Flat rate vs time and materials: a decision tree by trade and job type

Flat rate vs time and materials is usually argued as a philosophy. It's really a question about risk: when a job runs long, who pays for the extra hour? With flat rate pricing, you do. With time and materials (T&M), the customer does. Everything else, from close rates to tech pay to how angry people get at the invoice, follows from that one transfer.

So the useful question isn't "which is better?" It's "for this job type, can I carry the risk profitably?" This post gives you a decision tree to answer that by trade and job type, a worked simulation that shows when a flat price is safe, and the research on why customers will pay a little more for a number they know in advance.

What each model really is

Flat rate pricing sets the price of a task before work starts. Replace a capacitor: one price, whether it takes 25 minutes or 70. The price comes from your hourly selling rate times the expected hours, plus materials at your markup. Our glossary entry on flat-rate pricing has the short version.

Time and materials bills the actual hours at a stated rate, plus materials at cost or with a markup. The customer approves the rate, not the total. See time and materials.

The federal government, which buys more services on T&M terms than anyone, describes the core problem with it plainly in the Federal Acquisition Regulation:

That's why federal contracting officers may use T&M only when no other contract type is suitable, and only with a ceiling price the contractor exceeds at its own risk¹. Homeowners have the same instinct without reading the FAR. When the meter is running, every minute the tech spends walking to the truck feels like it's on their bill.

Why customers prefer a price they know

There's solid research on this, mostly from telecom and subscription pricing, where companies offer both flat and pay-per-use plans and can see what people choose.

Anja Lambrecht and Bernd Skiera studied this in the Journal of Marketing Research and found that many users choose a flat rate even when a pay-per-use tariff would have cost them less, a pattern they call the flat-rate bias². They traced it to three effects:

• The insurance effect. People pick the flat price to avoid variation in what they'll pay.
• The taxi meter effect. People enjoy what they're buying more when they aren't watching a meter tick.
• The overestimation effect. People overestimate how much they'll use.

Their most useful finding for a contractor: the bias didn't significantly increase churn, so the companies earned more both in the short and long run.

The taxi meter idea goes back to Drazen Prelec and George Loewenstein's work on the "pain of paying," which argues that a ticking meter reduces the pleasure of the ride itself³. Earlier still, Kenneth Train, Daniel McFadden and Moshe Ben-Akiva found in local telephone service that when per-call charges rose, enough households moved to flat-rate service that revenue barely changed⁴.

None of these studies is about furnaces. But the mechanism carries over cleanly: a homeowner with a broken AC is buying certainty as much as a repair. A known price removes one of the two things they're anxious about.

Who carries the risk: a worked simulation

Here is the part most pricing articles skip. A flat rate is a bet that the hours average out. Whether that bet is safe depends on two things: how much the hours vary for that task, and how many times a year you do it.

First, a selling rate built from public numbers. The Bureau of Labor Statistics puts the mean wage for HVAC mechanics and installers at $31.14 an hour (May 2025)⁵. In June 2026, wages were 68.6% of total employer compensation cost for private-industry installation, maintenance and repair workers, with benefits the remaining 31.4%⁶. Divide wage by that share and a tech who earns the mean costs about $45 per hour paid.

Now the example, with every assumption labelled. Say a tech is paid 2,080 hours and 60% are billable, overhead is $60,000 a year per tech, and you want a 20% net margin. Labor per billable hour is about $76, overhead adds about $48, so the cost of a billable hour is about $124 and the selling rate is about $155. The labor rate calculator does this with your numbers.

Take a task that averages 1.5 hours. Priced at exactly 1.5 hours, the labor portion is $232.50. Some jobs take 1.1 hours and some take 2.3. On any single job, you'll sometimes lose money. The question is how often a whole batch of those jobs loses money.

We simulated two kinds of task, both averaging 1.5 hours. A steady task, like a capacitor swap or a water heater element, where hours vary modestly. And a wild task, like chasing an intermittent fault or opening a wall in an old house, where the median job is quick but one in ten takes three times as long.

Chance that the batch loses money. Worked example: lognormal job hours, 20,000 simulated batches per cell, $155 selling rate against $124 cost per billable hour. Not industry data.

Read the table by row. On a single steady job, you lose money one time in five. Over 20 of them, almost never. That's the law of large numbers doing your risk management, and it's the whole reason flat rate works for common repairs: the long jobs and short jobs cancel out across a month.

Now read the wild columns. Even after 20 jobs, you still lose money on the batch about one time in eight at average pricing. A modest padding (pricing at 1.75 hours) helps, but for a task you do five times a year, you're gambling.

Two rules fall out of this:

1. Flat-rate a task only when you do it often enough that the hours average out within a month or a quarter. Twenty a year is a reasonable floor for a steady task.
2. For wild tasks, either price for the spread or don't flat-rate them. "Priced for the spread" means a higher flat price with stated exclusions, or a flat diagnosis followed by a flat repair once the scope is known.

The decision tree

Use these four questions, in order, for each job type in your price book.

1. 1

   Can you define the finished scope before work starts?

   If the tech can say exactly what will be done and what 'done' looks like, keep going. If scope depends on what's behind a wall, under a slab or inside a 40-year-old panel, go to T&M with a not-to-exceed, or split the job into a flat diagnosis and a flat repair.

2. 2

   Do you do this task at least 20 times a year?

   If yes, you have enough volume for the hours to average out, and enough history to know the real average. If no, you're guessing at both. Price it as a project quote or T&M.

3. 3

   How wide is the spread of hours?

   Pull actual hours for the last 20 or more. If the slowest jobs take less than about twice the median, flat rate it at a little above average. If the slowest take three times the median or more, add exclusions and stop points, or keep it T&M.

4. 4

   Who controls the scope once you start?

   Homeowners usually approve one fix. Commercial customers, property managers and general contractors often add work mid-job. When the customer controls the scope, T&M against a purchase order and a not-to-exceed protects both sides.

By trade and job type

Here's how that tree usually lands. Treat it as a starting point and test it against your own hours.

The HVAC pricing guide and plumbing pricing guide walk through building the flat-rate side task by task.

When time and materials is the honest choice

Owners who switch to flat rate sometimes swing too far and flat-rate everything. That's when it bites. T&M is the better choice when:

• The scope genuinely can't be known. You can't price the repair behind a wall you haven't opened. Quoting a padded flat rate anyway means good customers subsidize the bad surprises.
• The customer will change the scope. A property manager adding three units to the work order mid-day is easier to handle with a rate than with three new estimates.
• The contract calls for it. Many commercial agreements set labor rates per hour and markups on materials. Our post on your first commercial service agreement covers how to write those.
• The job is rare and wild at the same time. If you do it twice a year and the hours swing threefold, you don't have the volume to absorb a bad one.

Making T&M feel less like a meter

If you bill T&M to homeowners, borrow the best part of flat rate: certainty.

• Give a not-to-exceed. "We bill actual time at $155 an hour plus parts, and we won't go over $1,200 without calling you first." The customer still knows their worst case.
• Set a check-in point. "After two hours I'll show you what I've found before going further."
• Itemize the invoice in plain language. Hours by activity, not one lump.
• Get changes in writing. A signed change order before extra work starts protects you if the bill is disputed. Some states make written home improvement contracts and change orders a legal requirement above a dollar threshold, so check yours.

The hybrid most shops end up with

In practice, the strongest shops run three models side by side:

1. Flat rate for the 50 to 150 tasks that make up most of their service tickets.
2. A flat diagnostic fee that buys a defined scope of troubleshooting, followed by a flat repair price once the problem is known.
3. T&M with a not-to-exceed, or a project quote, for hidden-condition and commercial work.

The hybrid also fixes the pay question. If you pay techs a share of flat-rate task value, a fast tech earns more per hour, which is the point. If you pay T&M hours, you're back to paying for time. Whichever you choose, make sure the pay plan follows the pricing model rather than fighting it.

When you present a flat-rate repair, offering options is where the certainty pays twice. Our post on good, better, best estimates shows the math. In TradesWork, techs build each option from the price book on the phone and the customer signs the one they choose, so the price they approved is the price on the invoice.

Keeping flat rates honest over time

A flat-rate book is only as good as its last update. Three habits keep it from drifting:

The decision isn't permanent, and it isn't one decision. It's a small decision per job type, made with your own hours, and revisited when the numbers move.