AI Rank Monitor

Two tools in one — track how ChatGPT, Claude, Gemini, Perplexity & Google AI Mode rank your brand and your products against the competition

A step-by-step guide

01

Why AI Rank Monitor

Buyers ask AI before they buy — see whether AI recommends your brand and your products

Concept

From SEO to GEO

Your customers increasingly ask ChatGPT, Gemini, or Perplexity “what’s the best skincare brand?” — and act on the answer. AI Rank Monitor measures whether those models name and rank your brand — and your individual products — the way you used to track Google rankings.

SEO → GEO SEO is about ranking on Google. GEO — Generative Engine Optimization — is about being recommended inside AI answers. This tool is your GEO dashboard.

Real buyer questions It asks each AI the questions your customers actually ask — so a mention is earned, not prompted.

You vs competitors Every score is relative: where you stand against the rivals AI drops into the same answer.

Two tools, one platform

Brand Monitor vs Product Monitor

AI Rank Monitor has two tabs. Pick the one that matches the question you’re asking.

Brand Monitor Tracks your whole brand against competitor brands. “What’s the best skincare brand?” — category-level visibility.

Product Monitor Drills into one product line against rival product lines. “Which Cetaphil moisturizer beats CeraVe?” — product-level visibility.

When to use which Start with Brand Monitor for the big picture; switch to Product Monitor to defend or grow a specific product line.

How it works

One brand or product, five AI providers, one scoreboard

Give it a brand (or a product line). It writes buyer questions, sends them to five AI providers, reads every answer, and scores how often — and how highly — each one gets recommended. All five answer from a live web search and cite what they read, so every score traces back to real pages on the Sources tab. Google and Google AI Mode stay separate: Google AI Mode reads Google’s AI Overviews.

Brand / product

Your brand or
product line

→

5 AI providers

OpenAIAnthropicGooglePerplexityGoogle AI Mode

→

Scores out

VisibilityShare of VoiceRankingSentiment

1 · Ask Up to 20 buyer questions per analysis, written in your market’s language.

2 · Detect Every answer is scanned for mentions, list position, and (for brands) sentiment.

3 · Score Rolled up into Visibility, Share of Voice, and rankings.

Read the numbers

The numbers you’ll see

Defined once here. The result screens just show you where each lives — and where Brand and Product differ.

Visibility Score How much of the answer space you own, 0–100. Mostly how often an AI names you; where you rank nudges it. It’s a score, not a percentage: read it against your competitors’, not on its own.

Share of Visibility Your Visibility as a slice of everyone’s, the donut headline. Zero-sum: it falls when a competitor gains, even if your own score held. 16.4% = a sixth of the attention in your set.

Share of Voice Of every brand mention one AI made, the share that named you. Per AI, in Provider Rankings. Mentions only: position and sentiment don’t enter it. That’s what separates it from Share of Visibility.

Coverage vs Score Both matrices show 0–100 cells measuring different things. Brand’s is Coverage, the % of prompts that named you. Product’s is a rank Score. Never compare the two.

Sentiment The tone of the answers that named you, per AI. One majority verdict, not a scale: a “positive” can still hide a negative answer or two. Read those in Questions & Responses. Brand Monitor only.

02

Brand Monitor

Track your whole brand — from a name to live results

Step 1

Open AI Rank Monitor

The page opens on Brand Monitor. Two tabs sit up top — Brand Monitor (whole brand) and Product Monitor (a product line)
A dashed panel names the five AI providers each question runs against: OpenAI, Anthropic, Google, Perplexity, and Google AI Mode
Top right shows credits used and your organization; the User guide button opens this deck

Step 2

Enter Your Brand & Market

Pick the market (locale) first — the questions and brands AI surfaces depend on it, so the wrong market gives the wrong answers
Type a brand name — e.g. Cetaphil. A name works better than a URL: even with no website, AI Mode can profile it. A URL is optional
Advanced options lets you set Question language separately (default “Auto — match market”). Identifying the company is free

Company card for Cetaphil plus the auto-discovered competitor grid

Step 3

Confirm the Company & Pick Competitors

AI Rank Monitor profiles the brand — industry, description, keywords — even from just a name
It auto-discovers up to 12 competitors, all selected by default. Toggle any off, + Add competitor, or Bulk edit — they define your scoreboard
Click “Continue to Questions →”. Finding competitors is still free

AI-generated buyer questions, each with five provider icons, and the High Effort toggle

Step 4

Review the Questions

AI writes up to 20 buyer questions in your market’s language — none of them names your brand, so a mention is earned. Edit, remove, or “+ Add Question” (up to 50)
Flip High Effort Mode to ask each provider 5× instead of once — steadier scores, higher cost
“Start Analysis” fires every question at all five providers

Cost: 10 credits per analysis (20 in High Effort), charged when the analysis runs. Earlier steps are free

Live analysis progress with per-provider status icons and a live stat strip

Step 5

Watch It Run

Each question shows a live status per provider — a spinner for running, a green check for done, grey for skipped
A live stat strip counts up: prompts run, brand mentions detected, and competitors found
Up to 20 questions × 5 providers is up to 100 runs (×5 samples in High Effort); the count badge on the tab is the real number. Results open automatically when it finishes

Results · Brand Ranking

Where You Rank Overall

The donut splits total visibility across every brand. The centre shows your rank and Share of Visibility — Cetaphil lands Rank #3 at 16.4%
The ranked legend reads top to bottom with a bar and raw Visibility score per brand; a “You” pill marks you. It names the top 8 and folds the rest into Other, but pins you as your own row even when you rank below 8th
Footnote: every brand has a Visibility score (0–100); those scores add up to a total, and each % is that brand’s slice of it

Comparison Matrix heat-map of brands by AI provider with an average column

Results · Comparison Matrix

Who Wins on Which AI

A heat-map: brands down the rows, the five AIs across the columns, plus an average. Darker = stronger
Each cell is Coverage % — how often that AI named the brand across the prompts. 100% = named in every prompt; 0% = never. Toggle % / Count
The KPI up top is Avg Coverage across every brand. Scan a column to see which AI favours which brand

Results · Questions & Responses

What Each AI Actually Said

Read each provider’s answer verbatim. A green “Brand Mentioned” badge and a “Position #N” tag tell you where you landed
Switch provider tabs to compare wording across OpenAI, Anthropic, Google, Perplexity, and Google AI Mode
This is your evidence layer — the exact text behind every score, so you can see why a competitor outranked you

Provider Rankings table for one AI with mentions, average position, visibility, share of voice and sentiment

Results · Provider Rankings

Drill Into One AI

Pick one AI. The table ranks every brand by Visibility, with Mentions, Avg Pos, Share of Voice and Sentiment beside it. KPI: Average Rank
Stat cards summarise your standing there: rank, visibility, share of voice, average position, sentiment
Under the table sit three charts: a Share of Voice bubble (top-right = more voice, better rank), the Visibility breakdown — next slide — and Position across all AIs, a bump line where a gap means that AI didn’t rank you

Visibility breakdown bar showing Mention, Listed and Position making up the 0-100 score

Results · Provider Rankings

How the Visibility Score Is Built

The second chart under the Provider Rankings table — titled Visibility · {the AI you picked} — opens the score up. There’s no separate “Visibility” tab; it lives here, per AI
On the AI you’ve selected, Visibility (0–100) is Mention 70% + Listed 15% + Position 15%. Mention = how many replies named you (the main driver). Listed = a flat bonus every named brand gets, so it doesn’t separate them. Position = where you land in the list
Each brand is one stacked bar, and the dashed line marks the market average — so you can see who’s above or below it at a glance

Recommended Keywords with a keyword optimization score, quick wins, performance buckets and provider insights

Results · Recommended Keywords

Where to Improve

A Keyword Optimization Score (0–100) based on keyword frequency, competitive gaps, and provider coverage
Quick Wins are your top 3 opportunities: the words competitors most out-own you on
Performance Overview sorts every keyword by how much of its mentions you own
- Winning: you own most mentions of that word
- Competitive: you and your rivals split it
- Missing: rivals own it. This is the content backlog Quick Wins is drawn from
Provider Insights gives each AI its own card: the six keywords it leans on most, plus one tip. Use it to see which AI to write for
At the bottom, Full Keyword Analysis lists every keyword in a searchable table, filterable by category, sortable by Gap, with Export CSV. This is the list a content team works from

Sources tab with the platform reach chart, the platform-by-site heat map and cards of the most-cited sites

Results · Sources

Where the Answers Came From

Every AI now searches the web before it answers, and Sources — the last tab — lists the sites it read. The header counts them: “N citations across M sites”
Three charts sit on top: How wide each platform reads (citations and sites per AI), What each platform leans on (a site × AI heat-map), and Where to show up (the top 15 sites and which AIs reach each)
Below them, one card per site — the AIs that cited it, the count, and every cited page as a link. These are the pages worth influencing

Citation capture started 27 July 2026. Older analyses show “No sources recorded” — re-run to collect them

03

Product Monitor

Zoom into one product line and track it against rival products

Step 1

Discover Your Product Lines

Product Monitor starts from a brand name — e.g. Galderma. A brand name is required here; there’s no URL-only path
It extracts the brand’s product-line families from 3 sources: Website Scrape, AI Mode Search, and LLM Knowledge
Set the market first, just like Brand Monitor. Discovery is free

Step 2

Pick a Product Line

You get the brand’s catalog — e.g. Galderma · 23 product lines — each tagged with a category and a price tier (budget / mid / premium / luxury)
Aliases (alternate names AI might use) help match mentions back to the right line; Variants group its sub-products
Use the filter box, Add manually, or Re-extract, then hit Find Competitors on the line you want to track

Comparator products split into Direct and Indirect competitors with confidence badges and price tiers

Step 3

Choose Comparator Products

It finds rival product lines and validates each, split into Direct (head-to-head) and Indirect (adjacent) competitors
Each card shows the parent brand, a confidence badge, category, and price tier — so you compare like with like
Keep the ones that matter, then Generate Questions

Step 4

Review the Questions

Here the questions do name your brand — e.g. “What’s the best Cetaphil moisturizer for oily skin?” — because we’re detecting which product lines AI recommends
Edit, remove, or add your own (up to 50). High Effort Mode works the same as Brand Monitor
Start Analysis runs it — same 10 / 20 credit cost. The live progress screen is identical to Brand Monitor’s

Product Ranking donut with Cetaphil at Rank 1 and a ranked legend of product lines

Results · Product Ranking

Where Your Product Ranks

Same donut as Brand Monitor, but across product lines. Cetaphil lands Rank #1 at 33.2% Share of Visibility
The legend shows each line with its parent brand, a bar, and its Visibility score. It names the top 8 and folds the rest into Other, pinning your own line even when it ranks below 8th
Footnote: same share math as Brand Monitor. Every line’s Visibility score (0–100) adds into a total, and each % is that line’s slice of it

Product comparison matrix heat-map using rank-based Score % across the five AIs

Results · Comparison Matrix

Who Wins on Which AI

Same heat-map shape as Brand — but Product cells are a rank Score (0–100), not a Coverage percentage
Score reflects ranking position: #1 = 100, −10 per lower rank; 50 = mentioned but unranked; 0 = not mentioned. The toggle reads Score % / Position #, but that number is a rank score, not a share of anything
The KPI up top is Average Score across every line and AI

An expanded product question showing a provider's verbatim answer with a Product Mention badge and position tag

Results · Questions & Responses

What Each AI Said About Your Products

Read each AI’s answer verbatim, with a green “Product Mention” badge and a “Position #N” tag
Answers rank products with their parent brand — e.g. “Cetaphil by Galderma”
Switch provider tabs to compare across all five AIs

Product provider rankings table with mentions, average position, visibility and share of voice, no sentiment column

Results · Provider Rankings

Drill Into One AI

The table ranks every line by Visibility, with Mentions, Avg. Position and Share of Voice. Note: no Sentiment column in Product Monitor
Stat cards: product lines, your rank, visibility, mention rate, share of voice, avg position, total mentions
The same three charts sit under the table, product-scoped: Share of Voice bubble, the Visibility breakdown — next slide — and Position across all AIs

Product visibility breakdown showing mention rate and position score combining into the score

Results · Provider Rankings

How the Product Score Is Built

On the same Provider Rankings tab as the last step, scroll past the table to the chart headed “What builds the Visibility score”. Brand’s version sits in the same spot but is headed Visibility · {the AI you picked}
Product Visibility is Mention 70% + Position 30%, simpler than Brand (no Listed bonus, no keyword term)
Position score = 100 − (avg rank − 1) × 15, so every rank below #1 costs 15 points
That’s why a product ranked #1 across AIs can post a much higher Visibility than a brand with the same mention rate

Product recommended keywords with an optimization score, performance buckets and provider insights

Results · Recommended Keywords

Where to Improve

Same as Brand Monitor: a Keyword Optimization Score, Winning / Competitive / Missing buckets, and per-provider insights — scoped to your product line
Use it to spot terms the AIs tie to rival lines but not yours
Export the list to brief your content team

Product Monitor Sources tab showing the same platform reach, heat map and top-sites charts scoped to a product line

Results · Sources

Where Product Answers Came From

Sources is the last tab here too, and it is the same view as Brand Monitor’s — the three charts and the site cards, scoped to this product-line run
This is where retailer and review pages surface: if Lazada, Shopee or a review site is what the AIs read about your category, it shows up here with its citation count
Cited pages arrive with different detail per AI — Perplexity returns links without page titles, and Google’s link is a redirect. Both still render as the site, so no row is ever blank

Product vs Brand

How Product Monitor Differs

Same six report sections — but a few things change when you go product-level.

Starts from a brand Enter a brand, pick a product line, then compare it against rival lines — Direct & Indirect.

Visibility 70 / 30 Mention 70% + Position 30% — no Listed bonus, no keyword term.

Matrix = rank Score A rank score of 0–100 (#1 = 100, −10 per rank), not a Coverage percentage.

No Sentiment Product Monitor doesn’t score tone. Everything else mirrors Brand Monitor.

04

Manage, Export & Schedule

Every run is saved — revisit it, export it, and put it on a schedule

History

Revisit & Re-run

Open the History sidebar (the ☰ button) for every past analysis — brand or line, market, and when you ran it
Click any entry to reload its full results instantly — no credits. Each tab keeps its own history
Re-run this analysis refreshes the same setup; + New Analysis / Analyze another starts fresh. To track movement without remembering, put it on a schedule instead — next slides

Export

Take the Report With You

Export PDF sits in the left Report Sections rail, under the tab list — same button on both monitors, on the results view of a saved analysis. It downloads straight away; there’s no job to wait for and nothing is emailed
You get five sections: Brand (or Product) Ranking, Competitors / Comparators, Comparison Matrix, Provider Rankings, and Questions & Responses. The visibility donut is the only chart that crosses over
It’s a summary, not the full record: answers are cut at 1,200 characters and 8 per question, the matrix at 8 competitors, Provider Rankings at 10 rows, and there is no Share of Voice tile
Thai text renders properly — the report carries its own Thai font, so nothing turns into boxes

Two tabs stay portal-only and are not in the PDF: Sources and Recommended Keywords

Schedule

Put It on Repeat

From any finished analysis, click Schedule this analysis in the same rail. It re-runs the saved setup automatically and emails the report each time
Repeat offers Monthly or Weekly. Pick the day and time; the timezone is detected from your browser and can’t be changed here, and “Next 3 runs” previews the dates before you commit
Email the report to — tick org members or type any external address, up to 20. Works the same for a brand or a product line
A run starts around the time you picked rather than on the second, and it inherits High Effort from the analysis you scheduled it from

Creating a schedule is free — each run bills like a manual one at 10 credits (20 in High Effort). The dialog shows the yearly estimate before you commit

Schedule

Pause, Edit, or Stop It

The sidebar has a Schedules tab next to History. Each row reads its cadence and next date — Monthly · 15th, 09:00 · Next: 15 Sep — plus its last outcome. Brand and product schedules each sit under their own monitor; there is no combined list
Open one for Manage schedule: change the day, time or recipients, Pause and resume, or Run now for an immediate run (which costs a full run’s credits). Edits are locked while a run is in progress
Cancel schedule is final — it stops for good and drops off the list, so create a new one if you change your mind

Resuming a paused schedule recomputes the next run from today. It does not fire the run you missed

Completion email with the report-is-ready headline, the Download report button and the PDF attached

Schedule

What Lands in Your Inbox

When a scheduled run finishes, every recipient gets one email — subject “{Your brand} — AI Rank Monitor report” (or Product Monitor report), headed “…report is ready”. The body carries no numbers — it is a notification, and the report itself is the PDF
The PDF is attached, and a Download report button covers the case where it was too large to attach. That link is single-use and expires in 7 days
View in AI Rank Monitor opens the new run — where Sources and Recommended Keywords are waiting. On a product-line schedule the link lands on Product Monitor’s start screen instead; open that run from History in the sidebar

A failed run emails nothing at all. Out of credits and the run is skipped, not failed — the schedule stays active and you get one top-up email, then at most one more every 30 days

05

FAQ & Tips

The questions every first-time user asks

FAQ · Cost

Credits & Cost

What you pay, and when.

How much? A standard analysis is 10 credits; High Effort is 20. Identifying the company, finding competitors and comparators, generating questions, and discovering product lines are all free.

When am I charged? When the analysis runs. The setup steps before it cost nothing.

What about a scheduled run? The same 10 credits (20 in High Effort) — a schedule is billed per run, at the same price as running it by hand. So is “Run now”. The create dialog shows the yearly estimate before you commit.

Out of credits? You can still browse past results, but “Start Analysis” is blocked until your organization tops up. A scheduled run is skipped rather than failed, and resumes on its own after a top-up.

FAQ · Accuracy

Why Results Move

AI answers aren’t fixed — here’s how to read them right.

It’s a snapshot Models answer a little differently each time, and shift month to month. One run is a reading, not a verdict.

High Effort Mode Asking each provider 5× and averaging cuts the noise. Use it when a score sits on the fence, or for a report.

If an AI doesn’t answer Calls fail per prompt, not per provider — an AI might answer 7 of 10. Failed calls are dropped, not scored as zero, so they shrink the sample rather than dragging you down. Questions & Responses flags each one.

FAQ · Tips

Getting Sharper Results

Small inputs, big difference.

Set the right market The market (locale) drives the questions and the brands AI surfaces. Set the market you sell in — not the US default.

Curate competitors Drop irrelevant rivals, add the ones that matter. The competitor set is the yardstick every score is measured against.

Name, not URL A brand name is first-class and often more reliable than a URL — AI Mode can profile it without a website. Product Monitor needs a brand name.

“Session expired”? Your login lapsed — sign in again. The tool also needs a selected organization with credits to run.

Start Monitoring

Know what AI says about your brand and your products — before your customers do.

aissistance.us