Most contractors know when they’ve badly underquoted a job. It’s that nagging feeling halfway through the project — the materials cost more than expected, the job’s taken longer, the margin you thought you had has disappeared.
What fewer contractors track is the cumulative effect across a full year’s work.
Two kinds of bad estimate
There are two ways to price a job wrong, and they have opposite consequences.
Underestimating means you win the job but lose money on it. The materials cost more than you charged. Your labour was priced too tight. You forgot to include items. By the time you’re done, you’ve worked for less than your labour rate.
Overestimating means you don’t get the job — not because you were too expensive in an absolute sense, but because your quote came in higher than someone who priced it more accurately.
Both are problems. But they’re driven by different causes.
Why estimates go wrong
There are three main failure modes in manual estimating.
1. Scope gaps
A loft conversion that doesn’t include RSJ costs, or a kitchen refit that misses the waste disposal charges, or a bathroom where the tile adhesive and grout weren’t included. Scope gaps — items that should have been in the estimate but weren’t — are the most common source of underpriced jobs.
Experienced estimators make fewer scope errors. Newer estimators, or those pricing outside their usual trade, make more. But even experienced estimators, working quickly on high volumes, will miss items periodically.
An AI estimating system, trained on thousands of similar projects, makes far fewer scope errors. It includes the fixings. It includes the waste uplift. It includes the items that only get remembered after the job has started.
2. Stale prices
If you’re using a price book from six months ago, or your own mental list of what materials “usually cost”, your estimates are based on prices that may no longer exist.
UK construction material prices have been volatile since 2021. Timber, steel, insulation, and plasterboard have all seen significant swings. A price that was accurate in March might make your job unprofitable by June.
Live pricing — where every material is queried against actual merchant prices on the day of the estimate — eliminates this problem entirely.
3. Margin erosion
Even when scope and prices are right, margin can erode through rounding, through applying the wrong rate to the wrong item, or through simply forgetting to add overhead and profit to every line.
A 15% margin accidentally applied to only the materials (not the labour) might drop your effective margin to 8%. Over ten jobs, that’s a significant reduction in what you actually take home.
The maths of small errors
Here’s a worked example.
You’re pricing 40 jobs per year. Your average job value is £12,000.
Scenario A — Underpricing by 3% on average
A 3% pricing error sounds small. But on a £12,000 job, it’s £360. Across 40 jobs, that’s £14,400 per year of revenue you should have earned but didn’t.
Scenario B — Overpricing by 5% on average
If your estimates consistently run 5% above the market, you’ll lose a proportion of jobs to competitors. Lose 1 in 4 competitive quotes and that’s 10 fewer jobs per year at £12,000 average — £120,000 of work you didn’t win.
The sweet spot is accuracy. Estimates that reflect actual costs, at competitive market rates, using current prices.
What accuracy actually means
When PricingPro achieves a 96% accuracy rate, it means that across a sample of real estimates, the actual cost of the job came within 4% of the AI-generated figure.
That’s 4% variance — some jobs slightly over, some slightly under. It doesn’t mean every estimate is perfect. It means the systematic errors — scope gaps, stale prices, margin miscalculations — are largely eliminated.
For comparison, industry research suggests that manual estimates from experienced estimators, using up-to-date price books, typically run 5–15% variance depending on job type and complexity.
What you can do right now
If you’re still pricing manually, the first step is awareness. For every job you complete, compare your estimate to actual cost. Most contractors who do this honestly find they’re undercharging on a meaningful proportion of jobs — not because they’re bad at their trade, but because manual estimating has too many points of failure.
The second step is to modernise the process. Moving to live pricing and AI-assisted scope identification addresses the two biggest sources of error immediately.
The return on investment is usually immediate. Improving accuracy by 3% across 40 jobs per year recovers over £14,000 in revenue — more than enough to justify any subscription cost.