Pick the three to five competitors your buyers actually put on the same shortlist as you. Copy the text of each one’s homepage, main service page, a case study or two and a page of their reviews, and paste it into ChatGPT, Claude or whichever AI tool your company already pays for. Ask it to lay out, for each company, who they say they serve, the problem they lead with, the proof they offer and the claims they repeat. Then ask the question that does the most work: which of those claims does every company on the list make? Add your own website and ask the same questions about yourself. That’s a useful first read in about fifteen minutes.
When I posted a video about this last year, I called it a starting point rather than a replacement for full competitive research, and I still see it that way. The read is quick because it only covers the public half of the picture, what competitors say about themselves. Whether it’s worth your fifteen minutes depends on the decision you take it to.
This is for whoever decides how the company presents itself, whether that’s the owner, the CEO, the marketing leader or a sales leader who keeps meeting the same names in deals.
Decide what the analysis is for before you run it
A competitor read without a question turns into a comparison of features, and a feature comparison doesn’t tell anyone what to do differently on Monday. Before you paste in a single page, write down the decision it should inform. A few that a leadership team might bring to it:
- Which buyers should we lead with on the website and in sales conversations?
- What are we claiming that everyone else claims too?
- Is a new name in our deals going after our buyers, or a different set?
- Do we need to reposition, or is the message fine and the problem somewhere else?
The question also tells you who belongs on the list. If it’s about the buyers you lead with, include whoever those buyers compare you to, even when it isn’t a firm you’d call a competitor.
Your real competitor may not be on the list
One of the more useful competitor findings I’ve been part of didn’t come from comparing competitors at all. Focus Solutions, a healthcare IT company where I served as fractional CMO, kept getting filed as another managed IT provider. The discovery work my team did found that the real competitor was the category itself. As long as buyers put Focus in the managed IT bucket, the thing that actually made it different never came up. And when Focus lost, it was often not to another firm but to the buyer’s own internal team, who assumed the inefficiency was just the cost of doing business.
Neither finding would have come out of an AI summary of five competitor websites. An AI read can only compare the companies you hand it, and it compares them on what they say. That’s still worth doing, as long as you remember the list is your guess about who you compete with, and the buyer may have a different one. Doing nothing, hiring someone, or handling it in-house all compete with you when the buyer is the one choosing.
Running the fifteen minutes
- Gather the text yourself. Homepage, the main service or product page, one or two case studies, the About page, and recent reviews wherever buyers in your category leave them. Pasting the text in controls what the tool reads and lets you check its answers against the source. A tool that browses on its own can pull an old page or the wrong one.
- Ask for a table. One row per company: who they say they serve, the problem they lead with, the proof they offer (named clients, numbers, case studies), the claims they repeat, and anything they say about price. A table makes the overlaps visible.
- Ask what everyone says. “Which of these claims appear on every site?” If every company on the shortlist promises experience and responsiveness, those words have stopped helping anyone, you included.
- Ask what nobody says. Which buyer, problem or kind of proof doesn’t appear anywhere? Treat that as a lead to check before you claim it. Sometimes nobody says it because nobody can back it up, and sometimes because buyers don’t care.
- Read the reviews for the pain. What do customers complain about across the category? That tells you what hurts buyers in your market, and being plainly not that can be a stronger position than one more claim.
- Put your own site through the same questions. Your own row may be the most uncomfortable one to read.
Then ask it to quote the line behind every finding. AI tools sound just as sure when they’re guessing as when they’re reading, and a summary that says a competitor “focuses on enterprise clients” is only useful if you can see the sentence that made it think so. Check a few of those quotes against the pages yourself. If one doesn’t exist, treat the whole summary with suspicion.
What the read can’t tell you
Everything in that table is what your competitors want buyers to believe. It doesn’t tell you what buyers actually believe, why they pick one firm over another, or whether any of the claims are true. Those answers come from your own records, and they’re worth more than the public half:
- why you won your last several deals, in the customers’ own words
- the reasons on deals you lost, and who you lost them to, including nobody
- how your salespeople describe the competition when a buyer asks
- who prospects mention on sales calls when they talk about who else they’re considering
AI helps with this half too. It can read a stack of call transcripts or win-loss notes and pull out every mention of a competitor faster than anyone on your team. Before anyone pastes customer names, deal values or call recordings into a tool, though, find out what your company allows in which tools. That call belongs to whoever owns IT and security, and it’s easier to settle before the first upload than after.
Turning it into a decision
Put the two halves side by side and the questions get concrete. If the reasons your customers give for choosing you are the same ones every competitor claims, buyers have little to separate you on except price. If customers keep naming something no competitor talks about, that may be your position. If everyone in the category leads with the same thing, there may be room to lead with something else.
That last situation came up with a specialty materials manufacturer where I advised on the strategy. Every competitor in its category led with technical authority, and the category had a look to match: dark backgrounds, circuitry, clinical photography. Leading with more authority would have made it one more name in the row. The work leaned into warmth and humanity instead, carried partly by a mascot, which nobody else in the category offered, and let the structure of the site make clear which products the company manufactures itself. There’s no sales figure to point to yet. The first sign was the client’s own team sharing the mascot on their social channels the day after launch, without anyone asking them to. That shows the company believed in the new position, which matters, but it says nothing yet about buyers.
Not every finding should change anything. A gap in the market is sometimes a gap because no buyer wants what would fill it. Before you rebuild a website around something the analysis turned up, put it in front of a few customers and your own salespeople and see whether it lands.
How often to repeat it
Competitor messaging changes faster than your positioning should. Use what you find as one input to the positioning decision. Your position is meant to hold for years, and what a competitor says this quarter won’t. Rerun the quick read when something changes: a new name keeps coming up in deals, a competitor rebrands or gets acquired, or you’re about to rework your own website. A calendar reminder with no decision attached is how a fifteen-minute read turns back into a report nobody opens.
When to get help, and when not to
You can run the fifteen-minute version yourself, and I’d encourage it. Outside help earns its fee when a real positioning decision rides on the answer and the people inside the company are too close to read it neutrally. After years with your own messaging, it’s hard to see what a stranger sees.
A competitor review is part of the brand discovery work I sell, so weigh that advice with my interest in mind. If what you need is a quick read before a pitch, do it yourself. And if buyers understand you fine and still choose someone else, the fix is probably in the offer or the sales process, and a competitor study won’t find it.
When the analysis points at a bigger question, like who you should be for or what you should stop claiming, that’s the work I do as a brand consultant. If you’d rather talk it through first, you can start a conversation here.
