Built an AI-visibility tool that also gets you cited, not just tracks it
More buyers, not just in software, ask ChatGPT or Perplexity before they ever type a company name into Google. If your product or service isn't in that answer, you don't just miss a click, you miss a decision you'll never even know you lost, because nobody tells you "I asked AI and you weren't mentioned."
I tested this directly instead of assuming it. Checked 30 real companies (SaaS tools, but also services like background checks, hosting, payroll) for a live Reddit thread in their exact category, someone asking "what should I use for X." Found one in nearly every case. In the large majority, the company itself had never replied, while competitors got named freely by other users in the same thread. That's not just a lost Reddit reader. It's exactly the kind of source material ChatGPT and Perplexity increasingly pull from when they answer the same question for the next person.
Run real buyer questions against live models and see plainly whether you're mentioned, where you rank against competitors, and how that shifts over time. Pair that with watching Reddit for new mentions of your brand as they happen, automatically, not something you have to remember to check. That's the AEOrank side most people expect.
But tracking alone just tells you you're losing. It doesn't fix anything. The part that actually moves the number is finding the specific live threads where your category is being discussed right now, verifying they're genuinely still open (not already answered, not removed by mods), and getting a real reply into that exact conversation while it's still live, since that's the material the next AI answer gets built from.
If you want to see what this actually finds for your own Business, go check it out yourself at aeorank.tech.
Curious how other companies here are actually measuring this today: tracking AI mentions at all, or mostly flying blind on it?
I believe social media gives us a false sense of security. There are so many posts and videos about people who created an app and now make 80K a month, but in reality, most of it is just clickbait. The SaaS field feels overbuilt. In document tools especially, which NexusDocs focuses on. The first objection is always the same: I can paste this into ChatGPT. That objection is fair. A generic summary is not a company or a proper response. What saturation hides is that most of those tools do the same job: explain the text. They do not own a moment. They do not change a decision. They do not get a second document without a reminder. AI lowered the cost of version one. It raised the number of lookalike products. The scarce things are still the same: I’m building through that, not around it. Shipping was the easy week. Getting one person to use it is the actual business.
The part about connecting AI visibility to actual business outcomes is what interests me most. Being mentioned or cited is useful, but the harder question is what happens after that visibility.
I’d be interested in seeing this measured as a funnel: AI mention → click/visit → qualified action → conversion. Otherwise, AI visibility could easily become another metric that looks impressive without showing whether it actually influences revenue.
The live Reddit-thread angle is particularly interesting because it gives companies something actionable to improve rather than just another visibility score.
The Reddit finding matches what I see when I check why one company gets cited over another: the engines lean on threads and third-party pages because that's where claims look independent. A company's own site says whatever it wants; a thread where three users name the same tool unprompted reads as verified. The part most people skip is boring: get specific, checkable claims (numbers, comparisons, real customer outcomes) onto pages you don't own. One caveat from the testing side - what ChatGPT cites and what Perplexity cites overlap less than you'd expect, so a thread that wins one engine can be invisible to the other. Worth running the same buyer question through both before deciding where to spend effort.
The measurement half is solid. The reply half is where this gets people banned, because mods remove vendor answers fast and a shadowbanned domain is a worse outcome than being unmentioned, so the feature that matters is not whether the thread is open but whether that subreddit tolerates a vendor reply and whether the account is old enough to survive one. The citations that hold up over time come from original data nobody else has, since models keep quoting the source of a number long after a thread has scrolled away.
This is fascinating — and slightly terrifying for someone building in the exact opposite direction.I built an offline-first AI assistant. Nothing ever leaves the device. Zero API calls. Zero cloud.Which means: I will never appear in ChatGPT or Perplexity answers. Not because my product is bad, but because by design, there is nothing to scrape, nothing to index, nothing to cite.Your tool made me realize: I'm not just invisible to traditional SEO. I'm invisible to the next SEO. The one where AI answers replace Google results entirely.The irony? The whole selling point is "your data stays private." But "my product stays private" too — and not in a good way for discovery.Have you seen any offline-first or privacy-first products successfully break into AI-generated answers? Or is this a structural blind spot we'll have to solve differently?
Congratulations on the launch! One thing I noticed is that your homepage explains the features before the problem you’re solving. Many first-time visitors decide within a few seconds whether to stay. Leading with the customer’s pain point could increase engagement. Happy to share a few more ideas if that’s useful.
I also develop the free security first pdf tool name aeropdf .app but the issue is even after one year the traffic is very low and also the search engines give very low ranking.
I’m also planning to build a SaaS product after optimizing the project I’m currently working on.
Adjacent disclosure, I work on an SEO tool, so read this as someone in the same neighbourhood rather than a neutral party.
Building on ohad1976's removal point, because the mechanism is worse than it looks and I only know that because we walked straight into it.
Reddit filters at the account level, site wide, on something called Contributor Quality Score. Individual subreddits set automod to auto reject anyone below a threshold, and you are never shown your score anywhere in the interface. A newish account replying in a category thread is close to the worst possible risk profile for that filter.
The part that makes it a product problem rather than a customer problem: a removed comment looks completely normal to the person who wrote it. It sits there on the thread when they view their own profile. The only signal is an automod reply, and that lands under Notifications rather than the inbox, so if you are not specifically looking for it you will not find it. We lost two days before noticing that every comment we had left in one subreddit had been removed within seconds of posting.
So a tool reporting replied in 30 relevant threads can be reporting 30 while the number visible to an actual human is zero, and neither you nor your customer would know. To answer ohad's question directly, I would not just track removals, I would make survival the metric. Check each reply from logged out a few minutes after posting and report survived rather than posted. It is cheap, and nobody in this category does it.
One other thing worth separating in the pitch: whether a Reddit mention feeds a citation depends on which mode the assistant is in. Browsing assistants do live retrieval against an index, so a fresh thread can surface quickly. Non browsing answers come out of training data and move at the next model cut, months away. Both are real, they run on completely different clocks, and customers will judge you on the fast one while part of what you are selling sits on the slow one.
Have you seen the removal problem in customer accounts yet, or has it not surfaced precisely because nobody can see it?
One thing I’d separate is “am I mentioned?” from “am I mentioned in the prompts a buyer would actually ask right before choosing?” A tiny test: keep 20 fixed prompts split into compare / alternative / pricing / use-case buckets, run them weekly, and only count a win if the answer gives a real reason to pick you. Otherwise the visibility score can move up while the buyer still has no reason to click.
Does this mean we prepare a fixed set of questions and publish them once a week? And are we republishing the same questions each week?
Spot on regarding the LLM citation feedback loop. How do you distinguish between high-intent recommendation threads versus broad, casual mentions when scoring visibility?
Great breakdown, super interesting approach to AI visibility. Good luck with the growth!
Interesting, good article.
The line between tracking and getting cited is the real insight here, most AEO tools stop at the dashboard. But running aged accounts with fake personas across Reddit threads is the same failure mode I've seen on other platforms: once the network gets flagged, every account in it goes down together, not just the newest one. I'd bet on training the actual team to answer three threads a week in their own voice, slower, but nothing there for a platform to detect.
Interesting approach. I’m currently developing a Morocco travel project for the U.S. market, so I’m especially interested in how AI visibility can help small travel businesses get discovered. Thanks for sharing.
Spot on regarding Reddit being the backbone of modern RAG pipelines. Ever since OpenAI and Google inked data licensing deals with Reddit, brand presence in fresh threads has become prime real estate for LLM retrieval. Connecting passive AEO tracking to real-time Reddit engagement actually addresses the root cause of why brands get omitted. Curious—how are you handling thread freshness vs. sub-reddit self-promotion rules when alerting founders to jump in
Really interesting approach. I especially like the idea of moving beyond simply tracking AI visibility and actually working on getting businesses mentioned in relevant conversations.
I’m also working on an online home improvement and interior design platform, Interior Glamour, where we’re building useful, topic-focused content around areas like flooring, lighting, home decor, and interior design. It’s interesting to think about how AI visibility and citations can help this kind of niche content reach the right audience.
The “gets you cited” part is more interesting than simply tracking visibility. How are you measuring whether a citation is actually useful versus just appearing in the answer?