Google AI Overviews appeared for 100% of the software “alternatives” and “best X for Y” searches in a Semrush data sample.
That’s an important change for the kinds of searches B2B content teams have spent years trying to rank for.
But instead of how often AI Overviews appear, I think there’s a more important point to focus on: when someone compares your brand with a competitor, does your content give enough for them to make a decision?
As AI search is cementing itself as a part of that comparison process, what information are AI assistants finding and using to describe your brand?
AI Overviews are taking over software comparison searches
In a Semrush study, they tracked 400 commercial keywords in software, banking, and ecommerce from September 2025 to August 2026.
By the end of the study, every software “alternatives” and “best X for Y” keyword in its sample triggered an AI Overview. Software review queries reached 75%.
For the wider set of industries, AI Overviews appeared on nearly 90% of review-related searches, up from around 30% at the beginning of the study.
One interesting example from the Semrush study was the search for “Mailchimp vs Klaviyo.”
Google's AI Overview cited Klaviyo's own comparison guide when explaining why someone might choose Klaviyo. But for Mailchimp, it relied on a third-party article from Zapier.
Two brands appearing in the same comparison, with different websites supplying the information behind Google's recommendations.
The question marketers should be asking here is:
Who's providing the information AI search uses to compare your brand against competitors?
For your important comparison and alternatives keywords, examine which product differences Google highlights, which websites it cites, and whether those sources accurately represent your strengths and limitations.
You might rank well organically while a third-party website provides the evidence behind Google's AI comparison.
Understanding that difference could show content gaps your existing reports won't show.
AI search is becoming part of B2B vendor discovery
Semrush separately surveyed B2B professionals and found that, among 519 respondents who used AI for work, 62% used it during active vendor comparisons.
When an AI assistant recommended a vendor, 71% said they visited its website, 46% compared it with alternatives, and 38% checked software review platforms.
So appearing in an AI answer might get a brand considered. But the buyer still has questions to answer, particularly about fit, limitations, and whether the product can do what the recommendation suggests.
In the same survey, 53% said a vendor stood out in an AI answer when it closely matched their specific use case. Half valued a clear, detailed description.
That also gives content marketers something to think about.
AI assistants may help buyers discover and compare vendors, but what happens when they visit those vendors' websites?
Can they identify differences between the products, or are they reading similar info about productivity, efficiency, and saving time?
New research from Wynter suggests that distinguishing between competing B2B brands is harder than it should be.
Generic B2B content creates a problem with differentiation
Wynter surveyed 100 senior B2B SaaS marketers who had evaluated competing vendors.
64% said it was difficult to distinguish vendors from their websites and messaging.
Wynter then removed the brand names from five real CRM value propositions and asked the same group to match them to Salesforce, HubSpot, Zoho, Pipedrive, and Oracle.
They averaged 1.86 correct answers out of five. Random guessing would average one.
These were experienced marketers evaluating familiar software companies, and even they struggled to identify whose messaging was whose.
It can be difficult - visit a few B2B software websites, and you'll often find the same promises about saving time, improving productivity, and making things easier.
Those benefits might all be true. They still don't tell someone what makes one product more suitable than another.
Now consider AI summarizing several vendors that describe themselves in almost identical terms. There may be genuine differences in the products, but the marketing copy gives the assistant, and the buyer, very little concrete information to compare.
There's an SEO and GEO implication here too.
If your comparison pages contain the same generic claims as competing websites, you're providing little distinctive information for AI assistants to reference when answering specific questions about your products.
What should content marketers optimize for SEO and GEO?
I'd pay more attention to existing comparison, alternatives, and product pages before producing more.
A strong B2B comparison should explain:
Who each product suits. Be specific about company size, requirements, and use cases rather than claiming the product works for everyone.
What it can and can't do. Name feature limitations, plan restrictions, integrations, migration requirements, and other differences that can change a buying decision.
What the claims are based on. Link to current documentation, show relevant product evidence, and separate verified details from opinion.
What implementation involves. Explain onboarding, security requirements, likely costs, and dependencies where you have reliable information.
Make those differences easy to identify in the content too. Descriptive headings and clear comparisons can help readers scan the page, but they need to be supported by accurate information.
For AI search, I'd also review whether important product details appear consistently in your documentation, pages, and third-party sources. Conflicting or outdated information makes it harder to understand which claims are reliable.
Take software integrations as an example.
A comparison page claiming your product has ‘better integrations’ can be thin.
But explaining which integrations are native, which require third-party tools, what limitations exist, and which pricing plans include them gives buyers information they can use to compare products.
Those details also give AI search systems specific, verifiable information to reference when answering questions about how competing products differ.
The same applies to pricing, implementation requirements, feature limitations, and supported use cases. These claims should be backed by documentation rather than generic marketing descriptions.
Audit your comparison content for AI search
Start with 3 to 5 important comparison or alternatives pages.
First, review the search terms those pages target, particularly software alternatives, product comparisons, and use-case-specific questions.
Then use consistent AI search tracking to audit how those topics appear in Google AI Overviews and other relevant AI search platforms.
Look at:
AI citations: Which pages are being referenced? Is your content appearing, or are AI assistants relying on competitor websites and third-party reviews?
Brand descriptions: When your brand is mentioned, does the answer accurately explain its strengths, limitations, and intended users?
Missing information: Are competing products being compared using details that your own content doesn't address?
Content quality: Are your comparisons based on verified product differences, or are they repeating generic copy from company homepages?
Use your own product knowledge, documentation, sales questions, and customer feedback to find and target gaps.
You can use AI to help identify weak claims and missing details, but the evidence still needs to come from reliable sources.
I'd also monitor brand mentions and cited URLs separately, alongside organic search performance. Improving a comparison page doesn't guarantee an AI citation, and tracking both helps you understand whether visibility is changing.
If AI search is becoming a bigger part of how software gets compared, giving it accurate, distinctive information about your brand should be part of your content strategy.