AI Visibility Index · System · Wave 1 · August 2026 · UK · Home & lifestyle retail · Illustrative data
Northvale is easy to surface for people who already know the name, showing in 62 of 64 branded answers, but it appears in only 93 of 160 non-branded answers and is the top pick in 35 of 160, so open buyer questions still go to rivals like Meridian Home and Hearth & Co. That branded-versus-open split is the biggest pattern here, more important than any one engine result. Comparison is the clearest weak spot, and Claude shows Northvale in 17 of 40 answers while ChatGPT shows it in 27 of 40. Move first on non-branded questions that compare retailers, so engines can separate Northvale from the rival set earlier in the decision.
Reach and preference are different things: Meridian Home is named in 34 of the 40 category questions this wave, Northvale in 28. In the 27 answers where both appear, Northvale is the answer's top pick in 24% of them.
22%35 of 160
#1 Pick Rate
The dimension closest to revenue: how often Northvale is the number-one pick on earned questions.
Section 2
Score Anatomy
How the AI visibility breaks down. The top three matter most; the three below are supporting signals. Raw counts with their denominators, every number re-derivable from this wave's answer record. Rates here use the earned lane only. There's no composite score in this report: a blended index would hide which dimension actually moved.
#1 Pick Rate
Critical
22%(35 of 160)
How often you're the number-one pick on earned questions
Top 3 Rate
Moderate
38%(61 of 160)
How often you land in the top three options on earned questions
Visibility Rate
Moderate
58%(93 of 160)
How often you show up at all across earned questions
External Entity Signals
Critical
38%(3 of 8)
How present you are on outside sites AI trusts (Wikipedia, directories, reviews)
Site AI Readability
Strong
78%(62 of 80)
How easy your website is for AI to read and crawl
Brand Accuracy
Strong
76%(121 of 159)
How many of your visible mentions carry two or more verifiable facts the engine can cite
Denominators are frozen for the engagement. Every rate here is a countable ratio you can re-derive from the answer record: no weighting, no index, no score out of 100. The ladder reads in one direction, so 35 of 160 top picks always sits inside 61 of 160 top-three, which always sits inside 93 of 160 visible.
Section 3
Sentiment
How the engines describe Northvale when it does appear: the tone of every mention this wave, favorable versus neutral versus unfavorable.
54%
neutral, of 159 visible mentions this wave (of 240 answers across all lanes)
Neutral-dominant
Sentiment reads every mention, in every lane. Those 159 mentions are 93 on earned questions (the 160 answers the Score Anatomy section scores), plus 62 on brand-named questions and 4 on competitor-versus-competitor questions.
Favorable
60 (38%)
Neutral
86 (54%)
Unfavorable
13 (8%)
When Northvale is named, the engines describe it neutrally more often than not. That's rarely a reputation problem and usually an evidence one: there isn't enough on the public record for an engine to say why Northvale is the better choice, so it lists the brand without arguing for it.
Section 4
Visibility Rate By Buyer Intent
How often Northvale appears at each stage of the buyer journey. These three stages are 120 of the 160 earned answers this wave; the remaining 40 sit in intents that aren't journey stages, broken out in Funnel Coverage.
Early Research
75%
30 of 40 answers
"best places to buy homeware online uk"
Comparing Options
25%
8 of 32 answers
"Meridian Home vs Brightwell, which is better for linens"
Ready To Buy
52%
25 of 48 answers
"where to buy quality linens and homeware online"
Insight
Northvale is easiest to find while people browse and hardest the moment they choose. It shows up when a question gives the engine a reason to name it, but on comparing-options questions that hook disappears and so does Northvale. This is a discovery gap, not a desirability one. The lever is comparison content and third-party citations that give the engines a reason to name Northvale on the open questions too, where the buying decisions are actually made.
Section 5
Funnel Coverage
How visible Northvale is across the questions buyers actually ask, split by buyer intent and engine. Each cell is how many of that intent's earned questions the engine surfaces Northvale on, out of that intent's earned questions. Brand-named questions show coverage on their own questions: context, never part of any rate.
Buyer Intent
ChatGPT
Claude
Google AI
Perplexity
Local
8/10
6/10
8/10
8/10
Purchase Intent
8/12
4/12
7/12
6/12
Discovery
9/10
6/10
8/10
7/10
Competitive Displacement
2/5
1/5
3/5
2/5
Comparison
0/3
0/3
0/3
0/3
Direct Brand Context
16/16
15/16
16/16
15/16
Cells count only earned questions: 40 of the 60 asked. Excluded from every rate are 16 brand-named and 4 competitor-versus-competitor questions. Visibility on a question that names you is handed to you, and a head-to-head is a test of other brands, so those intents show coverage on their own question counts instead. Comparison is 0 of 3 on every engine, which is the single clearest gap in this wave.
Section 6
Your Must-wins
The 10 questions starred as must-wins at onboarding, each with its own result this wave, weakest first. These are the questions the playbook works first. 5 of 10 shown.
2 of the 10 must-win questions return zero visibility today; across all 10 Northvale is named in 22 of 40 answers.
"Meridian Home vs Brightwell, which is better for linens"
0 of 4 not named
"best online homeware store for a first flat"
0 of 4 not named
"where to buy quality linens and homeware online"
1 of 4 named, never the pick
"best home & lifestyle store to buy online"
2 of 4 1 as the top pick
"affordable homeware brands that last"
4 of 4 2 as the top pick
Sample view · truncated
5 of 10 must-wins shown
Your report carries all 10 must-win questions with the verbatim answer text from every engine, who won each one instead, and the playbook ticket that works it.
This is a preview
Get the full report
Every section below shows a sample. Your full report unlocks the complete playbook with every ticket costed and owned, the full readability and entity audit, the complete citation map, and ongoing monthly measurement, on your real data, in your portal.
Section 7
Action Playbook
All 18 owner-ready tickets, ranked on expected answer gains measured from this wave: the answers each fix can recover, weighted by buyer intent and how likely the work is to land. Now is your website work, first because engines cannot cite what they cannot read and it's the part you fully control. Next is external placements, which move answers fastest. Later is the long-lead programmes. 3 of 18 shown.
01
Now
days 0-14
Start here
Your highest-priority levers, the biggest measured gaps first.
10 tickets
Publish a comparison page that puts Northvale beside Meridian Home, Hearth & Co and Brightwell
Buyers ask 3 comparison questions and Northvale is named in 0 of the 12 answers they produce. The engines answer with the rivals alone, so Northvale is never in the shortlist the buyer walks away with.
Managed content · Q&A content · On-site · High impact
W1-01
Add 6 purchase-intent questions and answers to your site
On 4 of them the engines never mention Northvale at all; Meridian Home and Brightwell get recommended instead. Pages that answer buyer questions directly are the number-one thing AI cites from your own site.
Managed content · Q&A content · On-site · High impact
W1-02
Turn the 5 Discovery questions where you're mentioned into the number-one pick
Engines already name Northvale here: across these 5 questions they returned 20 answers, named it in 14 and picked it first in 4. The remaining move is position, not visibility: make the page that carries each answer the decisive source.
Managed content · Q&A content · On-site · High impact
W1-03
Sample view · truncated
3 of 18 tickets shown
The full playbook carries all 18 tickets across Now, Next and Later, each with an implementation brief, the owner who executes it, and the answers it's expected to recover.
Section 8
Site AI Readability
Everything the crawl can verify about how machine-readable your site is to AI engines. Unchanged honesty rule: a blocked or failed crawl produces zero structural claims.
The proof points that tell AI who you are and that you're real: your presence in public knowledge sources like Wikipedia, directories and review sites. We only show what we could actually verify this wave.
Wikipedia entity
Absent
No Wikipedia article references northvale.co.uk (weight: high)
Wikidata entity
Absent
No Wikidata entity lists northvale.co.uk as its official website (weight: high)
Review platform presence
Linked
Trustpilot profile exists and is linked from the site (weight: high)
Editorial roundups
Absent
Not named in the 5 editorial sources that carry most of this category's citations (weight: high)
Social profiles
Linked
4 found and linked from the site (weight: low)
Industry directory
Not linked
Category directory presence not yet checkable from this wave (weight: medium)
Google Business Profile
Linked
Profile exists with 240 reviews and is linked from the site (weight: medium)
Retail marketplace listing
Not linked
Marketplace storefronts exist but aren't linked from the site (weight: medium)
3 of 8 public trust signals are in place, checked against the live site and public knowledge sources this wave. 2 more exist but aren't linked from the site, and linking them is the fastest win here.
Section 10
Top Competitors (according to AI)
The brands AI names when buyers ask about the category without naming anyone, on the two grains that matter. Left is question reach: how many of the 40 earned questions name each brand at least once. Right is answer reach: how many of the 160 individual answers name them, so a brand all four engines agree on counts four times.
Question reach
earned questions naming each brand, of 40
Meridian Home
34
Hearth & Co
30
Northvale (you)
28
Brightwell
21
Verano
17
Olin & Vale
12
Answer reach
answers naming each brand, of 160
Meridian Home
124
Hearth & Co
96
Northvale (you)
93
Brightwell
71
Verano
38
Olin & Vale
22
One-on-one, counting every answer that names you both: Northvale is the pick over Meridian Home in 24% of 71 · over Hearth & Co in 46% of 63 · over Brightwell in 61% of 54 · over Verano in 74% of 31. Reach is close at the top of this table and preference isn't, which is the whole finding.
Section 11
Citation Intelligence
The domains engines trust for this category: where you're cited versus the competition, and the full cited-source ranking. Competitor-owned sources are shown as intel, never as targets.
1,860total citations across 312 unique domains this wave. Northvale's own domain accounts for 37 of them.
The sources that influence your answers
Engines read pages, then write answers. Showing the top 6 of 25 cited sources; the full list lives in your portal.
Citation source
Category
Volume
Northvale
goodhousekeeping.com
Editorial
148
Not mentioned
meridianhome.co.uk
Competitor site
131
Not mentioned
nytimes.com/wirecutter
Review
119
Not mentioned
apartmenttherapy.com
Editorial
96
Not mentioned
reddit.com/r/HomeDecorating
Community
88
Mentioned
trustpilot.com
Directory
74
Mentioned
What each category means for you: editorial is a pitch for inclusion · directory is claim or build a profile · community is earn genuine mentions · competitor site is a rival's own domain, excluded from outreach. An absence is an opportunity only on editorial and directory sources.
Sample view · truncated
6 of 25 cited sources shown
Your report carries the complete per-engine citation map, every source each engine cites for your category, plus the ranked list of which to earn next and what it takes to get on each one.
Section 12
Agent Analytics
AI-engine behavior measured on your own infrastructure. This wave's robots signal is below; full measurement activates when your server logs are connected.
Optional add-on
Current signal from this wave's crawl: 0 AI crawlers blocked by robots policy, and no explicit AI-crawler directives detected. Full crawler and referral measurement, meaning AI-engine fetches on your pages, verified bot identity, and the humans AI engines send you, activates when Northvale's server-log feed is connected (a 3 to 5 day hand-off; Cloudflare, nginx, Apache or an S3 drop).
Methodology & Instrumentation
Engines: ChatGPT, Claude, Google AI and Perplexity, web search, 60 answers each · Question set: 1 (frozen) · 240 answers total · Measurement window: 2026-07-28 to 2026-08-03
Scoring basis: raw counts with stated denominators, re-derivable from the answer record. No weighting and no composite index. Denominators: counted over earned questions, those where a buyer describes a need without naming you. Direct-brand and competitor-versus-competitor questions are reported separately and never enter a rate.
The limit we always name: this measures the retrieval layer, meaning what an engine fetches and cites in a live answer. No tool directly observes what a model absorbed in training, and anyone claiming to measure training-data exposure is overpromising. Engines also change models, which is why the instrument is re-run on a cadence rather than trusted as a snapshot.
This is an illustrative sample. Northvale is a fictional brand and every number here is synthetic, shown to demonstrate the structure of a real report. Nothing on this page describes a real company. In a live engagement every count re-derives to the specific answers behind it, on denominators frozen for the engagement.
Get your report
Tell us your site and category, and we'll run your AI visibility audit across all four engines.
✓
On its way
We'll run your AI visibility audit and send it across, with portal access to the full report. Pick a time to walk through it.