What the engines say
Six engines. We set the cadence. Position, share of voice and sentiment per category, week after week.
Whether AI names you, and who it names instead.
The AI visibility loop
Find Where AI Does Not Recommend You.
Then Research the Topic Properly.
- ChatGPT
- Gemini
- Google AI Overviews
- Google AI Mode
- Grok
- Perplexity
- Claude
An AI Answer Now Sits Between You and Your Buyer
"Best CRM for a twelve-person agency." "Which supplier makes corrosion-resistant valves for saltwater?" "Good noise-cancelling headphones for an open-plan office."
Ask any of these and you get a short list — usually three or four names, with reasons. Your buyer reads that list before they ever reach a search results page. You are on it or you are not, and nothing in your analytics tells you which.
Different Signals
Backlinks and domain authority do not decide what a model says. It answers from what it was trained on and what it retrieves at the moment it is asked.
Different Rules
Ranking first on Google does not put you in a ChatGPT answer. The two are built from different inputs, and one does not predict the other.
No Instrument
Your analytics can tell you a visit came from an AI assistant. It cannot tell you what the assistant said about you, or which competitor it named instead.
Every run is stamped with the model set that produced it, and the timeline flags the date that set changed — so you are never shown our own instrument drift as your progress.
What BrandKarma Does
We measure how the engines talk about you. We research one topic. We publish a small cluster. We measure again.
Monitoring is the instrument. The researched topical cluster is the intervention. Search Console and site traffic are how you know it worked. We will not help you flood.
How It Works
- 01Measure
Six engines, Search Console and site traffic. We set the cadence.
- 02Research
One topical cluster: a dossier of cited claims, not a keyword list.
- 03Write
A pillar and three to five pieces you would actually publish.
Then we measure again — same instrument, not a fourth step. Published pieces are pulled from one API. We will not help you flood.
Why they belong together
Each layer answers one question
Measure what the engines say, what the web said, and whether anyone arrived. Read those together. Then research one topic and publish a few pieces from that dossier. BrandKarma runs that on one cadence.
Measure
What the open web said
Classified mentions: tone, claims, a key quote and a suggested action. Not a raw dump.
We read each match. A suggested action, not a feed.
Whether Google sent anyone
Search Console. Sixteen months of daily site history, sitting beside the AI number.
How you know classic search moved.
Whether anyone arrived
First-party clickstream: sessions, pages and sources. No analytics cookies, so it adds no consent banner of its own.
How you know a visit actually happened.
Join
The digest that binds them
A written digest on a schedule, across engines, mentions, Search Console and traffic. One shutter, one next topic.
The coordination. You do not run the parts as separate jobs.
Write
Research, then a few cited pieces
One topical cluster: a dossier of cited claims, then a pillar and three to five pieces from that dossier.
Research spends a cluster. Writing from that dossier is the same cluster.
Then we measure again — same instrument, not a fourth step.
The researched cluster
Two topical clusters a month, done properly
Monitoring is the instrument. The researched topical cluster is the intervention. Search Console and site traffic are how you know it worked.
A dossier, not a brief
One topic, researched once. Cited claims, not a prompt stuffed with keywords. The facts that will go into the pieces are gathered before a word is drafted.
A pillar and a few supports
From that dossier: one pillar page and three to five supporting articles. Enough to cover the topic. Not enough to teach the models you are noise.
Thin ideas do not ship
Ideation rejects thin pages, duplicates, and “Complete Guide to X.” That gate is real. We will not help you publish thirty uncited posts a month.
What You Get
AI Visibility
Track What the Engines Say, Week After Week
Six engines included. We set a weekly LLM cadence. Google AI surfaces are checked daily. You get your average position, share of voice and sentiment per category, plotted over time and broken down by engine.
- Your average position and mention count per category, tracked over time
- Every brand the models name alongside you — ranked by mention count
- Sentiment and mention-rank distributions for the latest run
- Filter by engine, by category and by prompt type
- ChatGPT, Gemini, Google AI Overviews, Google AI Mode, Grok, Perplexity and Claude
Keyword Intelligence
See Your Mention Rate and Your Google Rank on One Card
For every keyword you track, BrandKarma shows how often you and each competitor you track are mentioned in AI answers, which domains the models cite, and where you sit in classic Google results — with difficulty, cost per click and volume alongside. One card, both channels.
- Mention rate for you and each tracked competitor, with the change since last run
- The domains models actually cite for that keyword, with counts
- Keyword difficulty, cost per click and monthly search volume
- Your current Google position and clicks, from Search Console
The join
What the Models Say, and Whether Anyone Arrived
Most AI visibility tools stop at the answer. BrandKarma puts the answer next to the outcome: AI mentions, first-party site traffic and classified web mentions on one dashboard, plus a written digest that reads those three against your Search Console history.
- Google Search Console — sixteen months of daily site history, pulled in the background when you link a property
- First-party web analytics — cookieless, so it writes nothing to the visitor's device and adds no consent banner
- Classified web mentions: tone, page kind, claims, a key quote and a suggested action
- A written digest that reads across all four sources, on a schedule
Two Things That Are Not a Dashboard
Run it from inside Claude
BrandKarma ships an MCP server. Point Claude or any MCP client at it and you can read your analytics, manage monitored prompts, edit brand configuration and research keywords — in conversation, without opening the app. How to run the loop from Claude.
A headless content API
Published content is available over a read API: fetch by brand, language, category, tag or slug, and render it however you like. Every article on this site’s blog is served through it. Read the API docs.
What We Can Show You Today
We have no named customer case study yet, and we are not going to invent one. Here is the instrument instead.
Six engines
Included. LLM analysis weekly. Google AI surfaces daily. Daily LLM is Enterprise.
16 months
Of Search Console site history, pulled in the background when you link a property
Every post
On this blog was written by BrandKarma and delivered through its own public API
Brand $99 · Company $199
Two researched clusters, or five. See pricing.
See What AI Says About You
Request a free AI Visibility Report. We run your top categories across ChatGPT, Perplexity and Gemini, show you who gets named instead of you, and send it within 48 hours. It is not a full cluster.
Request Your Free AI Visibility Report