3.7 Sonnet, and Perplexity Pro.
From those, we isolated ~118,067 links tied to commercial, informational, and navigational prompts about B2B SaaS products, including categories like MarTech, Project Management, HR, CRM, Collaboration, and more.
These citations were then deduplicated at the root-domain level, and each domain tagged into a single content bucket: Social & UGC, Affiliates & Listicles, News & Publishers, Commerce, Reference, or Academia, then scored.
We calculated two key metrics:
This approach highlights which publishers LLMs lean on most often in real SaaS buying scenarios, while smoothing out differences in citation volumes across models.
Across the models we analyzed, a small set of domains capture a disproportionate share of citations.
The practical implication? If your SaaS brand wants to be discovered in the AI customer journey, you need a citation strategy that spans three arenas: affiliate listicles, UGC forums, and high-authority publishers (spoiler alert: in that order, depending on the model).
Here are the most important takeaways:
Before we get into a more detailed breakdown of how LLMs cite B2B SaaS domains, we’ll begin with a high-level overview of the top 10 domains across all of the models we studied (Gemini, Claude, ChatGPT, and Perplexity).
What stands out from these findings?
To give a more accurate picture of which LLM citations your B2B SaaS domain could be showing up in, based on where conversations about your brand are happening, let’s convert this data into a heatmap:
When leading AI models spit out synthesized answers about B2B SaaS-related queries, they are citing a mix of validated and unvalidated sources; these sources can be organized into three distinct categories:
For brands, this signals a need to first be captured in these corners of the internet (organic conversations, ranking listicles, and trusted publishers) before being cited in LLMs.
As mentioned previously, not every LLM cites equally; some favor affiliate sites or listicle-type content, while others rely more heavily on social proof from users. Let’s review general personality and findings by model:
ChatGPT’s top 10 most cited B2B SaaS domains prioritize UGC, community, and review sites: Reddit, G2, PCMag, and Gartner are prominent.
Also, ChatGPT heavily pulls from user-generated content: reviews and vendor comparisons show up often in results.
Gemini’s top 10 most cited B2B SaaS domains tilt toward affiliate sources, listicles, and editorial roundups: PCMag, Capterra, and TechRadar are all featured prominently.
The takeaway? If your goal is visibility in Gemini, editorial review pages and buyer’s guides matter.
Claude can be described as a hybrid of UGC and distinguished publications; it leans on listicles and social proof (like those on Capterra and Reddit), but also consults high-end publishers like Forbes and TechCrunch.
The key takeaway here is the fact that Claude seems to mix both conversational sources and publisher credibility; for a Claude optimization strategy, focus on both ends of the spectrum.
Perplexity pulls from UGC, publisher content, and professional networks, particularly LinkedIn. It’s similar to ChatGPT in leaning on social proof and major publishers, but includes professionals’ networks more often.
For the sake of consistency, the ideal strategy for Perplexity optimization is to mirror what you’re doing for ChatGPT, but mix in a little of Claude; in other words, focus on UGC, but don’t forego sources like Forbes.
Alright, let’s break it down; so far, we’ve talked about the top 10 domains across models, and though the major players are fairly similar, there are a few oddballs to address.
Key Takeaway: ChatGPT has a pretty even spread across our three content categories. That being said, Reddit is the top player (surprise, surprise), with listicle sites PCMag, G2, and TechRadar trailing fairly close behind.
Key Takeaway: On Claude, reigning champion Reddit is usurped by Capterra, with a mix of listicle and affiliate sites, and news and publishers trailing close behind. Another long-time player, CNET, also makes an appearance. This emphasizes Claude’s reliance on vetted sources (but marketers shouldn’t forget about Reddit regardless).
Key Takeaway: Gemini is the odd one out in terms of model rankings. Reddit falls all the way down to sixth place, with the top five spots being dominated by only affiliate and listicle content. This underscores the importance of being included in the industry-wide sources that are trusted by users (and the algorithm).
Key Takeaway: Reddit, a familiar face, is back up top. Perplexity, unlike other models that favor more “official” affiliate or publishing sites, however, emphasizes social and UGC-style domains (namely, LinkedIn, not included in any other model’s top 10, and Wikipedia). This is likely due to the fact that Perplexity is built to crawl the live web first, giving it the most fresh and up-to-date information rather than structured content with traditional search power.
If you’re reading this study and thinking: This is all great information, but how do I turn it into actionable components of my AEO strategy moving forward? Without further ado:
You shouldn’t be targeting people looking for your product specifically; branded queries are virtually nonexistent in LLM interactions, especially in the top and middle of the funnel; 88.1% of searches that use AI are informational. Brands should therefore focus instead on informational and commercial queries that relate to your specific B2B SaaS niche (think “best workflow management tools” versus “ClickUp features”).
Your owned content needs to remain strong, yes, but not for the same purposes that we’re used to hearing (namely, SEO).
Your “Top 10 Ways to Streamline Processes” blog post shouldn’t be built to be directly cited in AI. Rather, putting out high-quality, helpful content is a strategy for getting included in listicles; those listicles are then included in AI, driving up your authority, and the cycle continues.
Get the conversation going in places where you have less control (it’s scary, we know). Join in on Reddit discussions. Suggest edits to relevant Wikipedia articles. Reach out to G2 or Capterra and tell them why your B2B SaaS product should be featured in their article.
Though it’s surprising, investing in quality video content could be a consideration. Whether you’re being reviewed by trusted creators or creating videos of your own, YouTube is one of the top 10 most cited domains for both Perplexity and Gemini.
B2B SaaS is an incredibly wide industry with many niche tools, creating niche problems, which require new niche tools that provide solutions to those new problems. Even if that’s not the space you operate in, users who are searching for products to try out or purchase are very specific about the functionality they need.
This specificity and “there’s a solution for every little problem” mindset make standing out in listicles that actually pertain to your product vital. Don’t try to be included in every article under the sun; focus on the ones that your product actually aligns with.
We’re not asking you to turn your entire search strategy upside down; we’re merely making a point that you should consider expanding your reach. That blog that you post on a few times a month? It’s not cutting it. Your Help Center and extensive library of documentation? It’s great for users who have already bought into your product, but it’s not driving SQLs. It probably hasn’t been for a while.
When a user asks ChatGPT, “what is the best payroll application for companies located in New York with employees located globally, but there are only 20 of us total?”, it’s highly unlikely that you have a blog post targeting that exact query.
The benefit of using AI to search is that users can overload it with keywords without a care for their search being too niche, or getting the dreaded “no results found” screen on Google; they’re comfortable being incredibly specific, using natural language, and having a back-and-forth conversation with the LLM.
And when the LLM returns a response, they sound like your smartest friend; they’ve “read about this platform called X”, or “seen Y publisher recommend Z tool”. LLMs will describe features, pricing, scalability, and even cite user reviews, all from the same screen (no click necessary).
This is a huge reason why things like affiliate listings, product reviews, PR coverage, feature announcements and the public buzz associated with them were once nice-to-haves; but that’s simply not the case anymore.
The more complete, vetted, helpful, positive, and relevant your entire online presence is, the more likely you are to be able to attribute some of those sweet, sweet conversions to chat.openai.com:
At the end of the day, AI search visibility comes down to one thing: citations. If LLMs aren’t citing you, you’re not a part of the answer (and you’re not getting in front of users). The SaaS domains that show up most often are embedded in affiliate roundups, community discussions, and trusted publisher coverage. That’s what models are pulling from.
So the question becomes: where is your brand being cited today, and where are you missing?
That’s where Goodie comes in.
This study was conducted and dissected as a means to show B2B SaaS brands who is dominating the AI search space, and drill down into the “why” behind the charts and tables.
Now that you have the “who” and the “why”, it’s time to move on to our favorite part: the “how”. Enter Goodie.
We’re more than the website where you found this article; we’re the tool that’s pioneering the way brands view and value LLM citations and the AI search space in general. We’re also a B2B SaaS organization (major plot twist). Here’s how we (“we” being us and you; welcome to the party) do it:
Goodie is your one-stop shop for all things AI search, and it all lives in one beautiful dashboard. From there, you can see where your brand is (or isn’t) being cited by LLMs. Filter this data by model, by category, or by timeframe to double-click even deeper into these insights.
Showing up in UGC, listicles, and news sources is crucial, but so is maintaining your own website. After all, you can’t create a strong digital presence without a central hub for it all to lead to. Goodie’s AI Crawler & Agent Analytics functionality shows you exactly how AI crawlers interact with your site; so you can make sure that the right information is being cited.
In an industry made of niches and sub-niches (and the occasional sub-sub-niche), finding the right way to describe your product to get it in front of the right users is half the battle; the other half is actually writing the content. Goodie does both.
Whether it’s a topic you’re missing out on entirely or just one that needs a little boost, Goodie will find it, display it, and create agent-ready content that gets you into LLMs and onto the domains that matter most.
Need to connect all of that data to your ROI? Whether it be session, conversion, impression, or revenue impact data, Goodie helps you track the entire customer journey from beginning to end. Our Traffic & Attribution dashboards are built to be understood, communicated to stakeholders, and shown to your executive team to prove AI search really is happening.
When users move to ChatGPT, Perplexity, Claude, and Gemini (which, yes, is still Google), you have to move with them, or risk losing out on valuable leads. It all starts with the understanding you gained from reading this study.
In short, the B2B SaaS space, though incredibly broad, is very much dominated by 10-15 domains, depending on the model. To show up in LLM responses, brands need to:
At the same time, this is not a case for completely forgetting about your own website and the content that lives on it. Your content must remain relevant, helpful, authoritative, and show expertise. Your structured data must make it easy for AI agents to crawl your website.
Your digital presence as a whole has to be honed, finely tuned, for AI search. Brands that hop on the train now will succeed; those that don’t will fade into obscurity within the 10 blue links.
AI has made quite the splash in SaaS (as it has in other industries). But there are two sides to the story. While AI is increasingly being used to power the B2B SaaS platforms we all use today, users are also increasingly turning to AI as a means to find B2B SaaS products that solve the specific problems they’re having.
Whether your focus is on the former or the latter, it’s important to understand the impact of AI on B2B SaaS.
AI is a threat only if you ignore it. If AI citations favor competitors or third-party reviews over your own assets, buyers can form decisions without ever seeing your product pages. The threat is visibility loss, not AI itself.
You 100% can (and definitely should) be optimizing your own website; your content, technical architecture, and structured data. But that’s only one part of it. LLMs don’t just pull from what they find on your website; they assess your value and legitimacy based on what they find on sources across the web.
This makes earning third-party citations (especially if they fall into the top 10 most cited domains in B2B SaaS bucket) equally as vital.
SEO is about improving your rankings in traditional search engines; AEO is about being part of the definitive answer to a user query. B2B SaaS brands should focus their efforts on being part of the source list that AI engines pull from, not from the top three spots on page one of Google.
.svg)
.svg)
.svg)
.svg)
