Pro free for 30 days.No card.Start free →

Estimating Project Time and Cost Without Guessing: A Practical Method

Guessing is the most expensive way to quote a job. Learn why estimates run low, how to use your own past projects as a reference, break work into small tasks, turn hours into cost, add a realistic buffer and give clients a range with stated assumptions.

By WorkBOS Content Desk6 min read
Calculator, pen and papers on a desk for budget planning (Photo by stevepb on Pixabay)
Photo: stevepb / Pixabay

Every small business owner has quoted a job, watched it take twice as long, and quietly absorbed the loss. Guessing feels fast, but it is the most expensive way to estimate. A guess that is too low costs you margin. A guess that is too high loses the job.

This guide shows a practical way to estimate project time and cost using your own past work, a few simple formulas and a deliberate buffer. It is not a substitute for accounting advice, and the numbers in the examples are illustrations, not benchmarks for your industry.

Why estimates come out too low

A UK government publication on optimism bias describes it as a built-in tendency to expect better results than similar past work justifies. It says this leads people to underestimate costs, timescales and risks, and that it affects experts and non-experts alike. The same document points out that overruns on large public projects have stayed high for decades, which suggests a systematic habit rather than bad luck.

The inside view trap

The report says planners often take an "inside view": they look at the specific pieces of this project instead of how comparable projects actually turned out. For a small business that sounds like "this website is only ten pages, so two weeks". The outside view asks a different question: how long did our last three ten-page websites really take, start to finish, including client delays?

Use your own history as the reference

The same publication recommends a method called reference class forecasting. Its steps, simplified for a small business:

  1. Pick similar past projects. The report talks about samples of 20 to 30 for large programmes. You will have fewer, so use every comparable job you have, even five or six.
  2. Compare estimate with outcome. For each one, note the original estimate and the real hours and cost.
  3. Find the typical overrun. If your jobs usually run 30% over, that is information, not failure.
  4. Add that uplift to a clean base estimate. Do not double count: remove any buffer already in the base number before adding the uplift.

The report gives an illustration: on a large base cost, a 10% contingency left only about 35% confidence of staying within budget. A flat "add 10%" rule can give false comfort when your real history says otherwise.

If you have no history yet

Start recording today. Log estimated and actual time on every job for a few months. Until you have data, use a wider buffer and tell clients the number is a range.

Break the job into small pieces

Big tasks hide surprises. Small tasks expose them.

Bottom-up estimating, step by step

  • List every deliverable and break each into tasks of roughly half a day to three days.
  • Include the invisible work: meetings, revisions, waiting for feedback, project admin, handover.
  • Estimate each task as a range (best case, likely, worst case).
  • Add the likely values for the base, then look at the worst-case total to see your exposure.
  • Ask the person who will actually do the work, not just the person selling it.

Duke University's project guidance suggests reviewing the schedule and risks with the whole team to confirm that time and resources are sufficient. That review is a cheap reality check on any estimate.

Turn hours into cost

Time is only part of the cost. A simple structure keeps you honest:

| Cost line | How to work it out | Example | |---|---|---| | Labor | Hours x fully loaded hourly cost | 60 hours x $45 = $2,700 | | Direct expenses | Software, stock images, travel, subcontractors | $400 | | Overhead share | Your monthly overhead divided by billable hours | 60 x $8 = $480 | | Buffer | Your historical overrun applied to the above | 25% of $3,580 = $895 | | Profit | Your target margin on top | Set before you quote |

"Fully loaded" means pay plus payroll taxes and benefits, not just the wage. Check the numbers with your accountant, because tax and cost treatment depends on your situation. For a wider decision framework, see our cost-benefit check before you hire, buy or expand.

Give a range and name your assumptions

A single number sounds precise and is almost always wrong. A better quote has three parts:

  • A range: "$6,000 to $7,500".
  • Assumptions: "assumes content is supplied by March 10 and two rounds of revisions".
  • What changes the price: extra pages, extra revisions, late feedback.

Assumptions turn disagreements into simple questions. When one proves false, you have a documented reason to revisit the price, which links directly to our guide on managing scope creep with clients.

Re-estimate as you learn

An estimate is a forecast, not a promise carved in stone. The Los Angeles public works project manual, for example, calls for updating estimates at set stages as design develops rather than relying on the first number forever. Do the same: re-estimate at each milestone, compare to the budget, and tell the client early if the range has moved.

A weekly check

  • Hours spent versus hours planned.
  • Work finished versus work remaining.
  • Forecast to finish (spent plus honest estimate of remaining).
  • Any assumption that has changed.

Common mistakes

  • Estimating from memory. Memory flatters. Use logged hours.
  • Forgetting non-billable work. Meetings, email and revisions are real time.
  • Padding silently. Hidden buffers distort pricing; state a range and your assumptions instead.
  • Estimating alone. The person doing the work usually knows where the traps are.
  • Never comparing estimate to actual. Without the comparison, you repeat the same error forever.

Tracking estimates against actuals in WorkBOS

The only way to improve estimates is to compare them to reality. WorkBOS lets you set up projects with tasks and budgets, then view project budgets against actuals as work and expenses come in, so each finished job becomes a data point for the next quote. When you are ready to bill, invoicing and basic accounting sit in the same workspace. The documentation explains project setup and the pricing page lists plans; the Free plan includes up to 5 users. This is general information, not financial advice.

Frequently asked questions

Q: How much buffer should I add to an estimate? A: Use your own history. If similar jobs usually overrun by 25%, add roughly that to a clean base. Without data, start higher and refine as you record results.

Q: Should I give clients a single price or a range? A: A range with stated assumptions is usually more honest and easier to defend. Fixed prices work best for well-understood, repeatable work.

Q: What is the difference between an estimate and a quote? A: An estimate is your best forecast and can change as facts emerge. A quote is a price you commit to. Say clearly which one you are giving.

Q: How do I estimate something I have never done before? A: Break it into smaller known parts, ask someone with experience, add a wider range, and consider a small paid discovery phase before committing to a full price.

Q: How often should I update an estimate? A: At every milestone, and whenever an assumption changes. Share the updated forecast with the client promptly.

Sources are linked in the text above: GOV.UK (Homes England optimism bias publication), Duke University Bass Connections, City of Los Angeles Project Delivery Manual.

Run it all from one place

Projects, clients, finance, people and AI employees in one system. Start free, no card needed.

Start free →