When AI answers your buyers, be the brand it names.
ChatGPT, Gemini and Perplexity are already answering your category's questions — with or without you. Aeloria measures exactly how you show up, hands your team a ranked list of fixes, and re-runs the same prompts to show what moved. Knowing where you stand is where this starts. Standing in a better place — month after month, on purpose rather than by luck — is what we’re built for.
Trusted by the brands AI talks about — and the agencies behind them
Global leader in yachting 
Leader in textile automation 
Leader in construction solutions 
What is Aeloria?
Aeloria is an AI reputation training system, built in Zürich. Nobody outside a frontier lab trains the model itself — but every answer an AI gives about you is assembled from evidence it can find, and that evidence can be measured, corrected and measured again. That loop is the training. GEO and AEO are two tactics inside it; reputation is the whole surface. Aeloria measures how AI assistants — ChatGPT, Gemini, Perplexity and more — describe, judge, cite and recommend your brand, converts every gap into a ranked, effort-scored fix list, and re-measures the identical prompt set after each fix ships. One composite Brand AURA score (0–100) rolls presence, sentiment, narrative, competitiveness and website readiness into a single number a board can track. Most tools stop at watching. A training system closes the loop.
Your SEO stack can't see where buying decisions moved.
AI answers are a different battlefield with different rules — and the evidence says the ground is being claimed right now.
of AI-cited links appear in Google's top 10
Across ChatGPT-class engines, roughly 88% of the links AI actually cites rank nowhere in Google's top ten for the same query. Rank tracking alone is blind to the answer layer.
Source: Ahrefs, 15,000-query cross-engine study, August 2025of AI category leaders keep the lead next month
Once a brand owns an AI answer category, it retains first place in 9 out of 10 month-over-month comparisons. In AI answers, incumbency compounds.
Source: Semrush, AI visibility is a topic-level game — 1,094 categories, July 2026of AI answer demand sits in categories with no clear owner
Only 15.2% of categories have a clear owner — and the highest-demand ones are the least settled. The brands that build AI visibility now inherit the compounding advantage; everyone else negotiates with the winner later.
Source: Semrush, AI visibility is a topic-level game — 1,094 categories, July 2026Every tool in this category can chart where you appear.
Aeloria is built around what happens next.
A closed loop from detection to proof — so your Monday morning starts with a ranked to-do list, not a dashboard to interpret.
Detect
Discovery Trails run the questions your buyers actually ask — across engines, across languages — and record every answer where someone else is cited instead of you.
Diagnose
Aeloria reads the pages the engine did cite and identifies what it lifted: the definition, the attributed statistic, the parseable structure you're missing.
Prioritise
Each finding becomes a specific action, ranked by expected impact against effort. A writer knows exactly what to ship — not forty generic best practices.
Re-measure
After you ship, the same prompts re-run on schedule — same set, same languages — so movement is measured, not assumed.
Your whole reputation. Not just your visibility.
Virtually every app in this market counts mentions and stops. But being mentioned is not the same as being chosen. What AI says about you, the tone it says it in, the story it builds, where you stand against rivals, and whether your own site feeds it the right material — Aeloria measures and trains all of it, in one system, from a board-ready number to the exact sentence an AI engine lifted from a competitor's page.
Brand AURA score
One composite 0–100 score — presence, sentiment, narrative, competitiveness, website readiness — tracked over time, per engine, per market. The number your board asks for.
Discovery Trails
Structured buyer-question protocols run on schedule across engines and languages — measuring leadership, share of mind and mentions per trail, not vanity prompt counts.
Fix Engine
Every gap becomes a ranked, effort-scored action with quoted evidence — and re-measurement of the identical prompt set after you ship.
Reputation Signals
A living graph of every theme AI associates with your brand — signal by signal, engine by engine — so you see the story AI tells before your customers do.
Competitive arena
Designated rivals tracked on the same trails: presence, sentiment, AURA deltas. Know exactly who's eating your citations and where.
Sentiment with receipts
Every sentiment signal is backed by quoted evidence you can open and read — no black-box scores you have to defend on faith.
See the story AI tells about you — node by node.
Every theme, claim and risk AI engines attach to your brand, mapped as a live graph with mention counts and sentiment. Filter by engine to see how ChatGPT's version of you differs from Gemini's.
- Positive, risk and negative signals separated per aspect
- Drill from any node to the underlying quoted answers
- Spot narrative drift before it hardens into "common knowledge"
Your site, graded the way an AI crawler experiences it.
Seven dimensions — accessibility, indexability, credibility, identification, scannability, wayfinding, consistency — scored page by page, with critical fixes ranked and explained in plain language.
- Finds the key pages AI can't reach or can't parse
- Flags entity inconsistencies that fragment your brand identity
- Every finding ships with a "see the fix" action, not just a grade
Measurement is where the others stop.
It's where Aeloria starts.
The tools we're most often evaluated against are visibility trackers — strong at counting mentions. Aeloria is a training system that acts on what it finds. Compared on public pricing and published documentation.
| Capability | Aeloria | Profound | Peec AI | Otterly AI |
|---|---|---|---|---|
| One composite brand score (board-ready) | ✓ Brand AURA | — | — | ◐ visibility index |
| Ranked fix list, scored impact vs. effort | ✓ | ◐ enterprise tiers | — | ◐ audit checklists |
| Forever-free plan to start | ✓ free tier | — paid from $99, annual | — 7-day trial | — 7-day trial |
| AI-readiness audit of your own site | ✓ Website Health | ◐ | — | ✓ URL audits |
| Multilingual measurement as a first-class feature | ✓ any language | ◐ | ◐ tier-gated | ◐ |
| White-label agency reporting | ✓ native +€99/mo | — | ◐ via Looker | ◐ via Looker |
| How AI answers are collected | ✓ licensed provider APIs | ◐ direct interface monitoring | ◐ | ◐ |
Based on public pricing pages, documentation and verified user reviews, July 2026. ✓ included · ◐ partial or tier-gated · — not offered or not confirmed. Coverage and update frequency scale by plan on all four platforms. Prompts run in any language you sell in; the Aeloria interface ships in six.
On collection method the difference is deliberate: some AI surfaces publish no official API, so covering them means automated sessions on consumer interfaces. We stay with licensed provider APIs — a slower-growing engine list, and a data history that stays yours.
Run an entire client portfolio from one AI-visibility cockpit.
Every client gets a Brand AURA score, a fix pipeline and a re-measurement schedule. You get one portfolio view, optional white-label reporting (+€99/mo flat, any number of brands), prospect audit tools for new business — and a recurring service line your clients can't get from their SEO retainer. From €599/mo with three brands included. Agency partners report returning more than 3× the platform cost within their first months — on retainers they price themselves — and not one of them has left us since.
- Portfolio dashboard: every brand, every score, every delta
- White-label client reports — your logo, your story, on schedule +€99/mo flat
- Sell outcomes: "we shipped nine fixes, you're now cited in six more buyer questions"
GEO advice is full of superstition.
We only ship what survives the evidence.
We track the peer-reviewed research and controlled experiments in this field — and grade every tactic before it enters our recommendation engine. Most vendors won't tell you this. We lead with it.
What the evidence supports — so we weight it
- Informational density: quotable definitions, attributed statistics, structured comparisons and worked examples — the content shapes independent studies consistently find AI engines lift.
- Retrieval & crawler access: making sure answer-time bots can reach, parse and extract your key pages — the layer most audits skip entirely.
- Category ownership: longitudinal data shows AI category leadership persists month over month once won — visibility compounds.
- Continuous re-measurement: peer-reviewed benchmarks show one-off content tricks erode as competitors adopt them. The loop is the moat, not the trick.
What we won't sell you
- "Add this magic file and AI will cite you" — controlled studies and the engines' own documentation say otherwise.
- Guaranteed citation uplifts quoted as universal percentages — the underlying studies don't support that promise, and we won't make it.
- Keyword stuffing and format tricks — measured at best as noise, at worst as actively harmful in modern LLM search pipelines.
- Black-box scores you can't defend — every Aeloria signal opens to its quoted evidence.
AI answers can't be gamed reliably — but they can be measured precisely and earned systematically. Aeloria exists to make sure your budget flows only into the levers with evidence behind them, and to prove it moved something.
Privacy-first. Because scraped data can be taken back.
The standard method in this category is to run automated sessions against consumer chat interfaces — the larger tools advertise it as an advantage. The providers' own terms are less enthusiastic. OpenAI's forbid any attempt to “automatically or programmatically extract data or Output,” or to “circumvent any rate limits or restrictions or bypass any protective measures.” Google's and Anthropic's say the same thing in their own words. It works until it doesn't.
How it ends
- A competitive-analytics company scraped a major platform's public profiles and argued for five years that public data was fair game.
- It never needed to lose in court to lose everything. The platform cut off access and the company shut down in 2018 — four years before the case was decided.
- The court then found breach of the platform's user agreement, plus unauthorised access where fake accounts had been used to reach protected pages.
- The consent judgment: $500,000, and destruction of all scraped data together with the source code and algorithms built from it.
hiQ Labs v. LinkedIn, N.D. Cal. — consent judgment, December 2022.
The Aeloria way
- We never scrape AI chats. Every answer we measure enters Aeloria through official provider APIs, under commercial terms we actually agreed to — and we'll confirm that in writing for your procurement file. (Website Health fetches the pages you point us at, as any browser would.)
- Nothing we've collected for you can be clawed back, because none of it was taken. Your baselines, your trend lines, your entire history of how visibility changed: yours to keep.
- It is also the more rigorous measurement. Same prompt, same model, same conditions on every run — so when your score moves after a fix ships, you know it was the fix, and not somebody's session history, location or interface test.
- Swiss-built, privacy-first by design — data posture as a feature, not a footnote.
No AI provider has shut off a GEO tool yet — and “yet” is doing a great deal of work in that sentence. The terms are already written, the precedent is already set, and the decision belongs to someone else, on their timeline, with no notice to you. It goes wrong in one of two ways. Access is withdrawn, the measurement stops, and your trend line breaks off mid-story. Or a court gets involved, and the history is destroyed along with the code built on it. Neither can happen to data that arrived under a contract. Before you trust anyone with years of your reputation history, ask them where their numbers come from.
Data, hosting and security — the detail your procurement team will ask for
Application, storage and compute on Google Cloud, in European regions. Aeloria is Swiss, based in Rüschlikon; Switzerland holds an EU adequacy decision, so data moves between the two without a workaround.
Aligned to the GDPR and the Swiss revDSG. You are the controller for the brand data you configure; Aeloria is the processor. Transfers outside the EEA run on Standard Contractual Clauses, with the EU–US Data Privacy Framework where it applies.
Every processor is listed by name and purpose in the DPA — hosting, billing, email, cookieless analytics and each AI provider we query. No unnamed “trusted partners”.
Workspace content is kept for the subscription plus 30 days, with export tools open across that window; backups purge within 90 days; logs at 12 months. Billing records are held 10 years because Swiss commercial law requires it.
TLS 1.2+ in transit, AES-256 at rest. Salted password hashes, role-based access on least privilege, logging of administrative access to Customer Data, and dependency vulnerability scanning.
One documented answer a year. The DPA gives you Annex II — the security measures as actually implemented — plus our hosting provider's documentation, and commits us to answering a written security questionnaire once every twelve months. On-site inspection where mandatory law or a supervisory authority requires it.
Read it rather than take our word for it: Data Processing Agreement · Privacy Policy · Terms · Cookies.
The answer economy is being claimed in real time.
So is this category.
AI-visibility software is where SEO tooling was in 2004 — except the land grab is measurably faster, and the winners compound.
Projected generative-AI market by 2033, growing at ~40.8% CAGR — every brand in it will need to manage how AI represents them.
Source: Grand View Research, Generative AI Market Report 2026–2033, April 2026of organizations use AI regularly or across most business processes — the buyer side of the answer economy is no longer hypothetical.
Source: Publicis Sapient, 2026 Global Enterprise AI Report, 1,550 decision-makers, June 2026PRO and Agency sit squarely in the mid-market gap where competitors are either enterprise-priced or monitoring-only — with a free tier feeding the funnel from below.
Measurement is commoditizing across the category. A closed measure→fix→prove loop is product architecture, not a changelog entry — that's the defensible layer.
Aeloria is Swiss-built, privacy-first and multilingual by design — the natural wedge into European enterprise, where data posture is a buying criterion, not a checkbox. The category's incumbents raised on dashboards; Aeloria is built on the layer that survives a budget review: shipped fixes and measured movement.
Questions we get asked — by humans and by AI.
What is an AI reputation training system?
What is generative engine optimization (GEO)?
What is answer engine optimization (AEO)?
Which AI engines does Aeloria monitor?
Why not Copilot or Google AI Overviews?
How is Aeloria different from Profound, Peec AI or Otterly AI?
Where does Aeloria's data come from?
How quickly does Aeloria show results?
The brands AI recommends tomorrow
are training their reputation today.
Send us your brand and your main market — we set up the free plan and run your first Brand AURA measurement. When you're ready to close the gaps, the Fix Engine is one upgrade away.