12 · The plan

The plan covers 90 days: quick fixes first, then the Profile and mobile speed, then the sites the AI tools read, then searches already close to page one.

example-property.co.uk · based on live checks on 8 July 2026, prepared by Robin Laires. Sequenced so the fixes with the most effect for the least effort go first.

The short version
What we found
Harbour & Key has 662 Google reviews at 4.6, about 550 articles and a £600 fixed fee. They may not yet be in the places Google and the AI tools read.
Why it matters
A seller choosing an agent may not see Harbour & Key on Maps, and the AI tools name it in 4 of 32 seller answers.
What could help
Quick page fixes in days, then Profile and mobile work in weeks 1 to 4, directory listings in weeks 2 to 8, and page-2 searches in weeks 4 to 12.
The strategy

The idea is to put the reviews, articles and fixed fee where Google and the AI tools read

Harbour & Key has the reviews, the articles and the fee already. The plan starts with fixes that need no one else, such as schema, headings and titles, then moves on to the Business Profile, the mobile build and the sites the AI tools cite. AI answers are noisy, so we would re-run the same checks each month and look at the trend.

Phase 0 · Days
Quick wins

Fix the page basics and the broken links.

Phase 1 · Weeks 1 to 4
The structural fixes

Put the Business Profile, the fees page and the mobile build to work.

Phase 2 · Weeks 2 to 8
The sources the AI tools cite

Get Harbour & Key onto the sites the AI tools already read.

Phase 3 · Weeks 4 to 12
What is already close

Move the searches that are near page one, and keep measuring.

The goals below are goals and not forecasts. Each one is a measurable state the plan is built to reach, so the work can be checked at the re-audit. The overview has the six findings behind these steps.

90
Days in the plan
4
Phases, from days to week 12
16
Actions in total
662
Google reviews at 4.6, the base the plan builds on
4 of 32
Seller AI answers that name Harbour & Key today
0 of 5
Maps searches with Harbour & Key in the top 20
Phase 0 · Days

Quick wins

Fix the page basics and the broken links.

Schema, headings and titles can go live without waiting on anything else, and everything after them builds on these.

  1. 1

    Structured data (schema)

    High
    One template change in the themeDeveloper

    Add RealEstateAgent, Offer (the £600 fixed fee) and FAQPage schema on the homepage, /fees/, /sell/ and the guide steps.

    What we saw: There is no RealEstateAgent, LocalBusiness, FAQPage or Offer schema on any of the 231 pages crawled, so the £600 fee and the 4.6 rating from 662 reviews may be hard for Google and the AI tools to read.

  2. 2

    Headings, titles and meta descriptions

    High
    DaysDeveloper

    Add H1s, titles and meta descriptions to the 53 core pages.

    What we saw: 42 of 53 core pages have no H1, including /sell/, /buy/, /fees/ and /valuation/, and 52 of 53 have no meta description. These are often the first things Google and the AI tools quote.

  3. 3

    Empty title tags

    Medium
    HoursDeveloper

    Fix the 8 builder partnership pages that have an empty title tag.

    What we saw: Pages like /barratt/ and /taylor-wimpey/ give Google and the AI tools a blank title.

  4. 4

    Broken inbound links

    Medium
    An afternoonDeveloper

    Add 301 redirects for the targets of the 36 broken inbound links.

    What we saw: 11 point at a sold listing, 7 at old /dvd/*.asp URLs from the 2000s, 7 at the valuation page that returns 403, and 2 at a mistyped market-update URL. The link value goes nowhere.

  5. 5

    The valuation subdomain

    High
    HoursDeveloper

    Ask someone inside the firewall to check enquiries.example-property.co.uk/valuation.

    What we saw: The subdomain returned 403 to every client we tested, and 7 external links point at it. A rule may be blocking real visitors, or the links may lead to a dead end. We cannot tell which from outside.

Phase 1 · Weeks 1 to 4

The structural fixes

Put the Business Profile, the fees page and the mobile build to work.

The 662 reviews at 4.6 do little in the Maps pack so far, and the mobile homepage takes 16.7 seconds to show its main content.

  1. 6

    Business Profile

    Critical
    Weeks 1 to 4Ops

    Rework the Google Business Profile: primary category, services, the £600 fee in the description, photos and weekly posts.

    What we saw: Harbour & Key is not in the top 20 for any of 5 live Maps searches, despite the largest review count of any Edinburgh agent we measured. The Profile may be under-filled for what it carries.

  2. 7

    Reviews at settlement

    Critical
    Weeks 1 to 4Ops

    Set up a way to ask for reviews at settlement, so new ones come in steadily.

    What we saw: 662 reviews is the biggest asset in this audit, and the Maps results still show Neilsons (4.8/194), Umega, Warners and VMH (4.9/321). More reviews, landing steadily, could be the most direct lever on the distance and prominence signals.

  3. 8

    The /fees/ page

    High
    Week 1Content

    Rebuild /fees/ as the page for “cost of selling a house in Scotland”: a 40 to 60 word direct answer at the top, the £600 figure stated plainly, a comparison with percentage fees, and an FAQ.

    What we saw: The only rankings on seller searches are informational ones at 27, 28 and 32 (“fixed fee estate agent scotland”, “how to sell a house in scotland” and “do i need a solicitor to sell my house in scotland”). /fees/ is 962 words, titled “Our Fees - Harbour & Key”, with no structured data, and the position 27 ranking is held by a guide step.

  4. 9

    Mobile JavaScript

    High
    Weeks 1 to 4Developer

    Cut the mobile JavaScript, with a first milestone of main content under 4 seconds.

    What we saw: Mobile performance is 44/100, and the main content takes 16.7 seconds against a 2.5 second target, with 746 KiB of unused JavaScript and a 4.0 MB page. The desktop page scores 82/100 at 2.0 seconds, which suggests the build is the cause. Most property searches start on a phone.

  5. 10

    Tag pages

    Medium
    DaysDeveloper

    Noindex the 215 tag pages.

    What we saw: 215 tag pages exist to support 550 posts. They are thin, near-duplicate pages that may use up crawl budget and add little the AI tools would quote.

Phase 2 · Weeks 2 to 8

The sources the AI tools cite

Get Harbour & Key onto the sites the AI tools already read.

The tools seem to build seller answers from directories, round-ups and forums, and mostly not from agents' own websites.

  1. 11

    Directories and round-ups

    High
    Weeks 2 to 8Ops

    Build profiles on getagent.co.uk, agentseeker.co.uk and yably.co.uk, refresh the allagents.co.uk profile, and get into the “best estate agent Edinburgh” round-ups.

    What we saw: ChatGPT leans on agentseeker.co.uk and yably.co.uk (8 citations each) and allagents (4). Perplexity leans on reddit.com (12) and getagent (4). Claude leans on top10propertyagents.co.uk (12). Harbour & Key's sales volume does not seem to reach these sources, so being named there could help move 4 of 32 toward double digits.

  2. 12

    Area hub pages

    Medium
    Weeks 2 to 8Content

    Build suburb and area hub pages for Edinburgh and Glasgow that last.

    What we saw: 356 of the top 500 ranked keywords come from /property/ listing pages that expire when the property sells, so the visibility does not build up. Area hubs could do the same job and stay.

  3. 13

    Answer capsules

    Medium
    Weeks 2 to 8Content

    Add 40 to 60 word answer capsules to the top 20 guides.

    What we saw: Nothing on the site is set up as a direct answer yet. The 2014 LBTT guide still ranks 27 to 31 for ADS terms at 590 to 720 searches a month. Rewriting it with capsule answers and question-style H2s, and keeping the page, could be the quickest test of the format.

Phase 3 · Weeks 4 to 12

What is already close

Move the searches that are near page one, and keep measuring.

Some keywords already sit on page 2, and a single run of AI answers is noisy, so the monthly trend is what matters.

  1. 14

    Page 2 keywords

    High
    Weeks 4 to 12Content

    Push the 31 page-2 non-listing keywords toward page 1.

    What we saw: 31 keywords already sit at positions 11 to 20, including first time buyer scotland (1,300 a month, #20), scotland first time buyer scheme (720 a month, #16), best areas to live in edinburgh (590 a month, #17) and the home report trio (390 a month, #14 to #17). These look like the closest wins.

  2. 15

    Glasgow coverage

    Medium
    Weeks 4 to 12Content

    Add Glasgow content and service-area settings that match where the firm works.

    What we saw: Harbour & Key covers Glasgow but is not in the top 20 for “estate agents” searched from Glasgow, and the site has little written for that market.

  3. 16

    Monthly re-run

    Medium
    Monthly, ongoingOps

    Re-run the 12-question baseline across the four AI tools and AI Overviews each month.

    What we saw: This audit is a single-run snapshot, and a single run is noisy. The trend over at least 4 monthly runs would show whether the 4 of 32 is moving.

What good looks like

Four goals for 90 days

MeasureNow90-day goal
Maps searches with Harbour & Key in the top 20 (of 5 tracked)0 of 52 of 5
Seller AI answers naming Harbour & Key (of 32)4 of 3210+
Informational searches (positions 27 to 32)Page 3Page 1
Mobile main-content load16.7 sUnder 4 s
Before the work starts

Three decisions to make now

The plan needs three calls before Phase 0 can start
  1. 1
    Approve the Phase 0 technical sprint

    Schema, H1s, titles and meta descriptions, the broken-link redirects and the firewall check are days of developer time, and everything else builds on them.

  2. 2
    Commit to the Profile and review work

    The 662 reviews at 4.6 are the biggest asset in this audit, and they do almost no work in the Maps pack today. Fixing that is largely an ops commitment and not a rebuild.

  3. 3
    Green-light the directory and round-up work

    The AI tools already cite agentseeker.co.uk, yably.co.uk, allagents, reddit.com and top10propertyagents.co.uk on seller questions. Being properly represented there could be the fastest route from 4 of 32 toward double digits.

The 20-minute walkthrough call is where these three decisions could become a scheduled Phase 0.

Method and limits

Every baseline figure on this page comes from the audit, measured on 8 July 2026. AI answers vary between runs, so a single run is a snapshot.

Real client work. Names and locations changed.