Walk into a new coffee shop in almost any big city, and there’s a good chance you already know what it looks like. Raw wood tables, exposed brick, and Edison bulbs hanging over the counter.
In 2016, the writer Kyle Chayka gave that look a name. In an essay for The Verge called “Welcome to AirSpace,” he described cafés from Odessa to Seoul that shared the same style. They weren’t chains, and nobody told them to match. Their owners had arrived at the same look on their own, shaped by the same apps, the same social feeds, and the same photos of what a good café is supposed to look like.
B2B SaaS content has the same problem. Nobody sets out to publish the same article as a competitor, yet so much of what ranks reads like it came off one production line. We’re all drinking from the same pond, and the water tastes the same no matter whose cup it’s in.
I read 22 articles on three B2B SaaS topics
To see how deep the pond goes, I searched three topics a B2B SaaS content team might write about: customer onboarding best practices, what revenue operations is, and sales enablement strategy. On September 25, 2026, I pulled the articles that came back for each search and compared their headings. After removing two pages that blocked access and one that timed out, I had 22 articles, most of them published by software companies.
Here’s what they had in common.
| Pattern | Articles | Share |
| Covered steps, best practices, or how-to | 22 of 22 | 100% |
| Opened with or included a definition (“What is…?”) | 16 of 22 | 73% |
| Explained the benefits or the case for it | 16 of 22 | 73% |
| Covered metrics or how to measure it | 15 of 22 | 68% |
| Covered tools or software | 13 of 22 | 59% |
| Included a heading promoting the publisher’s own product | 11 of 22 | 50% |
| Used language pointing to the publisher’s own survey or research | 3 of 22 | 14% |
Thirteen of the 22 used the same basic skeleton, with a definition, a section on its benefits, and a list of steps or best practices. Eleven added a metrics section on top. Half included a heading that pitched the publisher’s own product, which usually arrived right after the best practices.
The headings themselves often matched almost word for word. Here’s the same question, as five onboarding articles phrased it.
| Publisher | Heading |
| Dock | Why is customer onboarding important? |
| Zendesk | Why is customer onboarding important? |
| Livestorm | Why is customer onboarding so important? |
| Docebo | Why effective customer onboarding is important |
| Gainsight | Why Customer Onboarding Matters for Long-Term Success |
The definitions lined up the same way. Gainsight, Livestorm, Zendesk, and Thought Industries each had a section titled “What is customer onboarding?” For sales enablement, Celum, Mindtickle, Highspot, and the Sales Enablement Collective each asked “What is sales enablement?” None of these pages is wrong to answer those questions. But when four or five pages answer them in the same order, under the same heading, a reader has no reason to prefer one over another.
The citations repeated too. Forrester showed up in five of the articles and Gartner in four. Only three articles used phrases like “our survey” or “we found.” Two of those were citing their own industry reports, and the third was quoting a customer.
This isn’t a scientific study. I matched headings by keyword, and three searches can’t speak for every topic in B2B SaaS. But you don’t need a scientific study to recognize the pattern, because you’ve probably read all 22 of these articles already under different logos.
Where the water in the pond comes from
The articles I read were mostly competent, and some were very good. They sound alike because they were built from the same inputs, and inputs shape output more than talent does.
The same search results
The standard SEO brief starts with the top 10 results for a keyword. The writer, or the tool that builds the brief, lists the headings those pages share, notes the questions they answer, and turns that into an outline. It’s a sensible way to match search intent. It’s also a machine for producing the 11th version of the same article.
“Cover everything the top pages cover, then do it better” was the old skyscraper advice, and it assumed “better” meant longer or more thorough. When everyone follows it, every article gets a little longer while saying the same things.
The same keyword tools
Semrush, Ahrefs, and the tools built on similar data show every team the same keywords, the same difficulty scores, and the same volumes. Teams that pick topics by those numbers end up chasing the same terms in roughly the same order. That’s how three competitors in one category publish nearly identical guides within a few months of each other.
The same content scores
Content optimization tools score a draft against the pages already ranking. Used as a checklist, that’s useful for catching a missing subtopic. Used as a target, it pulls every draft toward the same terms, structure, and length as the competition, and it quietly punishes the parts of an article that the top pages don’t contain, which are usually the parts worth reading.
The same AI models
AI drafting makes the pond smaller. Models learned from the same public web, so when you ask one for an article on customer onboarding, you get the most likely version of that article, which is the average of what already exists.
The research backs this up. In a 2024 study in Science Advances, Anil Doshi and Oliver Hauser found that writers who used AI-generated ideas produced stories rated as more creative and better written, but those stories were more similar to each other than stories written without AI. The authors described it as a social dilemma, where each writer is better off while the group produces a narrower range of work. A Cornell study presented at CHI 2025 found something similar: AI writing suggestions pulled Indian participants’ writing toward an American style and stripped out their cultural details.
For a content team, that means AI can make each draft look fine while making your whole library blend into everyone else’s.
The same distance from the customer
This input does the most damage, and it’s the hardest to fix. Plenty of B2B writers, especially freelancers, never speak to a customer, a salesperson, or a product manager. They write from the brief, the search results, and whatever’s public. When those are your only sources, rewriting the pond is the only thing you can do.
The approval process often finishes the job. An opinion gets softened in review, a specific claim gets cut because legal wants to be careful, and a customer story gets removed because nobody can get sign-off in time. What’s left is safe and indistinguishable from the rest.
Why sameness costs more than it used to
Sounding like everyone else used to be a branding problem. It’s now a distribution problem as well.
Your buyers already struggle to tell vendors apart. In Wynter’s 2025 survey of 100 marketing leaders at B2B SaaS companies with $50 million or more in revenue, only 6% described their brand as very distinctive. If the people running the brands can’t see a difference, their buyers probably can’t either.
AI search raises the stakes further. When ChatGPT or Google’s AI Mode answers a question, it doesn’t need 11 versions of the same article. It needs one or two sources for each claim, and it has little reason to cite the page that repeats what every other page says.
I’ve seen this in my own AI visibility tracking. The pages of mine that get cited most often answer one specific question directly, and one AI tool’s citation linked straight to the single sentence at the top of a page that gave a clear answer. For the broad strategy questions, where I don’t have a page that says anything specific, AI tools barely mention me at all.
The honest objection: is it worth the effort?
I’ve tried to add something new to everything I write, and I’ll admit it isn’t always easy. Interviewing an expert takes scheduling, testing a product takes time and sometimes money, and original research can take days. On a tight budget, with a calendar full of deadlines, it doesn’t always feel worth it, especially when AI is going to absorb whatever you publish and serve it back to people without a click.
That objection deserves a real answer, and I think there are three.
First, AI absorbing your work isn’t the same as AI ignoring it. Answer engines cite sources, and the source they cite is the one that said something first or said it most clearly. If your article is where a statistic, a framework, or a customer example comes from, you’re the page that gets named. If your article repeats the consensus, you’re competing with every other page that does.
Second, what AI can reproduce is the pond. It can summarize the definition, the benefits, and the best practices, because those already exist in a hundred places. It can’t reproduce what your customers told you on a call, what happened when you tested a product, or what your data shows, until you publish it. That’s the part worth paying for.
Third, not every piece needs original research. The expensive inputs are the ones people picture first, but plenty of useful inputs cost very little.
A ladder of inputs, from cheapest to most expensive
When I write reviews and comparisons, I try to include SME interviews, real-life experience, or product testing. When I don’t have access to those, I draw from user-generated content, like reviews, forums, and community threads. That’s a ladder, and every rung takes you further from the pond.
| Input | What it costs | What it adds |
| User-generated content, like G2 reviews, Reddit threads, and community questions | An hour or two of reading | Real buyer language, objections, and complaints the top 10 often miss |
| Data you already have, like Search Console queries, sales call notes, and support tickets | A few hours, plus access | The questions your buyers ask in their own words |
| Your team’s experience | A short conversation | Specific examples and opinions only your company can give |
| Subject-matter expert interviews | 30–60 minutes per interview, plus scheduling | Depth, nuance, and quotes nobody else has |
| Product testing | Hours to days, sometimes a subscription | Firsthand evidence and screenshots that make claims credible |
| Original research | An afternoon for a small analysis, weeks for a survey | Data that other writers, and AI tools, cite back to you |
The analysis earlier in this article sits near the top of that ladder, and it took an afternoon, with no survey and no budget. I compared the pages that ranked for three searches and counted what they had in common. That’s often all original research needs to be.
A workable rule for a content team is that every piece gets at least one input from somewhere on that ladder. A comparison page might use G2 reviews and a sales call, a guide might use Search Console questions and a 30-minute interview, and a flagship piece might justify real testing or research. The point is that no article gets built from the search results alone.
What getting out of the pond looks like
One of my favorite articles didn’t start in a keyword tool. I was refreshing an article for Zapier about two-factor authentication when I noticed a short section that listed seven authenticator apps with almost no detail. I asked my editor whether we could turn it into its own piece, and she agreed.
That article, on the best authenticator apps, now ranks near the top of Google for its main keyword and saves Zapier more than $20,000 a year in ad spend it would otherwise need to attract the same traffic. The idea came from paying attention to something the search results hadn’t told me yet, and the article won its rankings by going deeper than a list of names. I broke down the whole process in this case study.
You can see the same thing in other writers’ work. A good example is Rosanna Campbell’s guide to writing B2B blog intros on Beam’s blog. Plenty of guides on this topic repeat the same tips about hooks, short paragraphs, and statistics. Hers stands out for four reasons.
- It follows its own advice. Instead of a definition, it opens with a University of Virginia study in which many people chose mild electric shocks over 15 minutes alone with their thoughts, then ties that to how readers feel about dull intros.
- It brings in people who do the work. Content leads from monday.com, Adyen, and Sphere explain how they write intros, in quotes you won’t find in the other search results.
- It shows instead of tells. Each step ends with a real intro from a named writer at a company like Ahrefs, Adyen, or Front, so you can see the advice working.
- It sounds like a person. The title promises an intro that “isn’t boring AF,” and the voice keeps that promise all the way through.
None of that needed a big budget. It took a few conversations, reading beyond the usual SEO sources, and a willingness to write with some personality.
The same thing applies at the level of a single paragraph. A comparison page that quotes what a customer said on a sales call, a guide that shows a screenshot from real testing, or a review that mentions a limitation nobody else admits all give a reader, and an AI tool, something they can’t find in the other 10 results.
What to change in your content process
Asking writers to “be more original” rarely helps, because the fix sits upstream of the writer, in what goes into the work.
Add one question to every brief. Ask “What do we know that the top 10 results don’t?” and don’t approve the brief until someone has answered it. The answer might be a customer quote, a product detail, a data point, or an opinion your team holds.
Budget for inputs, not just words. If you pay writers by the word or the article, you’re paying for length. Set aside time and money for interviews, access, and testing, and treat them as part of the cost of a piece.
Give writers access to the people who know things. One 30-minute call with a salesperson or customer success manager often gives a writer more usable material than a week of research online.
Use content scores as a floor. Let the tool catch missing subtopics, then stop. Never cut original material to raise a score.
Protect one opinion in every piece. Make it part of the review process that at least one clear point of view survives editing, as long as it’s accurate and defensible.
Check how much you overlap before you publish. Put your headings next to the headings of the top five results. If most of yours match, you’ve probably written another version of the same article, and it’s worth asking what you could add before it goes live.
Getting out of the pond
The cafés in Chayka’s essay weren’t bad. They were pleasant, well-designed, and interchangeable, and that was the problem. You could have been in any of them, and after a while you couldn’t remember which one you’d visited.
Your content probably isn’t bad either. Readers still expect a definition, the steps, and the basics, and there’s no need to throw those out. What they won’t remember is the version they could have read on any of the other 10 sites.
So before an article goes out, ask one question. What in this piece could only have come from us? If the answer is nothing, the article came straight from the pond, and your buyers and AI tools will treat it that way.
