Ask a small heating, plumbing or electrical firm how jobs get booked and most owners describe the same system without ever having called it a system: a customer messages a photo of the broken boiler on WhatsApp, someone quotes a price in the same thread, the job goes in a paper diary or a shared calendar nobody else can see properly from a van, and the invoice gets typed up — from memory, from the WhatsApp thread, from whatever's left on the job sheet — on a Sunday evening once the week has calmed down.
It works, in the sense that jobs get done and most invoices eventually go out. What it isn't, is free. It's three separate, unconnected records of the same piece of work — a chat message, a diary entry, an invoice — and somebody has to hold all three in their head long enough to turn them into one bill. That's not a quirky inefficiency. It's a specific, measurable chunk of the week, and UK trade bodies have actually measured it.
What the survey data actually says
The NOW Report 2026 — independent research commissioned by Powered Now and Installer Magazine, fieldwork run in April and May 2026 — put the figure at 277 hours a year lost to business admin per UK tradesperson: quoting, invoicing and chasing payment. That's roughly five and a quarter hours a week, per person, and it's a national average across firms with every level of tooling, from fully integrated job management software down to a paper diary and a biro.
The firms sitting at the paper-and-WhatsApp end of that range aren't paying the average. They're paying more, because the 277-hour figure is a straight admin-time number, and their process adds a step the survey doesn't separately count: every job gets re-keyed, not once, but typically three times — once as a WhatsApp quote, once as a diary booking, once as an invoice line — with no link between any of them, so an error or a forgotten extra at any step has to be caught by memory rather than by a record.
What that actually costs a small team
Take a realistic five-strong team: an owner-manager and four engineers, each one personally fielding enquiries, confirming jobs and chasing their own invoices through WhatsApp and a diary rather than a shared system. Applying the NOW Report's 277 hours a year across all five gives 1,385 hours a year spent on quoting, invoicing and chasing payment — admin that happens almost entirely in the evening or at the weekend, because none of it was designed to fit inside the working day.
Illustrative figures for a five-engineer heating and plumbing firm: 1,385 hours a year of quoting, invoicing and payment-chasing admin, at a loaded cost of £24 an hour for that time — roughly £33,240 a year, before anything goes wrong.
Add the re-keying error that three unlinked records almost guarantees: a forgotten extra, a misquoted price copied across from the wrong WhatsApp message, or a job that never gets invoiced at all because nobody wrote it down properly. Say that happens once a fortnight at an average £45 of work left unbilled or undercharged — another roughly £2,340 a year.
Call it around £35,500 a year, for a process that's never once appeared as a line item, because nobody bought it — it's just how the diary has always worked.
Why three unlinked records is harder than one slow one
A single slow system is at least consistent — the same record gets updated by the same people in the same place, even if it's clunky. WhatsApp-plus-diary-plus-invoice isn't that. Each one is a genuinely different medium with its own failure mode: a quote buried forty messages down a chat thread with no search worth trusting, a diary page that's illegible, lost, or simply not visible to whoever picks up the phone for a reschedule, and an invoice that only gets written once someone has reconstructed the other two from memory.
That's also why "chasing payment" shows up inside the same admin figure as quoting and invoicing rather than as a separate problem. An invoice typed up five or six days after the job, in a weekend batch, goes out five or six days later than it needed to — which means the clock on getting paid starts five or six days later too, and the eventual chase is partly a bill the firm raised against itself by how late the invoice went out.
It's the same pattern described in our piece on why most integrations break: nothing here fails loudly. No single step looks broken on the day it happens. It just compounds, quietly, across a year, until someone finally sits down and adds up the hours.
What the off-the-shelf options actually cost
This is not a build-it-yourself problem, and it's worth saying so plainly: a team this size has no business commissioning custom software when the job is "one shared record per job, read by everyone who needs it." Several UK-built tools already do exactly that for trade businesses, and their published pricing is the clearest part of this whole case.
Tradify's current UK pricing runs three tiers per user per month, billed monthly with no fixed contract: Lite at £34 covers job scheduling, quoting and invoicing and online payments; Pro at £37 adds custom branding, job photos, recurring jobs, appointment reminders and progress invoicing; Plus at £44 adds AI-assisted bill reading and reporting on top. For a five-person team on the Pro tier, that's £185 a month — £2,220 a year — against the roughly £35,500 the same firm is plausibly losing to paper-and-WhatsApp admin now. The software costs about 6% of what it would save.
Commusoft sits at a different point on the market: it doesn't publish pricing, typically starts around £59 a month and commonly runs £40-£70 per user once modules are added, and it's generally pitched at firms with 20 or more staff running fleets or maintenance contracts, usually on a 12-month contract rather than month-to-month. For a five-engineer firm it's almost certainly the wrong size of tool — more contract than the business needs for a problem a simpler, cheaper system already solves.
Where automation genuinely helps, and where it doesn't
The part worth automating is mechanical: turning an enquiry into one structured quote record instead of a chat message, letting a completed job sheet become the invoice without anyone retyping it, and sending that invoice the same day the job finishes rather than in a weekend batch — which is also most of what actually shortens the payment-chasing side of the admin figure, because the invoice simply goes out sooner.
What shouldn't be automated, and what none of these tools try to automate, is the judgement: how to price a genuinely non-standard job, whether to chase an overdue invoice hard or let a long-standing good customer run a few days over, which quote needs a phone call rather than a text back. The software's job is to make sure those are the only decisions left for a person to make — the same shift our piece on the real cost of doing it manually for now describes for any recurring task.
A reasonable first step
Before switching anything, track the real numbers for two weeks rather than trusting a national average: how long each job actually takes to go from enquiry to quote to invoice to paid, and where in that chain the delay or the re-keying actually happens. Most firms assume it's the invoicing step. It's just as often the quote sitting unconfirmed in a chat thread for three days, which is a different fix.
Then run a trial on live jobs rather than switching everything over at once — Tradify's is 14 days with no card required. Keep the paper diary running in parallel for a fortnight and compare: not which system feels nicer, but which one actually gets an invoice out the same day the job finishes, because that's the number that was costing the business money in the first place.
Common questions
We're a sole trader, not a five-person team — does the £35,500 figure scale down proportionately?
Roughly, yes. The NOW Report's 277 hours a year is a per-tradesperson figure, so a sole trader doing their own quoting, invoicing and payment-chasing is looking at around £6,650 a year in time alone at the same £24-an-hour loaded cost, plus whatever re-keying errors creep in from running quotes through WhatsApp and invoices through something else. That's still comfortably more than a single-user job management subscription costs for a year, which is usually £400-£530 depending on tier — the case doesn't disappear at smaller scale, it just gets smaller in absolute terms.
Isn't WhatsApp fine for quoting if we're just careful and well-organised about it?
Care doesn't fix the structural problem, which is that WhatsApp was built for conversation, not for records. Even a meticulous user can't easily search a year-old thread for what was actually agreed, filter messages by job, or hand a colleague read access to just one customer's history without handing over everything else in the chat too. The issue isn't discipline, it's that the tool has no concept of "a job" as a distinct, searchable, shareable thing — and a dedicated system does.
How do we actually choose between Tradify, Commusoft and everything else on the market?
Start with team size and contract appetite rather than feature lists, because those two facts eliminate most of the shortlist quickly. Tradify's month-to-month, per-user pricing and 14-day trial suit a small team that wants to test it on real jobs before committing. Commusoft's unpublished, module-based pricing and 12-month contracts suit a larger operation with fleets or maintenance contracts, where the cost of getting the fit wrong is spread across enough users to justify a longer sales process. A five-person firm belongs in the first category almost by default.
This site usually talks about custom builds — does that apply here too?
No, and it's worth saying why not. A custom build earns its keep when nothing on the market reconciles a genuinely unusual setup — several incompatible systems, or a process no off-the-shelf tool models properly, the kind of case our piece on build vs buy for a multi-branch vet group covers. A five-engineer trade firm booking jobs on WhatsApp has the opposite problem: a well-understood process that several mature, cheap, UK-built products already solve. Buying is not the compromise option here — it's just obviously correct.