By a quarter to seven most mornings, the yard behind Harrowgate Pike's head office on the edge of Wakefield is emptying. There are similar scenes at depots in Sheffield and Warrington. Forty-four vans head out between them, most still carrying the red flame painted on the first one in 2011, when the business was two engineers servicing extinguishers for pubs and corner shops.

The company has grown up fast since. Turnover has gone from £4.8m to £11.6m in four years. It employs 95 people, more than half of them engineers, and nearly two-thirds of its revenue now comes from housing associations, care groups, healthcare and education estates: customers with dozens of buildings each. In the past year two private equity-backed consolidators have been in touch about buying it.

Very little of that shows from the outside. The website, last properly rebuilt in 2017, still opens with "your friendly local fire safety team". The bid team builds tenders in Word, from four different templates. And the procurement teams who decide million-pound contracts, who read the paperwork almost as closely as the price, have started to notice.

Four from fourteen

The figure that stayed with me wasn't the turnover. In the past year Harrowgate Pike reached the final stage of fourteen tenders and won four. Three of the ten it lost came back with the same note: presentation and perceived scale. Contracts worth between £250,000 and £1m a year were going to firms that looked the part, even though Harrowgate Pike did the work perfectly well.

Fourteen final-stage tenders.

Lay the paperwork side by side and the reason is obvious. A quote in one typeface, an inspection report in another, a tender in a third. Clients asked for "a proper portal" 31 times last year. What they meant was paperwork that looked as serious as the work.

That part was quick to fix. Quotes, reports, tenders and invoices now come from one set of templates, in one typeface, under one name. The portal clients kept asking for became a report pack, sent the morning after every visit and summarised each quarter. It looks as serious as the work, and there is nothing to log in to.

Two companies under one roof

Part of the problem was a deal that never quite finished. Two years ago Harrowgate Pike bought Pennine Alarm Services, a Sheffield firm with a good book of alarm contracts. Pennine's customers found out about the sale from a new invoice, and two contracts went to tender early as a result. Thirteen of the 44 vans still carry Pennine's green and grey. In Sheffield half the paperwork still carries the Pennine name, and the engineers there are still on Pennine's WhatsApp groups.

To an estates director responsible for eighty buildings, that reads as two companies, or as one that hasn't quite got itself organised. It also left money on the table. Pennine's clients had never been offered a fire door inspection, and most of Harrowgate Pike's own customers had no idea the business now did alarms. Two firms were sharing an invoice and very little else.

So the job was to finish the deal rather than rebrand around it. Within a month Sheffield was on the same quotation tool, reports and email signatures as everyone else, and its engineers were in the same depot huddles. Pennine's alarm clients were offered a fire door inspection at their next service, by the engineer they already knew. Harrowgate Pike's door clients heard, from their own engineers, that the firm now looked after alarms too. The vans and the paperwork follow, Pennine first.

Three right answers

Dean Harrowgate and Michelle Pike started the business together and still own it jointly. Everyone at the top knew something was off; they disagreed about what. The managing director could see the brand costing them in tenders, and could see what a buyer would make of it too. But he worried about losing the honest local firm that won their first customers, who still bring in more than a third of revenue. The technical director wanted to know why money should go on branding while engineers waited on new tablets. The head of bids had a team of three building submissions by hand and a tender due in six weeks.

All three were right. That's usually the hardest kind of brief.

So the brief became three promises: keep the local firm, spend nothing on the brand that doesn't also help the engineers, and get the next tender out on time in the new design. The tablets were ordered in the first fortnight, loaded with the new inspection report and the quotation tool. The tender went out first, and on time.

The 6%

Before anyone discusses a logo, it helps to know how big the prize is. So the first piece of homework was sizing the market: every service the firm sells, against every sector it could sell it into.

Counted from the ground up, fire safety servicing across Yorkshire, the East Midlands and the North West comes to roughly £190m a year. That covers businesses with premises to look after, 4,558 care homes and several hundred thousand housing association homes, each with a typical annual spend. Counted from the top down, by scaling the UK market to the three regions, it comes to about £210m. Harrowgate Pike's £11.6m is about 6% of that.

Fire safety servicing across three regions.

That is a serious regional business by any measure, and nobody looking at its paperwork would guess it. The national groups, many of them backed by private equity and buying regional firms as they go, bid for the same contracts with considerably better-dressed submissions. The gap between what the company is and how it presents is costing it work, and it is also costing it value.

The second piece of homework was search. Around 18,000 people a month across the UK look for the services the firm sells; scaled to its three regions, that's roughly 4,400. Fire doors lead, helped by regulations that came into force in 2023. Almost nobody types a town: "fire safety Wakefield" barely registers. You bid on the service and let a map do the geography, a radius around each depot. Search doesn't win tenders, but it keeps the local book growing, and it gives a procurement manager somewhere credible to land.

Together the two numbers set the plan. The growth wasn't in a new market; it was in winning more of the one the firm already served. That meant bigger contracts through tenders, more services sold to the clients it already had, and search aimed at services rather than towns: a page for each service, and ads drawn around each depot.

What people search for.

Every touchpoint

The last piece of homework was a map. It traced every point at which a customer meets the company, from the first search to the renewal letter. There are 25 of them. Marketing runs perhaps six. The rest belong to engineers, the office, finance and the founders: the van in the car park, the engineer at reception, the inspection report, the invoice, the email signature at the bottom of a reply about a missed visit.

That was the most useful thing the map showed. A brand isn't the logo on the van so much as the sum of those 25 moments, most of them handled by people who'd never describe their job as marketing. So each of the 25 now has an owner and a standard: what the van looks like, how the phone is answered, what the report says, who signs the email. Most of what follows is the work of getting there.

Local, redefined

Which brings us back to the managing director's worry. "Local" had done a great deal for this business, and nobody wanted to throw it away. The trouble was what it had come to mean: friendly, nearby, small. To a procurement manager scoring five bidders against a matrix, friendly is a nice-to-have and small is a risk.

So we kept the word and changed what it promises. Local now means the same engineers, who know your buildings, every visit: accountable rather than cosy, across three regions. It holds for the shop on the corner and for a care group with sixty homes, and it is exactly what a multi-site buyer is trying to score. The line that came out of it is four words long: Local roots. Regional standard.

The name changes too, if only slightly. Harrowgate & Pike becomes Harrowgate Pike. Two surnames joined by an "and" is how a lot of fire safety firms start out, and the name still read that way: two people and a van. Losing the ampersand turns a partnership into a company, and both founders keep their names on the door.

The 2011 flame retires with it. The new mark is drawn from a fire door hinge: the hinge axis sets the baseline, the leaf sets the height of the letters, and the slant is the leaf swinging open. It reads as clearly on a tender cover or a tablet screen as it does on a van door.

The new mark.

Pennine is folded in rather than erased. For a year, Sheffield documents and vans read "Pennine Alarm Services, now part of Harrowgate Pike", and the engineers those customers already know tell them about the change in person.

The sequence matters as much as the words. The next tender goes first, because it's the cheapest thing to change and the quickest way to find out whether any of this wins work. The vans change one door panel at a time, the Pennine ones first, as each comes in for its service. Nobody loses a van for a day, and nobody loses an engineer for one either. A small fleet tool keeps track of which van is due which panel; it now also holds service dates, MOTs and engineer assignments for all 44, which is what the technical director had wanted in the first place.

A queue of one

Every business has one: the person everything waits for. At Harrowgate Pike it was the estimator, one person pricing the remedial work found by 58 engineers and installers. An engineer would find six failed doors in a care home on a Tuesday. On average, the client heard what fixing them would cost eleven working days later, if the quote hadn't slipped down the pile. In a market decided on responsiveness, the bottleneck wasn't the engineers or the price. It was one inbox.

So the first tool we built wasn't for customers at all. It's an internal quotation tool, and anyone can use it: an engineer in a car park, an apprentice, the office, the bid team between meetings. They enter the client, the building, the number of doors and the length of contract, and the tool does the rest. Prices come from a locked rate card. The inspection schedule follows the regulations for that type of building. The scope wording and terms are the same every time. A quote that used to wait for one person now takes a few minutes from whoever is nearest, and every one reads as though the same careful person wrote it. The estimator, freed from the queue, now spends the time on the large and complicated jobs.

The second tool removes even the typing. Engineers record a voice note or a short video on site and send it to one address. It's transcribed, each job is matched to the rate card, and a priced draft lands with the office, who check it and press send. The AI reads; it never sets a price.

I'm under no illusion that a firm like this will get the most out of tools like these by itself. Most don't. So the deal is that I do the heavy lifting: I build the tools, keep them running and change the rate card when prices move. The team's part is what they already do: talk, take a photo, forward an email. The operations director owns the rollout, each depot has a champion, and the apprentices go first.

Tenders, and only tenders

Once a strategy is agreed there's always a temptation to do everything: a blog, a podcast, a content calendar. Harrowgate Pike doesn't win work that way. Its contracts arrive as tenders, decided by a handful of people with a scoring sheet. So we resisted the content treadmill and put the effort into the route that pays.

That meant the tender library first: reusable sections for each sector, in one design, that the bid team can assemble in a day rather than a week. Then the parts of a bid nobody had treated as marketing. There's a social value statement covering the six apprentices, the local hiring and the routes planned to keep vans off the road, which housing and public-sector buyers score and the firm had never claimed. There's registration on the frameworks through which many of the larger contracts are actually let. And there's proper preparation for the final-stage interview, which is where ten of last year's fourteen were lost.

Every tender also needs pictures of engineers, vans and buildings, and time on site is the one thing this business can't spare. Its engineers are spread across three regions, booked weeks ahead, and paid to inspect doors rather than pose beside them. A day's photo shoot means a day of inspections someone else has to cover. So the imagery is generated with AI, built from reference photographs of the firm's own depots, vans and sites, and checked by the people who work in them. A new picture for a sector-specific bid now takes an afternoon rather than a booking, and nobody is pulled off a job for it.

After the handshake

Winning the tender is where marketing's part of the story usually ends, and where the money in a contract actually starts. A care group that signs for fire door inspections also has extinguishers that need servicing and alarms that need testing. It has doors that fail and need fixing, and more homes than the contract first covered. None of that gets sold by a campaign. It gets sold by whoever is in the building.

At Harrowgate Pike that's mostly the engineers: 52 of them, plus six installers, on site four days in five. So the inspection report on their tablets gained a few fields. If a door fails, the engineer photographs it, notes the defect and taps once to flag it for a quote. That lands straight in the quotation tool. A small bonus goes to the engineer when the job converts. It is, the managing director pointed out, the biggest sales team the company has ever had.

Everyone else has a part too. The bid team keeps a one-page plan for each contract showing which sites are covered, which services are held and when it renews. Each depot's service manager takes a quarterly report pack into a review meeting, which is where a missing service line comes up naturally. The founders keep one call a quarter with their twenty biggest clients. And every depot spends five minutes of its weekly meeting on what was found, what was quoted and what was won.

The rest of the estate

Winning one care home is rarely the prize. The prize is the estates manager behind it, and the seventeen other homes they look after. So the bid team now has a map of each client's estate. Every site comes from public registers, with a mark against each one Harrowgate Pike already services and for what. Each regional manager is listed with how many of their buildings are covered. The tool ends in a short ranked list of what to offer next. It starts with alarms at a home where the engineers already do the doors. Then come the homes run by a manager who already trusts them. New faces are last.

For one care group, that turned three homes into a list of eighteen and showed roughly £96,000 a year still to play for. None of it needs a cold call. It needs someone to ask the right person at the next quarterly review.

Who picks up

In a business this size, most of what a client thinks of the brand is decided by people who would never call themselves marketers. So everyone was given a part and the words to go with it. An engineer finishing a visit says who they are, what they checked, what failed and who will send the price, by name, within two days.

The office answers by name with the caller's site already on screen, because the phone system recognises the number. Known clients skip the menu and go straight to their depot. Out of hours, faults reach the on-call engineer. A missed call is answered within a minute by a text with a link to book or get a quote. Even the hold message does a job: twenty seconds on the fire door rules and the fact that anyone can now get a quote in two days.

The apprentices

Harrowgate Pike was never going to run a content calendar, and it doesn't need one. Its buyers check LinkedIn before a shortlist, and that's about it. So the company page does a modest job, posting contract wins, accreditations and the odd team photo from a set of templates. The founders and the sales team had their profiles rewritten in the new voice, because those are the pages a procurement manager actually clicks on.

The real voice belongs to the six apprentices. They're on site every day and they're comfortable on camera. What they're learning, how a door check works, what a failed closer looks like, the van at seven in the morning: that is the most convincing material the company has. Their posts double as evidence for the social value section of every tender: local people, trained properly. Their phones also supply the raw material for the company's imagery. A quick photo of a corridor, a plant room or a van at dawn, taken between jobs, is all the reference a generated picture needs.

They work to a one-page set of rules. No residents, patients or pupils on camera. No alarm panels, plans or keys. No client names without permission. It's taught on every new starter's first day, alongside setting up their LinkedIn profile, and revisited for five minutes at each depot's weekly catch-up, where the apprentices pick the moment worth sharing.

One job each

Mention marketing to most people and they think of content: posts, videos, a steady stream of activity. A lot of what businesses do online is busy rather than useful, and every hour of it costs something. So the rule here is that every channel has one job and doesn't stray from it.

Tender portals and frameworks find the big contracts. Search catches local demand. The website turns a visit into an enquiry. LinkedIn shows a buyer that the firm is real and regional. The vans and the engineers earn trust on site. Email grows and renews the clients the company already has. Every route ends in one system, so nothing sits in an inbox. Instagram, TikTok and the rest get nothing, because none of Harrowgate Pike's buyers are deciding anything there.

The website was the quickest part of the build, because by then there was very little left to decide. We knew who lands on it: an estates manager checking a bidder before a shortlist, or a facilities manager pricing fire doors for a handful of sites. We knew what each of them needs to find, and what they should do next. So the site is small on purpose: a page for each service and each sector, the accreditations, a quick way to reach the right depot, and a fire door cost calculator, because fire doors are what people search for most. Enter the doors, the type of building and the number of sites, and it returns an annual cost and an inspection schedule. The visitor gets an answer; the firm gets a properly qualified enquiry, in the same system as everything else.

The system of record

We didn't build a CRM, and I'd be wary of anyone who offered to. Firms like this already run job management software that does the hard parts: sites, door registers, schedules, invoices. The job was to make everything else talk to it. Quotes raised in the quotation tool land against the right client and site. Calculator estimates arrive as leads. Engineers' flags from the tablet become opportunities. Each night a plain export feeds the leadership dashboard and the estate map, so nobody rekeys a figure into a spreadsheet again.

That's also what makes the email work. Nothing goes out on a calendar. Messages go out because something happened. A booked visit sends a confirmation with the engineer's name and photograph, and a finished one sends the report pack the next morning. A failed door produces a priced quote within two days. If the quote is ignored, a polite follow-up arrives with the photograph of the door. Renewals are flagged at ninety days, and a regulation change goes only to the sites it affects. There's a ceiling of two non-service emails a month, and nothing promotional goes to a client with an open complaint. Every email is signed by a person the client has met.

The Monday meeting

Most founder-led businesses are run by conversation, and it works until it doesn't. At 95 people and three depots, Harrowgate Pike still ran on a two-person rhythm. There was a Monday catch-up most weeks, if the founders were in. Tender decisions often came the day before the deadline. A management meeting was pencilled in roughly monthly and cancelled roughly as often. Eleven WhatsApp groups did the rest. The only time all three depots were in one room was the Christmas party.

None of that was anyone's fault. It's simply what happens when a company doubles in size and nobody redesigns the week. So we redesigned it. Each depot now has a ten-minute huddle at half past seven. The leadership team meets for an hour on Monday around a single page of numbers, drawn from the systems rather than someone's spreadsheet. On Tuesday the bid team decides which tenders to chase while there's still time to chase them properly. Each depot ends the week with a half-hour catch-up. Once a month the whole company meets, hosted by a different depot each time, and the founders take one question from the floor. Every meeting has an owner, an agenda and a finishing time, and none ends without someone's name against an action.

The page everyone opens

The hardest audience in any field business is your own team. More than half of Harrowgate Pike's staff are on the road four days in five, and an all-staff email is the least read document in the building. So rather than ask people to look somewhere new, we put the news where they already look.

Every company laptop and tablet now opens on the same page: a plain search box, the thing people were going to open anyway, with sticky notes stuck over the top of it. To search, you have to move the notes out of the way, and while you're doing that you've read them. Today's priority sits in red: the inspection app goes offline at eight tonight. There's a note for each depot, the week's meetings, the quote-turnaround figure, a contract won, and quick links to the quotation tool and the client estate map. It works with a thumb on a tablet in a van, and there's nothing to log in to.

One of the notes is written by a machine. A regulation-watch tool reads the rule-makers every week: the Building Safety Regulator, the British Standards Institution, the certification bodies, the fire services. It turns any change into plain English, and the quality lead checks it before it goes anywhere. When the alarm standard was revised to cover cybersecurity on remotely accessed panels, the engineers had a note on the homepage and a five-minute talk at the Wednesday toolbox session. The clients with affected systems had an email explaining what it meant for their buildings. A change in the rules is, after all, a perfectly good reason to call a client.

Ready, not for sale

Nobody at Harrowgate Pike is planning to sell. But the sector is being bought up quickly. In the first half of 2026, private equity was behind 58% of deals in UK facilities and property services, and almost half of its bolt-on purchases were fire installation and maintenance businesses. One consolidator alone had bought 25 firms by the summer. Two have already called the managing director.

What those buyers pay more for is not mysterious. They want recurring revenue that is measured, not guessed at. They want contracts on standard terms, and systems one person doesn't hold in their head. Above all they want a business that runs without its founders. Advisers put firms of this size somewhere between 5.5 and 7 times profit. On Harrowgate Pike's numbers, the gap between the bottom and the top of that range is worth around £2m, before profit grows at all.

There's a lot of talk about AI premiums, and I'd rather be plain about it. A fire safety firm doesn't become a software company by using good software, and there's no published figure for what AI adds to a business like this. What the tools do is remove the things buyers mark down. The quotation tool means pricing no longer sits with one person. The job management system and the report packs mean twelve months of clean data on renewals and services per contract. The single brand means one integration story instead of two half-merged firms. None of it is done to sell the business, but all of it would count if they ever did.

Buying well

The other half of that conversation is that Harrowgate Pike may well do the buying. That's how a regional firm becomes a larger one, and it's how much of this sector has grown. The Pennine deal showed what happens when the people side of an acquisition is left to sort itself out.

So the next one comes with a playbook. On day one the founders visit, the staff get a letter, and every customer hears the news from the engineer they already know. By day 30 the new team is on the same quotation tool, reports and email signatures, and the cross-sell tracker has its customers in it. By day 100 the vans carry "now part of Harrowgate Pike" at their next service. After a year the old name retires, unless the data says it's worth keeping. The targets are chosen as carefully as the markets were sized: firms of £1m to £4m with mostly recurring work, proper accreditation and an owner ready to step back, inside an hour of a depot or worth opening a new one for.

The marketing part of a deal is the part that's easiest to forget: who is told what, in what order, and what the business looks like on the Monday after. Get that right and an acquisition adds value. Get it wrong and you end up with a second brand on the road for years.

What we're watching

Good marketing doesn't change what a business is. Harrowgate Pike was already a regional operation, with a capable team and fifteen years of inspection records behind it. The job is making sure the people signing the contracts can see that, without the shops on the corner feeling left behind.

So the measures are plain. The final-stage win rate, four in fourteen when we started. How often tender feedback mentions presentation. How long a remedial quote takes, which was eleven working days. How many of the three services each contract customer holds. The enquiry form, finished by 1.9% of the people who start it. And whether the long-standing local customers stay. If the new brand wins a care group and loses the corner shops, it has failed.