Here’s a fun experiment. Open ChatGPT (or Claude, or Gemini, or Perplexity, pick your robot) and type what your ideal customer would type. Something like “What’s the best [thing you sell] for a [type of company like your customers]?”
Are you in the answer?
If you just felt your stomach drop, you’re not alone. And this matters more than you think, because your buyers are already doing exactly that. In a G2 survey of more than 1,000 B2B software buyers in August 2025, half said they now start their research in an AI chatbot instead of Google. They also said AI chat was the top source they used to build a shortlist, ahead of review sites and vendor websites.
So the shortlist is getting made before anyone visits your website. Before anyone sees your beautiful homepage. Before anyone sees your logo.
The stat that made me laugh out loud
In 2026, Semrush surveyed about 500 US professionals who use AI tools at work and asked what makes a vendor stand out when AI recommends options. The answers:
- A close match to their specific use case: 53%
- A clear, detailed description of what the vendor does: 50%
- Clear benefits or outcomes: 38%
- Brand recognition: 7%
Seven percent. I’ve been saying nobody gives a f* about your logo for years, and now the robots agree with me. What gets you picked is being clear about who you help and what you do for them. Which, funny enough, is also what got you picked before the robots showed up.
What people are actually asking AI
The same Semrush study found buyers mostly use AI to explore possible solutions, compare vendors, understand a problem, and get recommendations. And they don’t type two-word searches like they do in Google. They describe their situation: their industry, their team size, their budget, what they’ve already tried, what they hate about their current tool.
That’s the big shift. Google rewarded you for matching keywords. AI rewards you for matching situations. If your website just says “innovative solutions for modern businesses,” there’s nothing for the AI to match to anybody’s situation. You’re invisible.
How to show up in AI answers: the checklist
Some people call this GEO (generative engine optimization) or AEO (answer engine optimization). Whatever. Here’s what actually helps:
- Say exactly who you’re for. “Payroll software for restaurants with 10 to 200 employees” beats “payroll made easy.” Put it on your homepage in plain words.
- Build a page for every use case. If you serve three industries, make three pages that describe each one’s problems in their language. That’s what gets matched when a buyer describes their situation.
- Answer the real questions. Pricing (at least a range), how long setup takes, what you integrate with, who you’re not a fit for. Buyers ask AI these things. If the answer isn’t anywhere online, the AI will guess, or pick a competitor who did write it down.
- Write honest comparison pages. “Us vs. Competitor X” and “Best tools for Y” pages get pulled into answers because buyers ask for comparisons constantly. Be fair. AI and humans both smell a hit piece.
- Get talked about somewhere other than your own site. Review sites, industry publications, podcasts, Reddit and LinkedIn conversations. AI tools pull from all over, and other people vouching for you counts more than you vouching for you.
- Keep your details consistent. Same description, same positioning, same facts everywhere: your site, LinkedIn, G2, directories. Mixed signals make for vague AI answers.
- Check yourself every month. Ask a few AI tools the questions your buyers would ask. Write down who shows up. Track it like you’d track rankings.
The plot twist
Notice that nothing on that list is a trick. It’s just good marketing: know your customer, describe what you do for them clearly, and earn a reputation. The companies that will win AI search are the same ones that would have won a conversation at a trade show. The robots just made it a lot harder to fake.
So go type that question into ChatGPT. If you’re not in the answer, you know where to start.
