On this page
What it actually is, at this size
Strip the acronym away and a CRM in a trade business is three things bolted together.
- A list. Every enquiry, from every source, in one place, with a name, a number, what they want and where it came from. That alone, before any automation, is more than most firms have.
- A set of columns. Where each enquiry has got to. New, quoted, waiting, won, lost. This is the pipeline, and it exists so that nothing can be simultaneously nobody's job and everybody's assumption.
- A set of rules. Things that happen on their own when something else happens. A call rings out, so a text goes. A quote is marked sent, so a follow-up is scheduled. A job is completed, so a review is requested.
That is the whole idea. What makes it valuable in the trades specifically is that the work is physical and the phone is not. An office-based business loses enquiries through carelessness; a trade business loses them through being up a ladder, under a van, in a loft or in a customer's kitchen with both hands occupied. No amount of discipline fixes that, because the constraint is physical. A rule that fires without anybody being available is the only thing that does.
The second reason it matters here is that trade sales cycles are lumpy. An emergency callout closes in ten minutes. An extension quote closes in eleven weeks. A firm holding both in a phone and a notebook will always favour the fast one, which is why builders and installers so reliably report that their five-figure quotes “went quiet”. They did not go quiet. Nobody spoke to them.
The three automations that pay for everything
You can build twenty. Three of them account for most of the return, and a firm that installs only these three and nothing else has captured the large majority of the available benefit.
One: the missed call gets a text
An inbound call that is not answered triggers an immediate text message to the caller. The alternative is voicemail, and voicemail in 2026 is a way of politely declining work: a substantial share of people will not leave one, and a caller who has your number on a list of three simply dials the second. Because the mechanics of setting this up properly are fiddly — and because doing it badly is worse than not doing it — we have written the whole thing out separately in the missed-call text-back guide.
Two: quotes get chased on a schedule
Every trade firm sends quotes and every trade firm intends to chase them. Almost none do it consistently, because chasing happens in the evening, after a day on site, in competition with invoicing and a family. So the system does it: a short message a couple of days after the quote checking the figure made sense, another a week or so later carrying something useful rather than a nudge, and a final one further out with a date attached.
The interval is where most firms get this wrong, and it must be set by the decision length of the job rather than by a template. A boiler quote is decided in a week; a loft conversion in two months. Running a seven-day sequence against a sixty-day decision means you have chased somebody three times before they have finished talking to their architect, and then gone silent for exactly the period in which they actually chose.
Three: reviews get requested automatically
The most under-built automation in the trades, because everybody knows they should ask and nobody remembers on the day. A job is marked complete; a message goes a few days later with a direct link to the review page, not to your homepage. That is it.
The argument for bothering is not vanity. BrightLocal's Local Consumer Review Survey 2026, based on a US panel of 1,002 consumers, reported that 97% read reviews for local businesses and that 68% would only use a business rated four stars or above. Apply the appropriate caution about a US sample, but the mechanism holds here: a steady flow of recent reviews changes what happens in local search results and in the moment somebody compares three firms. The difference between asking sometimes and asking every time is usually the difference between a handful of reviews and several dozen a year, and it costs nobody any effort once it is built. The wider approach sits in the reviews guide.
Pipeline stages, and why five beats fifteen
The commonest way a CRM build fails is being too clever. Somebody maps every nuance of the business into stages, and within a month nobody is moving cards because it takes four decisions to work out which column something belongs in.
Start with five: new enquiry, quoted, waiting on the customer, won, lost. Most trade firms can run on that for a year. Add a stage only when you can name a decision it will change. A survey-booked stage earns its place if you want to know how many surveys convert. A not-yet stage earns its place if you have customers waiting on planning permission or a season. A stage called “in progress” earns nothing, because it does not tell you what to do next.
Two rules keep a pipeline alive. Every card must have an owner and a next action with a date on it, and anything with neither is not a pipeline entry, it is a note. And lost must be used honestly, with a reason attached: too expensive, went elsewhere, never responded, out of area, job cancelled. After six months that field is the most useful management information in the business, because “never responded” being the largest category means something entirely different from “too expensive” being the largest.
One place enquiries land
A typical trade firm receives work through the office phone, two mobiles, a website form, Google Business Profile messages, Facebook, WhatsApp, and a supplier who passes leads on. Seven routes, no shared view, and no way of knowing whether any individual enquiry was ever answered.
Consolidation is unglamorous and it is usually the highest-value part of the whole build. Website forms post directly in. Google Business Profile and Facebook messages route in. Calls to the mobile are captured through forwarding or a tracked number. The person who answers the phone has one screen, and anything not on it does not exist. The test of whether it has worked is simple: can you answer, without asking anybody, what happened to every enquiry received last Tuesday?
Two practical notes on getting there. Forwarding the mobile everybody actually rings is usually the hardest part politically rather than technically, because it means the owner giving up the idea that they personally hold the important numbers. And WhatsApp is the awkward one: much trade enquiry traffic now arrives there, it does not consolidate neatly, and the workable compromise in most firms is a rule that anything arriving on WhatsApp gets logged as an enquiry within the hour by whoever received it. Not elegant, but a rule people follow beats an integration nobody maintains.
Knowing which marketing produced the work
Once every enquiry lands in one place with a source attached, a second benefit appears that has nothing to do with follow-up. You can finally tell which marketing produces jobs rather than which produces noise.
Most firms are guessing, and usually guessing in favour of whatever is most visible. Paid search is visible because there is an invoice. Word of mouth is invisible because there is not. Recording the source on every enquiry, and then recording the outcome, lets you calculate cost per job by channel rather than cost per enquiry, and those two numbers frequently rank the channels in different orders. A channel producing cheap enquiries that never close is worse than an expensive one that converts, and no advertising platform will ever tell you that, because it cannot see your pipeline. This is the same data that makes a search account improvable rather than merely adjustable, and the definitions are in the glossary.
Getting a team to use it
Systems in trade businesses fail on adoption, not on features. Four things decide it.
- It must be faster than what they do now. If logging an enquiry takes longer than writing it on a hand, the hand wins. Every field you add is a tax charged to the person least able to pay it.
- It must work on a phone, one-handed, outdoors, in gloves. Not on a desktop in the office. If it does not, it does not exist to the people doing the work.
- One person must own it. Usually whoever answers the phone most. Not the owner, who will be on site, and not everybody, which means nobody.
- The first version must be embarrassingly small. Five stages, three automations, one inbox. Add nothing for a month. Systems that launch complete get abandoned; systems that launch trivially small get extended.
When you do not need one yet
There are firms we tell to leave this alone, and it is worth being specific about which.
If you are a sole trader taking fifteen enquiries a month, answering most of them yourself and quoting the same week, a CRM will formalise something already working and add administration you have to do in the evening. A notebook and a recurring reminder to chase quotes on a Friday genuinely covers it. Revisit when you take on a second van or a person answering the phone, because that is the point at which information stops living in one head.
If you are turning work away, follow-up automation makes your problem worse by producing more enquiries you cannot service. The right move there is pricing or capacity, and we would rather say so than sell a build.
And if you already pay for job management software that includes a pipeline and messaging, switch on what you are already buying and measure it for two months before adding anything. A surprising number of firms run a second system alongside features they are already paying for and never enabled. If after that the gaps are real — enquiries from social channels never landing anywhere, missed calls going nowhere, quotes chased once — then a proper build is worth doing, and the sector detail is on the industry pages. Tell us how you currently handle a call you cannot answer and we will usually know within a minute whether you need this yet.
Rather have someone else do this?We run exactly this for trade and construction firms across the UK. Call and we will tell you what is realistic in your area, including when the answer is that it is not worth it.
Call 07443 392243WhatsAppCommon questions
Which CRM should a small trade firm use?
Less important than people think, provided it does four things: captures enquiries from every channel, shows a pipeline on a phone, sends text and email automatically from rules, and records the source of every enquiry. Plenty of platforms do that. The build — the stages, the rules, the wording — matters far more than the logo on the login screen, and a badly configured expensive platform loses to a well configured cheap one.
How long does it take before it is worth anything?
The missed-call automation returns value in the first week because it recovers callers immediately. The pipeline takes a month to be trusted and about three months to produce management information worth acting on, because you need enough closed enquiries for the lost-reason field to mean something. Review automation compounds slowly and is the one people wish they had started earlier.
Will automated messages make us look impersonal?
Only if they read like marketing. Two plain sentences that sound like the person who would have sent them, with no sign-off and no enthusiasm, read as organised. Long, cheerful, obviously templated messages read as a system. The wording is the part worth spending time on and the part most builds skip.
Do we still need to chase people manually?
For high-value quotes, yes, and the system should be making that easier rather than replacing it. Automation handles the persistence — the messages at day ten and day forty that nobody would have sent — and a human handles the conversation when somebody responds. The mistake is treating automation as a substitute for talking to a customer about a sixty thousand pound job.
What about data protection?
Keep the enquiry record and the marketing list mentally separate. Replying to somebody who contacted you about a job is ordinary business correspondence. Adding them to a newsletter afterwards is marketing and needs a proper basis and a record of it, plus an unsubscribe that works. Keep the list clean, honour opt-outs immediately, and do not import an old spreadsheet of names nobody can account for.