Not faults, so no score: openings found in your competitors' own data —
searches you nearly take from them, gaps only one of them owns, websites
that already link to them, sources the AI quotes instead of you. Pick one
or two per month and treat them as projects.
-
Competitors read and profiled
developer task
(B. Who you are actually up against)
What to do with it:
Profiling runs automatically, one competitor at a time, after the keyword comparison against that competitor is complete. It reads the competitor's homepage and its best-ranking pages, so it takes a few minutes each — the remaining ones will fill in.
What the data says:
36 of 91 competitor(s) profiled.
Why it matters: Whether each competitor's own pages have been read and turned into a written profile (its market, segments, offer and buyer personas). Numbers tell you who is winning; the profiles tell you what they are actually selling and to whom.
-
How much search territory you share
content work
(B. Who you are actually up against)
What to do with it:
Start with the competitors at the top of the overlap list: they compete with you on the most searches, so every position you take from them is traffic you actually wanted.
What the data says:
Median overlap 2 shared search(es) per competitor; academyfinepaintings.com overlaps the most with 2.
Why it matters: The number of searches you and each competitor both rank for. A high overlap means you are chasing the same buyers with the same content; a low one means the domain merely brushes your market and matters less.
-
Is the market concentrated or fragmented?
content work
(B. Who you are actually up against)
What to do with it:
No domain dominates these results, which is the easier situation: steady publishing against the gap list below takes territory without a fight.
What the data says:
The top 3 competitors hold 16% of all the search overlap — a fragmented market.
Why it matters: How much of the shared search territory the top three competitors hold. A concentrated market means a few domains own the results and you need to out-rank them specifically; a fragmented one means nobody dominates and consistent publishing wins.
-
The few competitors worth your attention
content work
(B. Who you are actually up against)
What to do with it:
Read the rest of this report through these few: their gaps are your backlog, their link sources are your outreach list, their AI citations are the answers to win back. Everyone else on the list is context, not competition.
What the data says:
Focus on: academyfinepaintings.com (sells what you sell (82% match), weaker than you — beatable with content, 2 search(es) shared with you); irisleegallery.com (sells what you sell (82% match), weake…
Why it matters: Nineteen domains is a list, not an answer. This picks the few whose business is closest to yours — weighing their similarity with your business, the AI citations they win against you, whether you can beat them on authority, and how much search territory you share — so you know who to actually watch.
-
The gaps actually worth writing
content work
(C. What they win and you do not)
What to do with it:
Write these in order. One page per search intent, answering the question completely — and link to it from the page of yours that already ranks closest to the topic.
What the data says:
702 qualified gap(s) worth ≈25,719,380 search(es)/month. Biggest: hydrangea, president garfield, venetian in italy.
Why it matters: The backlog filtered three times: searches under 60/100 difficulty — the ones a good new page can rank for without years of link building — on the topics you already rank for, and NOT owned by a competitor whose business was scored as different from yours. A reachable search from the wrong audience is a page that ranks and sells nothing.
-
What that missing traffic is worth
content work
(C. What they win and you do not)
What to do with it:
Use this figure to size the effort: if a handful of pages recovers a fraction of it every month, the content work pays for itself quickly.
What the data says:
≈$32,193,100/month of equivalent Google Ads spend sits in the qualified gaps.
Why it matters: What the qualified gap searches would cost to buy in Google Ads every month, using their real cost per click. It converts the backlog into a budget figure — the amount you are effectively leaving on the table each month.
-
Gaps that are also AI buyer questions (GEO)
content work
(C. What they win and you do not)
What to do with it:
Write these as question-and-answer pages: the question as the title, a complete self-contained answer in the first screen. That single shape serves the Google searcher, the AI assistant, and the buyer it quotes the page to.
What the data says:
79 gap search(es) match a stored buyer question, ≈865,320 search(es)/month. Biggest: famous art paintings, pop art, art and renaissance.
Why it matters: The gap searches crossed with the questions buyers in your market ask AI assistants (stored per market segment). A gap matching one of those questions is demand on both surfaces: the page that answers it can earn the Google position AND the AI citation — the highest-yield page this report can point at.
-
Which competitor owns the gaps
content work
(C. What they win and you do not)
What to do with it:
Read that competitor's market and segment profiles below, then work its gap list as one project — the searches will share a theme, and pages written together rank better than scattered ones.
What the data says:
1st-art-gallery.com alone holds 50 of the 714 gap search(es) (7%) — and its business matches yours at 81%.
Why it matters: Whether the demand you miss is spread across the field or concentrated in one competitor — and whether that competitor's buyers are yours. A backlog owned by a real rival is a roadmap; one owned by a different business is mostly noise.
-
Searches you win head to head
content work
(E. Head-to-head on the searches you share)
What to do with it:
You sit below your competitors on most of the searches you already share. These pages exist and already rank — improving them is far cheaper than writing new ones, so start here before the gap list.
What the data says:
No head-to-head comparison stored yet.
Why it matters: Among the searches where you and a competitor BOTH appear, the share where you sit higher. This ignores size entirely and measures the only thing the buyer sees: which of you is above the other.
-
First-page presence on shared searches
content work
(E. Head-to-head on the searches you share)
What to do with it:
Most of your shared searches sit past the first page, where they earn nothing. Pick the ones between positions 11 and 20 first — they are one solid update away from real traffic.
What the data says:
No head-to-head comparison stored yet.
Why it matters: Positions beyond the tenth bring almost no clicks, so a shared search only counts when you are on page one. This compares how often each side gets there.
-
High-volume searches you are losing
content work
(E. Head-to-head on the searches you share)
What to do with it:
Take these one at a time: open the competitor's ranking page next to yours, and match what it covers that you do not. These are the searches where a single afternoon moves real traffic.
What the data says:
No head-to-head comparison stored yet.
Why it matters: The shared searches with real monthly demand where a competitor sits above you. Each one is traffic flowing past a page you already own — the most concentrated loss in the report.
-
Who wins the searches that sell
content work
(E. Head-to-head on the searches you share)
What to do with it:
Prioritise the lost buy-intent searches whatever their volume: one hundred buyers beat ten thousand readers. The head-to-head table marks each search's intent so you can filter for them.
What the data says:
No head-to-head comparison stored yet.
Why it matters: The head-to-head record split by intent: searches advertisers bid on (cost per click of $0.30 or more — buyers) against the rest (readers). A poor overall win rate matters little if the searches you DO win are the ones that sell; a good one is worthless if every commercial search goes to a competitor.
-
Sources the AI trusts that Google does not show (GEO)
content work
(F. Who the AI assistants cite instead of you (GEO))
What to do with it:
These are your GEO targets: getting mentioned ON those sites puts you inside the answers even when the assistant does not cite you directly. Treat the recurring ones as an outreach list of their own — and when one of them is a seller of what YOU sell, it is a real rival Google never showed you: use the "Track as competitor" button next to it in the table below so the whole report starts comparing you against it.
What the data says:
27 domain(s) outside your competitor set are cited: google.com, galrie.com, mireillefineart.com.
Why it matters: Domains cited by the AI answers that are NOT among your Google competitors: reference sites, marketplaces, forums, media. They are competing for the same buyer attention on a surface where your usual competitor list does not apply.
-
How wide the AI test is (GEO)
content work
(F. Who the AI assistants cite instead of you (GEO))
What to do with it:
Promote more pages to widen the test: each promoted page adds its own questions, and the citation picture only becomes reliable across a few dozen of them.
What the data says:
30 question(s) tested against gpt-4.1-mini (web search on), producing 42 citation(s) across 28 domain(s) — last answer collected 2026-09-25.
Why it matters: How many distinct questions were put to the AI assistants, which models answered, and when. The test is generated from your promoted pages, so its coverage grows with the number of pages you promote.
-
What the cited pages all do (GEO)
content work
(F. Who the AI assistants cite instead of you (GEO))
What to do with it:
Apply the top habit to your promoted pages first — it is the one this market's winning pages share most, so it moves more citations than any generic best practice.
What the data says:
41 citation analysis(es) written; the recurring habits of the winning pages: question-shaped headings & lists (41×), figures, sources & citations (34×), a direct answer up front (28×).
Why it matters: Each lost citation gets a written analysis of why the cited page won. Tallied together they stop being anecdotes and become the playbook of your market: the habits the pages the AI trusts keep sharing — and therefore the changes most likely to win citations back.
-
How much your link profile could grow
outreach (ongoing)
(G. Link sources they have and you do not)
What to do with it:
Work the list steadily rather than in one campaign: a handful of new linking domains per month compounds, and looks natural while it does.
What the data says:
10 outreach prospect(s) in the sample, next to the 31 site(s) already linking to you (top 100 each).
Why it matters: The prospect list next to the link sources you already have. Both lists store the top 100 domains, so this sizes the sampled backlog — the real opportunity is at least this large.
-
Distance to the traffic leader
content work
(H. Search visibility compared)
What to do with it:
Open the leader's profile below and compare its segments with yours: the parts of the demand it serves and you do not are where most of that traffic difference is stored.
What the data says:
ebay.com collects ≈325,235,526 visit(s)/month across 35,946,769 search(es) — 2212486570.8× your traffic (leaving out 70 reference giant(s) — ahumbleplace.com, artprice.com, artrenewal.org — whose tra…
Why it matters: Which competitor collects the most Google traffic, and how far ahead of you it is. Unlike authority, a traffic gap can be closed with content alone when the leader's advantage comes from coverage rather than links.
-
Their traffic on YOUR territory
content work
(H. Search visibility compared)
What to do with it:
This gap closes search by search: the head-to-head section below lists exactly which shared searches that competitor wins and by how much — its close fights are where territory comes back cheapest.
What the data says:
No head-to-head comparison stored yet.
Why it matters: Whole-domain traffic mixes in demand that has nothing to do with you. This compares the visits earned on the searches you actually share — your side against each competitor's, summed over the sampled head-to-head rows. It is the number that measures the real fight.
-
Authority does not gate the AI answers (GEO)
content work
(I. Authority: their strength against yours)
What to do with it:
Do not let the authority gap above discourage the GEO work: citations are won with extractable answers, question-shaped headings and self-contained pages — content structure, not links. Start with the lost questions in the AI citations section while the outreach list runs in parallel.
What the data says:
The AI answers cite 28 domain(s); most are small sites whose authority is not even tracked — the AI surface does not require the links Google does.
Why it matters: The link deficit above applies to Google — not to AI answers. Assistants cite the page that answers the question best, whatever its domain's link weight, so this checks how many of the domains the AI cites carry no more authority than you. Every one of them is proof the AI surface is a fight you can start winning today, while the link work catches up.
-
Competitors compared with your business
developer task
(J. How close they are to your business)
What to do with it:
Scoring runs automatically after a competitor has been read and profiled, and needs your own descriptions to compare against — fill those in on the website page if they are empty, and the remaining competitors will be scored over the next runs.
What the data says:
33 of 91 competitor(s) scored against your business.
Why it matters: Whether each competitor's published positioning has been compared with yours — what you sell, the market you are in, the segments you serve and the buyers you address. Without that comparison, a competitor list is only a list of domains that share your words.
-
Direct rivals in the field
content work
(J. How close they are to your business)
What to do with it:
Read these first and treat the rest of this report through them: their gap keywords are demand from YOUR buyers, and the sites linking to them are the ones that should be linking to you.
What the data says:
6 direct rival(s) and 15 adjacent domain(s) out of 33 scored — closest: academyfinepaintings.com, irisleegallery.com, 1st-art-gallery.com.
Why it matters: How many of the domains Google ranks next to you are actually selling the same thing to the same buyers (70% similarity or more). These are the ones worth watching; the rest share your searches without competing for your customer.
-
Your closest competitor
content work
(J. How close they are to your business)
What to do with it:
Open its profile further down and read its segments and personas next to yours. Where its description is more specific than yours, that specificity is what is winning the shared searches.
What the data says:
academyfinepaintings.com matches your business at 82% — authority 162/1000, 2 search(es) shared with you, ≈257 visit(s)/month site-wide.
Why it matters: The single domain whose business most resembles yours, with the strength it brings. Whatever your market, this is the company your buyers are most likely to be choosing between you and.
-
Where the overlap is tightest
content work
(J. How close they are to your business)
What to do with it:
Where you are closest, your pages read like everyone else's and positions come down to authority. Where you are most distinct is where content can win on its own — say that difference explicitly on your commercial pages instead of leaving it implicit.
What the data says:
Closest on Buyer personas (33% on average), most distinct on Products & services (18%).
Why it matters: Which of the four compared dimensions — offer, market, segments, buyers — your competitors resemble you on most, averaged across the field. It tells you where you are indistinguishable, and on which dimension you still have room to stand apart.
-
Offer described for each competitor
developer task
(K. What they sell and who they sell it to)
What to do with it:
The offer profiles arrive with the rest of each competitor's analysis over the next runs.
What the data says:
Written for 36 of 91 competitor(s): academyfinepaintings.com, christies.com, 1st-art-gallery.com and 33 more.
Why it matters: Whether each competitor's products and services have been described from its own pages. Knowing exactly what they sell is what separates a real positioning decision from a guess.
-
Buyer personas described for each competitor
developer task
(K. What they sell and who they sell it to)
What to do with it:
The persona profiles arrive with the rest of each competitor's analysis over the next runs.
What the data says:
Written for 36 of 91 competitor(s): academyfinepaintings.com, christies.com, 1st-art-gallery.com and 33 more.
Why it matters: Whether the buyers each competitor writes for have been identified: who they are, what problem they want solved, and what makes them buy. Two competitors selling the same thing to different buyers are not really competing.
-
Market described for each competitor
developer task
(L. Their market and how they cut it up)
What to do with it:
These profiles are written automatically, one competitor at a time. The missing ones fill in over the next runs — each requires reading that competitor's homepage and its best-ranking pages.
What the data says:
Written for 36 of 91 competitor(s): academyfinepaintings.com, christies.com, 1st-art-gallery.com and 33 more.
Why it matters: Whether each competitor's own pages have been read and turned into a description of the market it operates in: what that market is, the demand it answers, how it works and who else serves it.
-
Market segments described for each competitor
developer task
(L. Their market and how they cut it up)
What to do with it:
The segment profiles arrive with the rest of each competitor's analysis. Once several are written, read them side by side — the segment missing from every list is usually the opening.
What the data says:
Written for 36 of 91 competitor(s): academyfinepaintings.com, christies.com, 1st-art-gallery.com and 33 more.
Why it matters: Whether the demand each competitor addresses has been cut into named segments. Comparing segment lists across competitors is how you spot the piece of the market everybody serves — and the piece nobody does.