Most AI marketing tools are one of two things: agents that do something behind the scenes, or reporting and audits.
OppAlerts does the actual work. It monitors news, blogs, and podcasts for your brand, your competitors, and your industry. It continually prospects for link and partnership opportunities. It helps you research, write, and optimize your content. It is a tool your team can use every single day.
OppAlerts may also be the first AI search optimization tool built for AI search rather than adapted to it. SEO optimizes for keywords. AI search runs on vector similarity and context, and that changes the work for every member of the team.
Optimizing for an LLM pulls in the whole marketing team, not just the SEO. The account manager, the copywriters, the link builders, and the PR team all need different things out of the same campaign, and OppAlerts gives each of them the tools for their part of it: over a hundred types of link and partnership opportunities watched continually, content scored against more than 25,000 taxonomy categories, the social accounts, email addresses, and phone numbers for every site it returns, and rank tracking in ChatGPT, Gemini, Google AI Mode, Google, Google’s AI Overview, and Bing.
OppAlerts comes out of a 25-year career across marketing and engineering that started in 2001: over 1,500 SEO campaigns, then more than a decade engineering everything from search engines and marketing tools to embedded systems.
A campaign holds the industry, a description of the business, the buyer personas, your owned brands, and your competitors. You get 30 of them: one per client, or one per market segment of your own business.
An AI answer changes with who is asking. A procurement manager and someone booking a vacation get two different answers, so each persona is asked separately, on every model.
Then thousands of LLM prompts and search queries run behind the scenes and build a map of your business and your industry. Everything else in the application reads that map: what to monitor, which opportunities are relevant, and what your content is scored against. That is research your team no longer does by hand.
This is where you see who the models actually name in your market, and how many of them agree. Every persona is asked by every model, many times, and the answers are pooled into one list.
Use it to get your team up to speed on a client’s market, to see where you stand against every competitor, and to find the sites the models trust that you had not considered.
The Models column matters as much as the score. A site named by all six models holds up when one model is updated. A site named by one does not.
The News Index took in over 6 million documents in August: pages, podcast episodes, and video, downloaded, scraped, and indexed so it can be searched two ways, by keyword and by vector similarity. Paste an outline of what you are writing and it returns what was published closest to it in meaning.
Two filters sit on top of that, and the second one is the part I have not seen anywhere else. PageRank is how strong a site is now. PRoT, PageRank over Time, is how long it has been strong.
That difference decides who is worth approaching. Two sites at the same PageRank today are not the same prospect if one took nine years to get there and the other took three months. The one with the history carries more trust, with search engines and with LLMs.
This is the prospecting and monitoring a link builder never has enough hours for. Six reports, all built off the campaign map, all updated every day.
You open the report and start the outreach. The finding is already done.
Sites that link to your competitors and not to you, grouped by how many of them they link to. The list of sites linking to nine of your ten is a different list from the sites linking to two, and they are never mixed.
Pages that name your brand and do not link to it. These are the fastest links you will get, because somebody already decided to write about you.
New links as they are found, so you can see what landed and what your last campaign actually earned.
Every page that names you, linked or not. A page that names you without linking still goes into a model’s context.
The same search, pointed at everyone you are up against, so a placement they won is a placement you can pitch.
Where they are being talked about, semantically matched to your market rather than to their brand name alone.
Over a hundred specific types, in fifteen groups: guest posts, resource pages, calls for papers, award submissions, event sponsorships, podcast guest slots, expert roundups, trade association directories, grants, corporate giving, newsletter ads, op-eds.
You set the parameters: the keywords it has to carry, the PageRank and the PRoT it has to clear, the links it has to have. What matches lands on your dashboard, recounted every day, so you see a guest post opening in your market the morning after it appears, not months later.
This goes far past guest posts on purpose. Awards, sponsorships, events, and research partnerships all put your brand into the text the models learn from. No traditional SEO tool watches for those.
Setting up a campaign sends the application out to find the pages that match it, and what it brings back goes into the Prospect Index. Over a million URLs are in it now, and it grows constantly.
You do not have to wait for your own campaign to fill up. The whole index is searchable from the first day, by keyword and by meaning, with the same PageRank and PRoT filters on it.
A search engine wanted the words people typed. An LLM places text somewhere in a category space, so the Content Optimizer turns your draft into a vector and compares it against 25,622 categories across 11 published taxonomies: Google Ads Topics, IAB Audience, IAB Content, Microsoft Advertising, IPTC Media Topics, NAICS, and more.
These are not their scores. There is no partnership with any of them. The taxonomies are used as a guide to how internet content is actually categorized, and the similarity score is OppAlerts’ own.
This is what optimizing content for AI search looks like for a writer. You know whether a draft is semantically aligned with your topic before you publish it, not after, and you can run your existing pages through it the same way.
Type a host and you get everything the crawl found for it: every social account and how many pages carried it, email addresses, phone numbers, contact pages, and about pages. No campaign is needed, so it works on a site you found somewhere else.
The database holds almost 40 million sites. The contact hunting that eats an afternoon of outreach is now a lookup.
Beside it, the PageRank history of that host, its www variant, and its domain, for every crawl in the database. The addresses and the numbers are blurred in the screenshot. They are not blurred in the application.
Prompts run in ChatGPT, Gemini, and Google AI Mode. Keywords run in Google, Google’s AI Overview, and Bing, with keyword research and related searches alongside them. A history page shows one prompt or one keyword over time: what each source named, and what it cited.
You choose the schedule. Every day, or only the days you want, as long as each prompt or keyword runs at least once a week.
Rank tracking is one small piece of OppAlerts. Similarly priced tools are mostly rank tracking with some analysis added. That is why this section is near the bottom of the page.
Every campaign has one. The opportunities found each day, the AI answers that named your organizations, the tracked keywords in Google’s top positions, the PageRank of every organization month by month, and the links found over time. Recounted every day.
This is how you stay on top of an account without asking anyone for a report. Open the campaign, see what is new, and go do the work that matters.
The dashboard makes it easy to get to the exact opportunities you are monitoring. Brand Link Gap, Brand Mentions, or any opportunity type: click the number and you see the results found that day.
Stays on top of the account
Research and write it
Prospecting runs for them
Watches and responds
Both kinds of ranking
This is the largest discount there will ever be.
After all five beta customers sign up, it becomes a waiting list. Whether it opens again in October is not decided.
Tell me what you work on and we will set up a call and a demo. I answer these myself.
No. One payment, and it covers you through December 31st, 2027. Nothing renews on its own, and there is nothing to cancel.
No. This beta is more than 50% off, and it is the largest discount there will ever be.
You go on the waiting list. Whether the beta opens again in October is not decided.
Yes. Fill out the form above to schedule a demo.
Two things, and they are separate. AI prompts run in ChatGPT, Gemini, and Google AI Mode. Search keywords run in Google, in Google’s AI Overview, and in Bing, so you get organic rank tracking and AI Overview citations from the same keyword. Keyword research and related searches come with it. The campaign research behind the scenes runs across several frontier models, and that list changes as the models do.
30 campaigns, and 10,000 tracked prompts or keywords a week. That counts per prompt, per source, per run. One prompt asked 5 times a week in ChatGPT is 5 of the 10,000. Add Gemini and it is 10. Keywords count the same way across the three search sources.
PageRank over Time. PageRank says how strong a site is now. PRoT says how long it has been strong. Two sites at the same PageRank today are not the same prospect if one took nine years to get there and the other took three months, and the filter is on every index search and every result card.
The API and MCP will be done by the end of October at the latest. I am aiming for the end of September.
We get on a call and we work. I help your team with AI search, the application, and your workflows, and I take your feedback and build it.
Yes. Tuesday, Wednesday, or Thursday, whichever suits your team.
Man, I’m so excited to share this. If you’re a marketer and can’t keep track of everything that’s happening across your industry, that’s a…
Change one word in a prompt, a word that has nothing to do with what you are asking, and the AI recommends a different…
You will never, ever, ever have consistent AI ranking reports. If you don’t thoroughly understand why, you don’t understand AI search. Imagine this prompt:…