City market report A research page, not an office page
London 6 local businesses scannedAn AI search readiness study by MoxSEO, a remote digital marketing agency for London businesses
Looking for a digital marketing agency in London? MoxSEO is an SEO and AI search team that works with London businesses remotely; we have no office there. Before advising anyone, we scanned 6 London business websites on 9 checks. Their median score was 70.5/100, 9th of 20 cities. 1 of 6 passed "author and dates declared".
Every London homepage had exactly one H1 and structured data, but only 2 of 6 used question-form headings, the lowest share of any city. thegoodmarketer.co.uk scored 91. pearllemon.com was the only site to name an author and dates.
London sample, by row
Published, CC BY 4.0Share of checks passedNot scanned
- 91thegoodmarketer.co.uk
- 73pearllemon.com
- 73makeagency.co.uk
- 55influencedigital.co.uk
- 68londonmarketingcompany.co.uk
- 64passion.digital
Six London websites, scored row by row
Every business in the London sample, with its score out of 100 and the result of each of the nine checks. These are diagnostic observations from one homepage scan, not endorsements or a ranking of the businesses.
| Website | Score / 100 | Answer | Q-heads | One H1 | Lists | 3+ H2 | Paras | Author | Cites | Schema | Opening words | JSON-LD blocks | AI crawlers in robots.txt |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| thegoodmarketer.co.uk | Direct answer near top: Pass | Question-form headings: Pass | Exactly one H1: Pass | Lists or tables: Pass | Three or more H2 sections: Pass | Paragraph length: Pass | Author and dates declared: Fail | Cites external sources: Pass | Structured data present: Pass | 32 | 5 | None named | |
| pearllemon.com | Direct answer near top: Fail | Question-form headings: Fail | Exactly one H1: Pass | Lists or tables: Pass | Three or more H2 sections: Pass | Paragraph length: Pass | Author and dates declared: Pass | Cites external sources: Pass | Structured data present: Pass | 8 | 4 | None named | |
| makeagency.co.uk | Direct answer near top: Pass | Question-form headings: Fail | Exactly one H1: Pass | Lists or tables: Pass | Three or more H2 sections: Pass | Paragraph length: Pass | Author and dates declared: Fail | Cites external sources: Fail | Structured data present: Pass | 32 | 1 | None named | |
| influencedigital.co.uk | Direct answer near top: Fail | Question-form headings: Fail | Exactly one H1: Pass | Lists or tables: Pass | Three or more H2 sections: Pass | Paragraph length: Pass | Author and dates declared: Fail | Cites external sources: Fail | Structured data present: Pass | 104 | 2 | None named | |
| londonmarketingcompany.co.uk | Direct answer near top: Pass | Question-form headings: Fail | Exactly one H1: Pass | Lists or tables: Fail | Three or more H2 sections: Pass | Paragraph length: Pass | Author and dates declared: Fail | Cites external sources: Pass | Structured data present: Pass | 67 | 1 | None named | |
| passion.digital | Direct answer near top: Fail | Question-form headings: Pass | Exactly one H1: Pass | Lists or tables: Pass | Three or more H2 sections: Pass | Paragraph length: Pass | Author and dates declared: Fail | Cites external sources: Fail | Structured data present: Pass | 22 | 2 | None named |
- Score range
- From 55 (influencedigital.co.uk) to 91 (thegoodmarketer.co.uk). Median 70.5, above the pooled 148-site median of 68, 9th of 20 cities.
- Most common signal
- Exactly one H1; Three or more H2 sections; Paragraph length; Structured data present: 6 of 6 sites.
- Least common signal
- Author and dates declared: 1 of 6 sites, against 16.2% pooled.
- Answer and provenance
- Direct opening answers on 3 of 6 sites, a named author and date on 1, citations on 3, structured data on 6. A concise answer helps retrieval; named responsibility and sources make claims checkable.
- Crawler handling
- None of the 6 London sites blocked a named AI crawler in robots.txt. None named an AI agent at all, so their robots rules are silent on AI crawlers.
- Strongest site
- Highest score: 91, by thegoodmarketer.co.uk.
Four audit stages, nine scored checks
The audit order has four steps. Under each sit the checks the study scored on every homepage, with the pooled 20-city pass rate beside the London count. With 6 sites, one business moves the London figure by 17 points, so read the counts, not just the percentages.
| Signal | What it means | 20-city pooled result | London |
|---|---|---|---|
Step 1Confirm accessRobots rules, response status, canonical output, rendering, internal links and indexation for the page types tied to real demand. | |||
| AI crawler access | Whether robots.txt lets AI crawlers fetch the site. The study parsed rules for 14 AI crawlers. Recorded, not scored. | 9 of 148 block one or more22 name AI agents; 14 of those block none | 0 of 6 block |
Step 2Test answersWhether priority pages answer the customer question early, use descriptive sections and expose facts that can be extracted accurately. | |||
| Direct answer near top | An opening paragraph of 25 to 90 words that answers the question the page exists for. | 3 of 6 (50%) | |
| Question-form headings | Headings phrased the way people ask, so a section maps to a query. | 2 of 6 (33%) | |
| Exactly one H1 | One main heading, so the page's subject is unambiguous. | 6 of 6 (100%) | |
| Lists or tables | Facts set out in structures a machine can lift without rewriting. | 5 of 6 (83%) | |
| Three or more H2 sections | Enough descriptive sections for each to answer something on its own. | 6 of 6 (100%) | |
| Paragraph length | Paragraphs kept short enough to read, and quote, as one unit. | 6 of 6 (100%) | |
Step 3Verify provenanceBusiness identity, authors or reviewers, dates, citations, qualifications, policies and local evidence, in proportion to the risk of the claim. | |||
| Author and dates declared | A named author and a published or updated date: who stands behind the page, and when. | 1 of 6 (17%) | |
| Cites external sources | Links out to the sources a claim rests on. | 3 of 6 (50%) | |
| Structured data present | Schema.org markup that states the business, page and facts in machine-readable form. | 6 of 6 (100%) | |
Step 4Measure actionsQualified discovery connected to calls, enquiries, appointments, applications, demos or visits, without presenting attribution as perfect. | |||
| Qualified actions | Whether search visits turn into the step a customer can actually complete. | Not part of the homepage scanMeasured per site, in an audit | Not scanned |
What 6 sites can show, and what they can't
The scan records visible technical and editorial signals from 6 London business websites at one point in time. It shows recurring implementation gaps and good questions for an audit. It cannot prove market-wide prevalence, customer preference, future AI citations or the commercial impact of any single signal.
- What this page shows
- Every London row from the published file, its nine check results and how the city compares with the other 19.
- What it doesn't claim
- That these 6 sites represent every London business, or that a higher score means more traffic or citations.
- How sites were picked
- From Google Places rather than at random: 6 to 9 sites per city, 148 in all, across 7 countries.
- Still useful for
- Auditing your own site. Recommendations come from its rendered pages, business facts, journeys and analytics, not from a city average.
One file, 6 of its rows
1 dot = 1 websitePooled CSV, 148 rows20 cities, published, CC BY 4.0
London, 6 rowsInside the published CSV
Download the pooled file: city-market-study-2026-07-27.csv. How evidence is handled: methodology. All findings: the 20-city study.
What to fix first in London
Start with the checks London sites failed most often, then test your own pages before assuming the pattern applies to them. Begin with the service, location and conversion pages behind your most valuable decisions, and ignore the city comparison wherever your own evidence points to a bigger problem.
Put a name and a date on the work
1 of 6 London homepages passed "author and dates declared" (17%), against 16.2% across all 148. Only pearllemon.com did. It is the cheapest provenance signal to add: a named, accountable person and a visible published or updated date.
17%of London sites passPhrase sections as the questions customers ask
2 of 6 London homepages passed "question-form headings" (33%), against 66.2% across all 148. Only thegoodmarketer.co.uk and passion.digital did. A heading that matches a real query gives a search engine or AI system a clean section to lift.
33%of London sites passOpen each priority page with the answer
3 of 6 London homepages passed "direct answer near top" (50%), against 41.2% across all 148. Put a 25 to 90 word answer first, then the detail that supports it.
50%of London sites passCheck what crawlers actually receive
No London site in the sample blocked an AI crawler in robots.txt, but robots rules are only the first gate. Verify the rendered page after scripts, templates, consent tools and plugins have run, not what the editor shows.
Connect discovery to a step customers can take
Calls, enquiries, appointments, applications, demos or visits. Measure qualified actions, and say where attribution is uncertain.
Then match the page to the customer journey
The London market includes finance, legal services, technology, property, healthcare, and international business. Each needs different proof on the page.
Name the regulated entity, the products, fees, risks and who reviewed each page and when. Keep rates and calculators current, and say where figures come from.
- Regulated entity named
- Fees and risks stated
- Reviewed and dated pages
Show the firm, the lawyers responsible for each practice area, the jurisdictions covered, how fees work, and an enquiry route that screens before it sells.
- Responsible lawyers named
- Jurisdictions stated
- Fee basis explained
A technical evaluator and a buyer should be able to verify the same offer. Connect product, use-case, integration, documentation, security, pricing and sales pages.
- Integration and docs pages
- Security and pricing in reach
- One path to a demo
Clarify the legal entity, service areas, inventory or project status, location evidence, reviews, fees, and the exact action a prospective customer can take.
- Legal entity and service area
- Project status and fees
- One named next step
Make services, providers, locations, clinical review, update dates, sources, privacy boundaries and appointment routes visible, without overstating medical outcomes.
- Named clinical reviewers
- Dated, sourced pages
- A clear appointment route
Twenty cities, London in context
The pooled study is the comparison point, not a league table. Each city figure comes from 6 to 9 sites picked from Google Places, so read it as the shape of a market. London's nearest comparison, Dubai, had a median of 66 against London's 70.5.
- London median score, against 68 across all 148 sites
- 70.5/100
- 9th by median score among the 20 cities in the pooled file
- 9 of 20
- London homepages declaring author and dates, against 16.2% pooled
- 17%
- ChennaiIndia43% author and dates, 57% direct answer near top
- BengaluruIndia38% author and dates, 38% direct answer near top
- KolkataIndia33% author and dates, 67% direct answer near top
- AhmedabadIndia33% author and dates, 56% direct answer near top
- SingaporeSingapore29% author and dates, 29% direct answer near top
- TorontoCanada29% author and dates, 29% direct answer near top
- GurgaonIndia22% author and dates, 22% direct answer near top
- LondonUK17% author and dates, 50% direct answer near top
- MumbaiIndia17% author and dates, 33% direct answer near top
- New YorkUSA12% author and dates, 62% direct answer near top
- DelhiIndia12% author and dates, 50% direct answer near top
- JaipurIndia12% author and dates, 38% direct answer near top
- PuneIndia11% author and dates, 44% direct answer near top
- SydneyAustralia11% author and dates, 44% direct answer near top
- NoidaIndia0% author and dates, 43% direct answer near top
- DubaiUAE0% author and dates, 38% direct answer near top
- BhopalIndia0% author and dates, 33% direct answer near top
- ChandigarhIndia0% author and dates, 33% direct answer near top
- IndoreIndia0% author and dates, 33% direct answer near top
- LucknowIndia0% author and dates, 25% direct answer near top
- HyderabadIndiaScanned separately. No values published, so no position.n = 13
Source: city-market AI readiness study, July 2026. Six cities recorded 0% on attribution. Per-city samples are small; the study calls these figures indicative. City names link to each city's page.
Digital marketing for London businesses, scoped to one site
MoxSEO works with London businesses remotely. We have no office or address in London and do not pretend to. Every engagement starts from evidence on your own site, the way this page starts from evidence on 6 local ones.
- Local SEOBusiness information, location pages, reviews and conversion paths
- AI SEO and GEOExtractable answers, provenance and AI crawler access
- Technical SEORendering, canonicals, indexation and internal links
- International SEOCountry targeting, hreflang and market-by-market pages
- Digital marketingContent marketing and the wider programme around search
- Local SEO packagesPublished starting points for local programmes
Test the same signals free: AI Answer Extractability CheckerAI Crawler Access CheckerNAP Consistency Checker
Questions about London
MoxSEO is an SEO, AI search and digital marketing team that works remotely with businesses across London. We have no office or address in London. This page is our research on 6 local business websites, which we use as starting evidence in a London audit, alongside the evidence on your own site.
The 6 London homepages we scanned had a median score of 70.5 out of 100, above the pooled median of 68 and 9th of 20 cities by median. Scores ran from 55 (influencedigital.co.uk) to 91 (thegoodmarketer.co.uk). The sample is small and picked from Google Places, so treat it as indicative, not a census.
"Author and dates declared": 1 of 6 passed, against 16.2% across all 148 sites. "Question-form headings": 2 of 6 passed, against 66.2% across all 148 sites. Both are fixable in a template or a content edit, which is why they lead our London audit order.
None of the 6 London sites blocked a named AI crawler in robots.txt. None named an AI agent at all, so their robots rules are silent on AI crawlers. Across all 148 sites, 9 blocked at least one of the 14 AI crawlers the study parsed.
London: median 70.5, 17% named an author and date, 50% opened with a direct answer (6 sites). Dubai: median 66, 0% named an author and date, 38% opened with a direct answer (8 sites). Both samples are small, so a difference of one site moves the percentages a lot.
Local SEO, AI SEO and GEO, technical SEO, international SEO, content and the wider digital marketing programme, all delivered remotely. Work starts with an audit of your own pages; in London the first fixes are usually "author and dates declared" and "question-form headings", the two checks local sites failed most.
The other 20 city reports
All researchFigures per city: share of sampled homepages declaring author and dates, and opening with a direct answer. Every page is a research report on that market, not an office. How evidence is handled: methodology.
Test your site,
not the city average.
Bring your site, your goals and your constraints. In a free 30-minute call, Sakshi Kumari or Ashish Khan, the specialists who run MoxSEO, check access, answers and provenance on your priority pages and tell you the most useful first move.
A call is not an engagement. We never guarantee rankings, traffic, revenue or AI citations.