Fire extinguisher, alarm, emergency lighting and fire door servicing for businesses across the North West. 85 staff.
Verdict: 8 searches (5 to improve; 1 new; 2 left) bring 1,011 visits a year now. Allowing for the chance of reaching each target, the work should add 1,164 visits and 14 enquiries a year, about 4 new customers: £4,755 of gross profit once reached (£1,610 on first contracts alone). It takes 50 hours of writing and 26 of technical fixes (£4,580): a first-year net of -£1,014 (-£2,797 if half as many targets are reached), then £2,115 a year after upkeep. Worth most: “fire extinguisher servicing manchester”; best for the money: “fire extinguisher servicing manchester”. Technical: 5 failing; 1 not known.
| Search | Position now → target | Action | Extra enquiries a year | Gross profit a year | Cost | First-year net |
|---|---|---|---|---|---|---|
| fire extinguisher servicing manchester 590 searches a month · ready to buy · Extinguisher servicing | 8 → 3 best rival 2 | Improve page | 6.0 | £2,032 | £440 | £1,084 |
| emergency lighting testing 480 searches a month · ready to buy · Emergency lighting | 11 → 5 best rival 3 | Improve page | 2.4 | £816 | £330 | £282 |
| fire door inspection 720 searches a month · ready to buy · Fire doors | 22 → 8 best rival 6 | Improve page | 1.9 | £661 | £950 | −£454 |
| fire alarm maintenance contract 320 searches a month · ready to buy · Fire alarms | 14 → 5 best rival 4 | Improve page | 1.8 | £603 | £550 | −£98 |
| fire risk assessment cost 880 searches a month · comparing options · New: fire risk assessment prices | – → 5 best rival 3 | New page | 1.2 | £404 | £660 | −£357 |
| how often should fire extinguishers be serviced 1,300 searches a month · looking for information · Extinguisher servicing guide | 6 → 3 best rival 1 | Improve page | 0.7 | £239 | £220 | −£41 |
| fire safety training for staff 1,000 searches a month · comparing options | – → – best rival 7 | Leave | 0.0 | – | – | – |
| co2 fire extinguisher 2,400 searches a month · looking for information | 35 → – best rival 4 | Leave | 0.0 | – | – | – |
“Fire extinguisher servicing manchester” is worth most: people ready to buy, and Mardale is already eighth. The information search brings far more visits but almost no enquiries. The new fire-risk-assessment page and the fire-door jump from 22nd to 8th don’t pay in the first year: do the three best searches first and judge the rest after six months. Fire-safety training isn’t something Mardale sells; the CO2 extinguisher search is people wanting to buy one online: both left. The agency’s “double the traffic” isn’t far off in visits (about 1,160 more on 1,010), but that’s about four new customers a year.
| Check | Status | Fix |
|---|---|---|
| Every page you want found is indexed (Search Console: Pages report) | Pass | – |
| No important page blocked by robots.txt or a noindex tag | Fail | 2 h |
| Pages load fast on a phone (Core Web Vitals: good) | Fail | 12 h |
| Works properly on a phone | Pass | – |
| Secure (https) everywhere, with no mixed content | Pass | – |
| No broken internal links or redirect chains | Fail | 4 h |
| Each page has its own title and description | Fail | 3 h |
| One page per topic: no near-duplicates competing | Fail | 3 h |
| A sitemap submitted and up to date | Pass | – |
| Business name, address and phone the same everywhere | Not known | 2 h |
Slow pages on phones and near-duplicate service pages hold every ranking back: fix them before the rewrites.
Searches and positions come from Search Console and a keyword tool; the share who click at each position from the Clicks sheet. Each target counts only at the chance of reaching it (50% improving a page, 30% a new one), building up over 6 months. Enquiries come at the site’s 1.5% per visit, 1.5 times that for searches by people ready to buy, 0.5 comparing, 0.1 looking for information; 30% become customers, each worth the first order × margin + 50% of the later work. Local map results, which decide many ‘near me’ searches, aren’t modelled (see the Local Search and Reviews Plan). Searches can change; competitors move too.
Rankings take months and aren’t in your control. Check positions monthly in Search Console; if a target hasn’t moved in six months, stop spending on it.
Fictional, continuing the example. Mardale chose the guided route and went ahead with all six searches, the three best first. Work on the pages started on 2 November 2026. The agent was set up in an AI assistant the business already uses, connected read-only to Search Console and the CRM (its IT confirmed this before it committed). On the second working day of each month it reads the position of each of the six searches being worked on, the visits and enquiries from search and the Pages report, enters them in a new dated copy of the plan, runs the checks and sends the managing director a short note. It uses the owner's lines on Inputs: a search more than 3 places below where it started, or still more than 3 places short of its target after 6 months; enquiries per visit more than 30% below the plan's 1.5%; any important page not indexed. Nothing left the business, and no decision changed, without one of its own people approving it.
| Date | What the agent flagged | What was decided |
|---|---|---|
| 2 Dec No flag | November read: 1,830 visits and 29 enquiries from search (1.6% a visit, against 1.5% in the plan). No search over a line. | Nothing needed. The managing director read the monthly note. |
| 5 Jan Flag | The new fire risk assessment prices page, published on 14 December, was excluded from search by a noindex tag left over from the test site. | The web developer removed the tag the next day and asked for the page to be indexed. It was indexed by the next run. |
| 2 Feb Flag | Enquiries per visit from search fell to 0.9%, below the 1.05% line: 18 enquiries from 1,960 visits, and none from the website form after 19 January. | The office manager found the form had stopped sending after a plugin update. Fixed the same day; she rang the seven people whose submissions had not arrived. |
| 2 Mar Flag | “Fire door inspection” had slipped from 22nd at the start to 27th, more than 3 places. | The managing director brought forward the fire door checklist and two case studies already in the plan. The target stayed at 8th. |
| 5 Apr No flag | Two searches reached their targets: “fire extinguisher servicing manchester” 3rd and “how often should fire extinguishers be serviced” 2nd. “Fire door inspection” was back to 14th. | Nothing needed. Good news shared with the team. |
| 1 Jun Flag | Six months after work started, “fire alarm maintenance contract” was 9th against a target of 5th, more than 3 places short. | The managing director kept the page and its target, asked the two trade bodies Mardale belongs to for a link, and will look again when the plan is re-run in July. |
At the check-in on 18 December, about day 45 and after the first monthly run, Geoff and the managing director read the first note and the log together. No rule needed changing, and Geoff's upkeep of the set-up ended there. What the agent added. The method is unchanged: same plan, same checks, same decisions. The agent made sure positions and enquiries were read every month, which the plan asks for and busy months skip. Two flags were problems nobody had noticed (a page hidden from search and a broken form), each found within days of the month end; one put a search back on track; one, at six months, asks the question the plan says to ask. The benefit here is time to act; no £ figure is claimed for it. The plan is re-run with the new positions in July.