Job Mix: The Growth Lever Hiding in Which Work You Say Yes To

Most service businesses try to grow by adding jobs. Fewer look at which jobs they are adding. Two shops with identical revenue can end the year with very different bank balances, because one of them spent the year doing work that pays and the other spent it doing work that merely keeps everyone busy. Job mix — the proportion of your calendar given to each type of work — is a growth lever you can pull without spending another dollar on ads.

Revenue hides what margin reveals

Every service business already tracks revenue by job, because that number falls out of the invoice. Far fewer track what the job actually cost to deliver: labour hours including the drive, materials, the equipment it ties up, the office time spent scheduling and chasing payment, and the share of jobs of that type that come back for a warranty visit. Those costs are not evenly distributed. A small repair with an hour of travel each way can carry less margin than a job billing three times as much.

You do not need a costing system to see this. Take one quarter of completed jobs, group them by the categories you already use on your invoices, and for each group write down the revenue, the total hours booked against it, and any materials cost you can pull. Divide revenue minus direct cost by hours. What you get is margin per hour by job type, and it is usually the first time the picture stops being flattering. There is normally at least one category that everybody assumed was fine and is not, and one that nobody thought about that quietly carries the business.

Two adjustments make the number honest. Count travel time as part of the job, not as overhead — it is the reason short jobs at distance lose money. And attribute callbacks and warranty returns to the original job type rather than treating them as a separate cost bucket. Work that comes back twice as often is more expensive than work that does not, and the invoice never shows it.

What to do with the categories that lose

The instinct after this exercise is to drop the worst category. That is usually the wrong first move, for three reasons. Some low-margin work is a gateway: a small diagnostic call that regularly turns into a larger job is buying you the larger job. Some is a filler that keeps a crew productive during hours you cannot sell anything better into. And some is contractual — a maintenance obligation attached to accounts that pay well overall.

So separate the categories that lose money into three piles. The first is work that is priced wrong: it is fine work, delivered efficiently, and you are simply charging too little for it. Reprice it. The second is work that is delivered wrong: the price is defensible but the job takes longer than it should, or generates callbacks, or requires the most expensive person on the team. Fix the delivery — a checklist, a different tech, a minimum booking window, a rule that these jobs get batched by area rather than scheduled on demand. The third pile is work that is genuinely not yours: it does not lead anywhere, it does not fill gaps, and no amount of pricing or process makes it good. That is the pile you stop selling — quietly, by removing it from the website and the ad account, not by firing customers.

Deliberate demotion is usually better than a hard cut. Raise the price on the weak category rather than banning it. Some customers will pay the new price, which solves the problem, and the rest will go elsewhere, which also solves the problem. You keep the option and you stop losing money on it either way.

Growing the mix you want

Once you know which categories carry the business, marketing gets simpler and cheaper. The ad account should be weighted toward the queries that bring the profitable work, not spread evenly across everything you are technically capable of doing. The pages on your site that describe your best work deserve the depth — photos, pricing ranges, real detail — while the marginal services can live on a single line. Your intake script can offer the good work first when a caller is undecided between options. Existing customers who bought a profitable service once are the most likely buyers of it again, and that list costs nothing to email.

Expect this to take a couple of quarters, not a month. Job mix moves slowly because demand arrives on its own schedule, and because your team has habits about what to quote. It helps to set a single target — the share of booked hours going to your best two categories — and check it monthly rather than trying to manage every line at once.

What to automate, and what stays yours

The measurement automates well. Pulling jobs from your CRM or field-service software, tagging them by category, joining revenue to hours and materials, and producing a margin-per-hour table every month is a report, not a judgement call. So is flagging jobs that ran materially over their estimated hours, and flagging repeat visits to the same address inside a warranty window so callbacks land against the right category. If you are tagging jobs by hand today, that is the piece to hand off first — a model reading job descriptions can categorise them consistently enough to make the report usable, and consistency matters more here than perfection.

What does not automate is the decision about what each category is for. Whether a low-margin service is a loss leader that earns you better work or just a loss is a question about your customers and your market, and the data cannot answer it — the profitable follow-on job is often booked months later under a different job type. Repricing is a judgement about what your market will bear and what your competitors charge. Dropping a service is a decision about identity: what you want to be known for and who you want calling you. Automate the arithmetic so the conversation happens over real numbers. Keep the conversation.