Estimate Versus Actual: The Number Most Removals Firms Never Measure

The problem with learning from experience

Experience is supposed to make you a better estimator. In practice it often does not, because the feedback loop is broken.

A surveyor estimates a job. The job happens, sometimes weeks later, run by a different crew. If it overran, the operations team absorbed it. The surveyor may never hear about it, and if they do it is as an anecdote rather than a number.

So the estimating error persists. Not through carelessness, but because nothing ever told the estimator they were wrong, and by how much, and in which direction.

Twenty years of that produces twenty years of the same error, held with increasing confidence.

What to measure

Volume. Estimated cubic feet or metres against what was actually moved. This is the master number, because everything else follows from it.

Time. Estimated hours against actual hours on site. Track loading and unloading separately if you can, because they fail differently.

Crew. Planned crew against crew used. An extra body sent at short notice is a cost and a signal.

Vehicles and loads. Planned against actual. A second trip is the most expensive form of estimating error.

Materials. Estimated against consumed. Usually a small number, consistently underestimated.

Total cost. The sum of all of the above against the quoted price, which gives you the actual margin against the expected margin.

The patterns that appear

Once you have a few months of data, the same handful of patterns show up in almost every firm.

Variance by surveyor. Different surveyors have different consistent biases. One runs ten per cent under on volume. Another prices access generously. Neither knows it.

Variance by property type. Flats routinely take longer than the model allows because of stairs, lifts and carry distance. Large detached houses often have more volume than the bedroom count suggests, because of lofts, garages and outbuildings.

Variance by job type. Storage jobs frequently overrun because handling happens twice. Part-load and shared jobs are commonly underestimated.

Variance by customer type. Corporate jobs have more waiting time. Elderly clients often need more support than the survey anticipated.

Variance by season. Peak season jobs overrun more, because crews are tired, vehicles are stretched and there is no slack in the day.

None of these are surprising once stated. All of them are invisible without measurement.

What to do with the findings

Adjust the model, not the person. If a surveyor is consistently ten per cent under, the useful response is to show them the number, not to criticise the judgement. Most estimating bias corrects itself immediately once it is visible.

Build the patterns into your rates. If flats consistently take fifteen per cent longer, price them that way rather than absorbing it.

Use it in the survey. A surveyor who knows that lofts are the most commonly missed space will check the loft. Our survey guide covers the method in full.

Watch it over time. The point is the trend. A firm whose variance narrows quarter by quarter is getting better at the single most important commercial skill in this trade.

What good looks like

There is no universal benchmark, because it depends on job mix and how you measure.

What matters is your own trend. Variance that narrows is a business learning. Variance that stays flat is a business repeating itself.

The firms that do this well end up quoting more confidently, not less, because they know what their numbers actually are rather than hoping.

Why almost nobody does it

Three reasons.

The data sits in different places. The estimate is in the quote, the actual is in the job sheet or the crew's memory, and joining them is manual work nobody has time for.

There is no obvious moment to do it. Nothing in the working week says "compare last month's estimates to last month's actuals."

And it can be uncomfortable. It is a measurement of your own accuracy, and the first month's findings are usually worse than expected.

The firms that push through that are the ones whose pricing improves. It is also one of the five questions that distinguishes a CRM from a system that actually runs your business.

Frequently asked questions

What is estimate versus actual in removals?

The comparison between what a job was estimated to require and what it actually consumed: volume, hours, crew, vehicles and materials. The gap between the two is variance, and tracking it is how estimating accuracy improves.

How do you measure removals quoting accuracy?

Compare the surveyed volume against the volume actually moved, and the estimated hours against the hours actually worked, on every job. Report it by surveyor, by property type and over time. The trend matters more than any single job.

What is a good variance figure for removals estimates?

There is no universal benchmark because it depends on job mix and measurement method. The meaningful measure is your own trend over time. Variance that narrows quarter on quarter indicates the estimating is improving.

Why do removals estimates go wrong so consistently?

Because the feedback loop is broken. The surveyor rarely learns what the job actually took, so the same error repeats indefinitely. It is a structural problem rather than a competence one.

Should I share variance data with my surveyors?

Yes, framed as information rather than criticism. Most estimating bias corrects itself immediately once someone can see it. Presented as a performance failing it produces defensiveness rather than improvement.

Sources and further reading

• Moovi guide to running a removals survey
• Moovi guide to pricing a removals job