Almost every service business is underpriced somewhere and doesn't know exactly where. Parts cost more than they did last year. The van costs more to fuel and insure. The technician you hired at one wage is now worth another. Meanwhile the price book — if it exists as a document at all — has the same numbers it had two summers ago, and the gap between what a job costs and what it bills gets absorbed silently by the owner.
The reason owners delay is not laziness. It's that raising prices feels like risking the base: the repeat customers, the property managers, the fleet accounts, the neighbors who refer. That fear is real but it is usually mispriced. In practice, a well-executed increase loses a small number of the least profitable customers and improves the business on nearly every other axis, while a badly executed one — improvised per quote, unexplained, applied unevenly — damages trust with the customers you most wanted to keep.
As of August 2026, the difference between those two outcomes is almost entirely process. This is a playbook: read your own numbers first, change the price book once rather than negotiating job by job, decide deliberately who gets grandfathered, pick the timing, give your technicians a script, and measure close rate afterward so you know what actually happened. It applies across trades — the mechanics are the same whether you run a locksmith van, an HVAC fleet, or a cleaning crew.
Start with your own data, not the market
The instinct is to survey competitors. That's the second step, not the first, because a competitor's price tells you what they charge, not what they earn — and copying a number from a business with a different cost structure is how underpricing spreads through a local market.
The first step is to establish what your jobs actually cost you. That means, per service type:
- Parts consumed at current cost, not the cost in your catalog from last year. Stale part cost is the single most common source of invisible margin loss, and it compounds because markup is a percentage applied to the wrong base. The full treatment is in parts markup and margin.
- Labor hours actually logged, not the hours you quoted. The delta between quoted and logged is where flat-rate pricing quietly turns into charity work.
- The trip — drive time, fuel, and the fact that a technician on the road isn't billing.
- Payment cost, because card processing is a real percentage off every ticket.
Put together, that's job costing, and it is the only defensible basis for an increase. It also tends to surprise owners: the finding is rarely "everything is underpriced by 10%." It's usually that two or three specific services are underwater — often the ones involving long drives, small tickets, or parts that moved sharply in price — while the rest are fine. That precision is worth having, because a targeted increase on four line items causes a fraction of the friction of an across-the-board bump.
You need the data in a form you can read. If your job history lives in a stack of paper and a bank statement, this step is a weekend of reconstruction, which is one of the more concrete arguments in spreadsheets versus field-service software. A system that records parts, labor, and payment against each job produces this report on demand — see field-service reports and KPIs for what to look at.
Only after that do you look outward. Competitor pricing is a sanity check on where you land, not the input that determines it. The SBA's guidance on managing business finances frames the discipline the same way: continuous tracking of income and expenses is what lets an owner make decisions on evidence rather than instinct.
Change the price book once, not per quote
The most damaging version of a price increase is the one that never gets written down. The owner decides rates should be higher, mentions it to the crew, and from then on each quote is a small improvisation — a little more when the customer seems comfortable, the old number when they hesitate.
This fails in three ways at once. Customers compare notes and discover inconsistency. Technicians have no defensible answer when challenged, so they discount to end the discomfort. And you can't measure the effect of a change you never actually defined.
The alternative is a single maintained price book that every estimate draws from, updated once, on a date, as a deliberate act. If you don't have one, build it before you raise anything — the structure is laid out in the service price book guide. A real price book covers:
- Each service as a named line with a set price or a defined rate structure.
- Parts priced by rule from current cost rather than typed per quote.
- The trip or service-call charge as its own visible line.
- Any after-hours, emergency, or distance modifiers, defined rather than ad hoc.
Once that exists, an increase is a data change, not a behavioral change. Every quote issued after the effective date carries the new number automatically, every technician quotes the same figure, and the change is auditable — you know exactly what changed and when, which is what makes the measurement step later actually possible.
This is also the moment to consider whether a straight increase is even the right instrument. Offering a good-better-best structure often raises the average ticket more than a flat percentage would, because it moves the customer's decision from "yes or no at this price" to "which of these," and a meaningful share choose the middle or top option. That mechanic is worth understanding before you commit — see good-better-best option pricing. Similarly, whether you're on flat-rate or hourly changes how an increase lands, which is covered in flat rate versus hourly pricing.
Decide who gets grandfathered, on purpose
Not every customer should be moved on the same day, and the distinction is straightforward.
One-off and new work moves to the new price immediately on the effective date. There's no relationship being disrupted; the customer is simply seeing your current rate.
Recurring and contract customers get notice and a defined transition. A monthly maintenance agreement, a property-management account, a recurring cleaning route — these customers committed based on a number, and honoring it through the current term or a stated notice window costs you comparatively little while preserving the relationship that makes them valuable in the first place. Sixty to ninety days of notice, or the next renewal date, are both defensible.
Nobody gets grandfathered indefinitely. This is the trap. An open-ended exception becomes a permanent second price list that nobody can explain, and eventually you're running two businesses at different margins with the same crew. Give the accommodation an end date at the moment you grant it.
If those recurring accounts are billed through automated invoicing, the mechanics of a staged transition are easy — you're changing a rate on a schedule, not chasing customers individually. See recurring invoices for service businesses for how that's structured.
Time it around your season, not your frustration
The trigger for most price increases is emotional: an owner looks at a job that lost money and decides today is the day. That's the right realization and the wrong timing.
Two windows to avoid. Peak season is the worst time to introduce a variable, because you have no operational slack to absorb pushback, retrain quoting, or handle a customer who wants to talk it through. And immediately after a service failure — a comeback, a missed appointment, a public complaint — links the increase to a problem you caused, which is the one association that converts a routine adjustment into a lost account.
The window that works for most trades is the shoulder between seasons, or the start of a calendar year, with notice given a month ahead. Seasonal businesses in particular should tie the change to the cycle they already run their cash flow on — the reasoning is in seasonal cash flow for service businesses.
Give the crew a script
Your technicians will absorb the entire customer reaction, and if they don't have language, they will improvise — usually by apologizing, occasionally by discounting on the spot to escape the moment.
A workable script has three parts and takes about eight seconds:
- State the number. "That job is $340."
- If asked why it changed, one sentence. "Our rates went up in September — parts and labor both moved this year."
- Return to the work. "Want me to get started?"
What to keep out of it matters as much. No apology — it invites negotiation. No detailed cost breakdown of your business — it's not the customer's problem and it sounds defensive. No "the owner made me." And critically, no authority to discount at the truck. If you want a discretion policy, define it explicitly ("you may waive the trip charge on a same-day rebook") rather than leaving each technician to invent one under pressure.
Run the script once as a five-minute exercise with the crew before the effective date. The first time a technician says a new price out loud should not be in front of a customer.
How the three approaches actually compare
| Improvised, quote by quote | Across-the-board bump | Data-driven price book update | |
|---|---|---|---|
| Basis for the number | Owner's read of each customer | A round percentage on everything | Per-service cost and margin data |
| Consistency | Varies by tech and by day | Uniform | Uniform and documented |
| Customer perception | Arbitrary, invites haggling | Blunt but understandable | Explainable in one sentence |
| Effect on underwater services | Usually still underwater | Fixed, but overshoots healthy ones | Corrected precisely where needed |
| Technician confidence | Low — no defensible answer | Moderate | High — the price is the price |
| Measurable afterward | No | Partly | Yes, by service and close rate |
Measure close rate, not just revenue
An increase is a hypothesis, and revenue alone won't test it. A higher average ticket can hide a falling win rate for a full quarter before the volume drop shows up in the bank.
The number to watch is close rate by service type: quotes issued versus quotes accepted, for 60 to 90 days before the change and 60 to 90 days after. Read it alongside average ticket and total gross margin.
Three outcomes and what each means:
- Close rate flat, ticket up. The increase was overdue. Consider whether you left more on the table.
- Close rate down slightly, margin up. Normal and usually fine — you shed the least profitable end of the demand curve. Check which customers stopped converting before reacting.
- Close rate down sharply on specific services. Those particular line items overshot, or the quotes aren't being followed up. Fix the follow-up before you roll back the price — an unaccepted quote is often a communication failure rather than a pricing one, which is the point of estimate follow-up and win rate.
That last distinction saves a lot of unnecessary retreat. Owners routinely blame a price change for a soft month when the actual cause was quotes that nobody chased. Both cash-flow research and the failure data point in the same direction here: QuickBooks' small-business cash-flow research puts late and unpaid invoices among the most common cash-flow problems reported by owners, and BLS business-employment data shows roughly a fifth of new establishments closing within their first year — attrition that is far more often about cash discipline than about demand.
The bottom line
Raising prices well is an operations problem wearing a psychology costume. The anxiety is about losing customers; the actual work is knowing which services lost margin, changing them once in a document everyone quotes from, deciding deliberately who transitions when, choosing a calm week to do it, giving the crew eight seconds of language, and then watching close rate honestly enough to tell whether it worked.
Businesses that run that sequence tend to find the increase was smaller than it needed to be. Businesses that skip it either never raise prices at all and slowly starve, or raise them arbitrarily and confuse the customers who were most loyal. The base you're afraid of losing is generally more durable than it feels — it just needs the change to look like a decision rather than a mood.
Related reading: Building a service price book · Job costing and true profit per job · Good-better-best option pricing. For a complete machine-readable feature and pricing reference, see our LLM reference page.