Estimating job duration so schedules don't slip

Why schedules slip

Most late afternoons start with a bad estimate in the morning. A repair booked as an hour takes two, the next customer gets a late text, and by 3 p.m. the last job of the day is rescheduled. The tech is blamed, but the real problem is the number the office used when it booked the call.

Offices usually estimate from memory or from the price book. Neither is a time standard. Memory remembers the easy jobs, and price book hours are often set to support a price, not to predict a day. The fix is to measure how long each type of job actually takes in your company, then plan with those numbers.

What a time standard is

A good standard has three parts: the job type, the planning time and the conditions that change it. It is not a target for the tech. It is a forecast for the dispatcher. Treat it that way in conversations with the team, or techs will learn to work to the clock rather than to the job.

The timestamps you need

You cannot build standards without clean timestamps. Most field service software records them if techs tap the status buttons. Make those taps part of the job, and check the data weekly for missing or obviously wrong entries.

Building standards from your own data

1. 01

   Pick your job types

   Use six to twelve types that drive most of your volume. Too many and you will not have enough data for each; too few and the averages hide real differences.

2. 02

   Pull 90 days of jobs

   Export arrival and departure times by job type. Drop jobs with missing taps and jobs that were cancelled on site.

3. 03

   Calculate the median and 80th percentile

   The median is the middle job. The 80th percentile is the time four out of five jobs finish within. Spreadsheets do both in one formula.

4. 04

   Choose a planning time

   For predictable work such as tune-ups and recurring stops, plan with the median. For diagnostics and repairs that often run long, plan with the 80th percentile.

5. 05

   List the modifiers

   Ask senior techs what makes a job run long: attic units, crawl spaces, older equipment, two-story homes, pets, large properties. Add fixed minutes for each.

6. 06

   Publish and review

   Put the standards in the booking screen or a one-page sheet for CSRs. Review every quarter and after any price book change.

Worked example: an HVAC service board

Example (fictional): a mid-size HVAC company pulled 90 days of job data and found its booking times were close for tune-ups but far too short for diagnostics. The new standards are below.

The board now holds fewer diagnostics per tech per day, which felt like a loss at first. In practice the shop stopped rescheduling one or two late-afternoon calls per tech each week, and the office stopped sending apologies. Completed jobs per day stayed flat while on-time arrivals improved, because the plan finally matched reality.

Buffer rules that keep the day on time

Even good standards miss. Buffers absorb the miss without pushing every later job. Write them as rules the dispatcher can apply without judgment calls.

• Add a 15-minute buffer after any diagnostic, because diagnostics turn into repairs.
• Do not book a fixed-time appointment directly after a job planned at its 80th percentile.
• Leave the last 45 minutes of each tech's day unbooked on peak-season days.
• Book jobs with known modifiers (attic, crawl space, older system) earlier in the day.
• After a second consecutive overrun, the dispatcher re-plans the rest of that tech's day immediately.
• Use arrival windows, not exact times, for anything after the first job of the day.

Intake questions that change the estimate

The best time to adjust a standard is on the phone. A few questions at booking can add or remove 30 minutes before the tech ever leaves the shop. Script them into the intake so CSRs ask them every time.

• Where is the equipment? Attic, crawl space, closet, roof or basement.
• Roughly how old is it, and when was it last serviced?
• How large is the home or property, and how many stories?
• Is there easy parking and access, or a gate code, pets or a building manager?
• Has anyone else looked at this problem already?
• Can you send a photo of the model label or the problem area?

Photos are the biggest help. A clear shot of the model label or the damage lets the office pick the right job type and lets the tech bring the right parts, which also protects the time standard from parts runs.

Different trades, different units

Not every business should measure in minutes per job. Lawn, pest and cleaning companies get better forecasts from minutes per size band: a cleaning standard might be set per bedroom and bathroom count, a mowing standard per thousand square feet of turf, a pest service per linear foot of perimeter or per property tier. Measure the first visit separately, because initial cleans and initial treatments usually take far longer than the recurring service that follows.

Install work needs a different unit again. Roofing, HVAC changeouts and garage door installs are best planned in crew-days with named phases, such as tear-off, dry-in and finish, rather than minutes. Build those standards from completed projects by size and complexity, and keep them on the install calendar, not the service board. Electrical and handyman companies with very mixed work often do best with a short list of time blocks (one hour, two hours, half day) chosen at intake from a few scoping questions.

What an overrun costs

Labor is the largest cost of a missed estimate. In June 2026 private employers paid an average of $46.89 per hour worked in total compensation, with $32.82 in wages and $14.07 in benefits^[1]. For a single trade, the median HVAC mechanic earned $29.33 an hour in May 2025 before any employer costs^[3]. Trade wages are climbing: average hourly earnings for production workers at specialty trade contractors reached $39.05 in September 2026, up 4.7% on a year earlier^[4].

An overrun also costs the next customer's goodwill and sometimes the job itself. That is why the standard should err on the long side for unpredictable work. A tech who finishes early can pick up a short job or restock. A tech who finishes late pushes the problem onto three more households.

Do not turn standards into a stopwatch

There is a trap here. If time standards become targets, techs will cut corners to meet them, and the work gets worse. ENERGY STAR says improper installation can reduce HVAC system efficiency by up to 30%, and sealing and insulating ducts can improve efficiency by as much as 20%^[2]. Rushed work that skips those steps creates callbacks and unhappy customers. Coach on quality and process; use standards only to plan the board.

Share the data with the team. When techs see that the office now books diagnostics at 105 minutes instead of 60 because of their own job history, they trust the board and are more likely to tap their timestamps accurately. That feedback loop is what keeps standards honest. The next step is to use those standards in capacity planning and in your same-day service rules.