See if Google's AI recommends your business
Most local businesses have no idea whether AI assistants recommend them. Growl measures it.
Paste your website URL. Growl generates the questions a real customer in your category and location would actually type, puts them to Google's AI with live web search connected, and shows you the answer unedited - including the businesses it named instead of you, and the pages the AI cited as its sources.
The measurement is free. About 30 seconds, no signup and no card.
Built by a solo developer. Growl was pointed at its own website first, scored zero out of ten, and that result was published rather than hidden.
Asks Google's AI the questions your customers actually type, generated from your category and location
Shows the AI's answer word for word, including every business it named instead of you
Lists the pages the AI cited as sources, so you can see what it actually read
Checks whether your robots.txt disallows Google-Extended, which removes a site from grounded results entirely
Standard ($29/month): 4 questions instead of 2, re-measured every week, emailed only in the weeks something changed
Deep Audit ($79 one-time): 8 buyer questions, the cited pages opened and scored against your site, delivered by email
You run a local business and have never checked whether an AI assistant names you at all
You were told to optimise for AI search and want a measurement before spending money on it
You want to see which competitors an AI recommends when someone asks about your category in your area
You want to know which of your pages an AI actually reads before you write more of them
52 people supporting this launch
4.8
6 ratings
AI video with native sound, paid per clip. Nothing stored.
A calmer Mac menu bar, one click away.
Enterprise Software. Intelligent Automation. Global Impact.
Comments
23I built Growl as a marketing tool and it turned into a measurement tool because of what the measurement said. I spent about a year running automated social posting and generating more than a hundred AI-written articles. Then I read the analytics: zero sessions from social, two search visitors in thirty days. So I pointed the tool at my own website and asked Google's AI whether it would recommend a business like mine. My site scores 94 out of 100 for AI-readability. The AI named me in 0 of 10 questions. I published that result instead of hiding it, and it is what Growl now does for anyone else: paste a URL, see the AI's actual answer, and see which businesses it named instead of you. Free, no signup, no card. I have no paying customers yet. I would rather write that than invent one. If you run it on your own site, I would genuinely like to hear what came back - including if it was zero.
yoo. good design and find it useful
Showing which businesses the AI named instead of you is more actionable than a score on its own; does the answer stay stable if you rerun the same question a day later?
Late answer, but I went and measured it instead of guessing - and same day, not a day later. I asked Google's AI the same four questions twice, word for word, in two separate sessions, across 24 US cities. It named 292 businesses, and 126 of them showed up in only one of the two runs. About 4 names in 10 moved with nothing changing on the outside. Every name, both runs, city by city, is published at growl-ai.com/guides/same-question-twice so you can check it instead of taking my word. I don't know why they move - I only measured that they do.
Create a natal chart from your birth date, birthplace, and birth time. If you do not know your exact birth time, you can still continue and receive the available chart information.
creator of braincell.lol
nice, i also have a similar product in the making
Founder, Builder, CEO
I think this makes it very much easier to get especially chatgpt to notice the product!
Branded QR codes that don't look like a barcode
Publishing the 0/10 on your own site instead of quietly fixing it first and leading with a better number is the part that actually builds trust here — most people would've held that until it looked good. The robots.txt / Google-Extended check is the detail I wouldn't have thought to ask about, but it seems like exactly the kind of thing that silently tanks a site's AI visibility without anyone noticing — is that turning out to be a common culprit when businesses score low, or is it usually more about the site just not answering the question an AI would look for?
Building PZERO, saving you money - pzero.studio
Checking whether Google AI already surfaces your business is a useful early signal before you pour more into content. If you want to spend less on AI at work, take a look at pzero.studio.
Ho
Good idea and design was good and no Break nothing
The readability-versus-recommendation split is the useful part here. Scoring 94 for AI-readability and 0 for being named at all is the exact gap most "is my site AI-friendly" checkers miss, because they grade the page in isolation instead of putting it inside a real competitive answer. Showing the businesses the AI named instead, plus the sources it cited, turns a vague worry into a concrete list of who to displace and where. One caveat I would want addressed: these answers are non-deterministic, so a single 0 out of 10 can be run variance as much as signal. Do you run each question once, or sample it a few times so the score reflects a trend rather than one draw?
You were right to push on this, and the honest answer at the time was: once. So I ran it twice. Same four questions word for word, two separate sessions on the same day, 24 US cities - 292 businesses named, 126 of them in only one of the two runs, about 4.3 in ten. So a single 0 out of 10 is partly variance, and that is an uncomfortable thing for me to publish, because I sell the tool that takes the reading. Full table, every name, both runs: growl-ai.com/guides/same-question-twice. While recounting it I found three of my own rows were not businesses at all - headings the model wrote that I had recorded as names. They are still in the table, marked, with the strict recount beside them: 121 of 288.
This is such a smart pivot! With AI search changing the landscape so rapidly, local businesses are flying blind right now. The fact that Growl shows you exactly what the AI answers is incredibly valuable.
This is such a smart pivot! With AI search changing the landscape so rapidly, local businesses are flying blind right now. The fact that Growl shows you exactly what the AI answers is incredibly valuable.
This is such a smart pivot! With AI search changing the landscape so rapidly, local businesses are flying blind right now. The fact that Growl shows you exactly what the AI answers is incredibly valuable.
This is such a smart pivot! With AI search changing the landscape so rapidly, local businesses are flying blind right now. The fact that Growl shows you exactly what the AI answers is incredibly valuable.
Growl offers a refreshingly honest tool for local businesses navigating the new frontier of AI search. Its free, no-signup visibility test is a brilliant starting point, showing you exactly what Google's AI says about your business. The platform's transparency is its strongest asset, openly sharing its own study and admitting its paid plans don't guarantee AI recommendations. Instead, it provides actionable marketing steps and clear, simple pricing without the agency fluff. For solo owners who lack a marketing team, Growl is a practical, no-nonsense solution to understand and experiment with your AI presence.
Building ClipWise to help creators and teams edit, analyze, and repurpose video content faster.
Ran my website through Growl and got some fantastic actionable insights. Showing the actual unedited AI output and cited pages makes optimizing for AI search so much easier!
Thank you for actually running it - you are the only person in this thread who did, and that is worth more to me than the upvotes. Two things I would like to know if you are willing: which businesses it named instead of you, and whether that list looked like your real competitors or like strangers. And one caveat I owe you: since your run I measured how much a single reading moves. Same four questions, word for word, twice on the same day, 24 US cities - 126 of the 292 businesses named came back in only one of the two runs. So please treat one run as a rough reading rather than a verdict. The full table is at growl-ai.com/guides/same-question-twice.
LetsGetSurfaced is an AI visibility and SEO platform that shows you exactly what AI assistants say about your brand and what to fix this week.
The Google-Extended robots.txt check is a smart addition. Have you thought about extending the same live-query approach beyond Google's AI? We've found that ChatGPT, Claude and Perplexity often name completely different businesses for the same query, since each pulls from different sources and weighs citations differently - a business could score zero on Growl but still get mentioned regularly on Perplexity, or vice versa. Curious whether that cross-model gap shows up in your data too.
Honest answer: I can't tell you, because I have only ever asked one engine. Everything I publish is Google's AI (Gemini) with grounding on - not ChatGPT, not Claude, not Perplexity, not AI Overviews. I would rather measure one engine and say so than claim four and sample none, so the cross-model gap you describe is simply not in my data. What I did measure is how much a single engine moves on its own: the same four questions, word for word, twice on the same day, across 24 US cities - 292 businesses named, 126 of them in only one of the two runs. If one engine drifts that much between two runs on one day, I would expect your cross-model gap to be larger still - but that is a guess, and I am not going to publish it as a finding. Data: growl-ai.com/guides/same-question-twice. If you have run the same question across models, I would genuinely like to see your numbers.
It would be good to know if the questions made to the AI also have the language in question, since that can change the outcome of the results, great product nonetheless
I think this feature could be accomplished with Codex or Claude, and it doesn't seem like a website is needed for detection. I'm not sure if my understanding is correct