How do you work out what estimating costs your firm?
Estimating is quiet, senior work. A request comes in, a partner or associate director opens a spreadsheet from a job that felt similar, adjusts the lines, argues with themselves about contingency and sends it. For a firm that quotes often, that adds up to a significant slice of the most expensive time in the building, and a lot of it goes on enquiries that never convert.
There is a second cost that rarely gets counted: the margin lost on work that was underpriced because the estimate was built from memory rather than from what similar jobs actually took.
To see the first cost, count the estimates your firm prepares in a typical month, multiply by the senior hours each takes, and convert to a weekly figure. Multiply that by the estimators' charge-out rate and by 46 working weeks. The unbillable hours calculator runs it for you. The second cost you can test by comparing a handful of past estimates with the hours recorded against those jobs. The audit will also tell you if estimating is where your hours go first.
Why does every estimate start from a blank sheet?
The knowledge needed to estimate well is already in the firm. It sits in the time recorded against past jobs, in the variations that followed, in the write-offs at billing and in the heads of the people who ran the work. It just is not in a form an estimator can use at speed.
So each estimate is rebuilt from scratch, or from a copy of one estimate that may not have been right either. Two partners pricing the same job can arrive at very different numbers, and neither can easily show why. The work of learning from past jobs is done in theory at every project review, and then again, badly, every time someone quotes.
What does rebuilt fee estimating look like in practice?
The estimator describes the job: type, size, client, stages, known risks. The process then does the groundwork.
- It finds comparable past jobs from your own records and sets out what they were estimated at and what they actually took, stage by stage.
- It drafts a priced breakdown using your current rates and grades, your standard assumptions and your usual allowances.
- It lists the exclusions and assumptions that normally apply to that kind of work, so they are stated, not forgotten.
- It flags the lines where past jobs most often ran over, so the estimator can decide whether to allow for it this time.
The estimator reviews the reasoning, changes what their judgement says should change and approves the fee. The approved estimate and its assumptions then feed the scope document and the proposal, so the price and the promise match.
The loop matters as much as the draft. When a job completes, its actual hours are compared with its estimate and stored, so every finished job makes the next estimate better.
What will estimators notice first?
Speed, then confidence. A first draft appears in minutes, and it comes with the evidence behind it. Estimators spend their time on the questions that deserve it: is this client likely to change the brief, is the brief unusually loose, is this a job we want at a thinner margin. Junior staff can prepare drafts for partners to review, which spreads a skill that has usually lived with a few people.
Consistency improves as well. When two directors price similar work, they start from the same history, and differences become deliberate choices rather than accidents.
Which firms benefit most from faster estimates?
Firms that quote frequently, where each quote takes senior time and margins depend on getting it right. That describes engineering consultancies pricing design and advisory stages, architecture practices working to RIBA stages, surveying practices quoting for surveys and valuations, and environmental consultancies pricing assessments and monitoring. Accountancy firms quoting for non-recurring work face the same problem in a different vocabulary.
What does a 30 day estimating rebuild include?
Mapping starts with your estimators, not your spreadsheets: we watch several estimates being prepared, gather the templates, rate cards and assumptions they rely on, and test how much usable history your time records hold. Building then takes place on that history and those rate cards, inside the spreadsheet or system your estimators already use, and every draft is checked against real enquiries arriving during those weeks. Estimators are trained in the fourth week and new enquiries are priced through the rebuilt process from then on. Support runs for 30 days after go-live, and during that time the first completed jobs start to feed back into the history.