Why Groundworks Firms Are Still Losing Money on Estimates (And How AI Is Fixing It)

Estimating is one of the most time-consuming parts of running a groundworks or civils business. It requires experience, knowledge of current rates, and hours of work. Often for a job you do not even win.

Most firms accept this as normal. They sit with an estimator who has 20 years of knowledge in his head, a spreadsheet built years ago, and a process that has not changed much since.

The problem is not a lack of skill. It is that the process does not scale. When your best estimator is on holiday, or when three enquiries land at once, quality drops and turnaround slows. You either rush quotes and win the wrong work, or lose jobs because you were too slow.

AI is starting to change this. It does not replace experienced estimators. It gives them a tool that thinks at their speed and works from your own rates.

The Real Cost of Slow Estimating

Before looking at the solution, it is worth being clear about what slow and inconsistent estimating actually costs a business.

Time is the obvious one. A detailed rate buildup for a civils package covering earthworks, drainage, concrete and subbase can take three to six hours when done properly. For a firm quoting five to ten jobs a week, that is a significant overhead.

But the hidden costs are just as damaging.

• Inconsistency. When rates come from memory or different versions of a spreadsheet, two estimators will price the same job differently. Over time, this creates margin variance you cannot explain.

• Knowledge risk. If your main estimator leaves, retires, or is ill for a month, how do you maintain output and quality? Most firms do not have a good answer.

• Missed opportunities. Slow turnaround means lost tenders. Clients, particularly main contractors, move on quickly.

• Underpriced work. When you are rushed, you miss items. A forgotten allowance for temporary works, surface water management, or muck away can wipe out your margin on a job.

None of this is unique to groundworks. But the nature of civils work, with varied scopes, complex ground conditions and fluctuating plant and labour costs, makes estimating particularly hard to standardise without the right tools.

What an AI Estimating Tool Actually Does

The term AI gets misused a lot. So let us be clear about what this means in the context of estimating for groundworks firms.

An AI estimating tool, built properly for your business, is trained on your own historical quotes. It learns your rates, your markup structure, your typical allowances for specific work types, and how you build up costs for the jobs you regularly price.

When an enquiry comes in, you feed the tool the scope as a written description, a bill of quantities, or key takeoffs, and it produces a structured rate buildup. Not a generic template. Your rates, applied consistently, with the logic your estimators already use.

Here is what that looks like in practice.

Rate Buildups from Your Own Data

The tool does not pull rates from a generic database. It works from quotes your firm has already priced. Earthworks gang rates, plant hire allowances, disposal costs, concrete rates, drainage labour. All drawn from your history and applied to new scopes.

Structured Output, Ready to Review

The output is not a rough number. It is a line-by-line buildup that your estimator can review, adjust, and sign off. The tool handles the first pass, the time-consuming part, and your team applies judgment at the end.

Scope Identification and Flagging

A well-built estimating tool will also flag items that are commonly missed or that carry risk given the scope. Temporary works, dewatering allowances, reinstatement, access constraints. Things that might not be in the client brief but need to be priced.

Speed Without Cutting Corners

A first-draft buildup that would take your estimator four hours can be produced in minutes. That does not mean skipping the review. It means your estimator spends their time on the parts that actually need human judgment: ground conditions, programme risk, subcontractor selection. Not entering line after line of standard rates.

A Real Example: Groundworks Firm in Oxfordshire

One of the groundworks firms we have worked with is based in Oxfordshire. They operate across residential developments, commercial groundworks, and infrastructure packages, typically ranging from smaller plot works up to six-figure civils contracts.

Their estimating process was largely manual. The director handled most quotes himself, working from a spreadsheet that had evolved over years. It worked, but it was time-heavy, and the firm was struggling to scale enquiries without it eating into time needed elsewhere in the business.

We built a custom AI estimating tool trained on their historical quotes. The tool can take a written scope or key quantities and produce a structured rate buildup across all major work sections: excavation and earthworks, imported fill, drainage, concrete, subbase, and associated preliminaries.

The rates it uses are not generic. They come from the firm's own pricing history, adjusted by work type and scope size. The output mirrors the format the firm already uses, so there is no re-formatting or translation required before it goes into a tender.

The result is a first-draft buildup that takes minutes to produce rather than hours. The director reviews it, makes adjustments where his site knowledge or current market conditions require it, and sends the quote. The process is faster, the output is more consistent, and the risk of missing items is lower.

This is not a replacement for experience. It is a tool that makes experience more productive.

Who This Works For

An AI estimating tool is not the right fit for every firm. It is most valuable where there is an existing body of historical quotes to train from, and where estimating volume is high enough that the time saving has real commercial impact.

The firms that benefit most tend to share a few characteristics.

• They are pricing regularly, at least a handful of jobs per week, and the estimating load is a genuine constraint on capacity.

• Their work types are consistent enough that historical rates are meaningful. Firms doing highly repetitive scopes such as residential groundworks, drainage packages, subbase and concrete get the most from this.

• They have historical quotes they can share. Even a year's worth of tenders gives the tool enough to work from. More is better, but the volume required is lower than most people expect.

• The estimating knowledge sits with one or two people and the business wants to reduce that dependency. This is one of the strongest use cases: using AI to capture institutional knowledge before it walks out of the door.

If your firm prices highly bespoke or one-off scopes where every job is genuinely different, the tool is less effective. But for the majority of groundworks and civils businesses, there is enough repeatability to make AI estimating work well.

What the Setup Process Looks Like

One of the things that puts firms off exploring this is the assumption that the setup will be complicated, expensive, or require significant IT involvement. In practice, it is simpler than most expect.

The process starts with an AI audit. A structured review of how your business currently handles estimating, what data you have available, and where the biggest opportunities are. This is free, takes around 30 to 40 minutes, and gives you a clear picture of what is possible before any commitment.

If the audit identifies a strong fit, we build the tool around your existing quotes and pricing structure. You do not need to change how you estimate. The tool integrates with your current process rather than replacing it.

There is a one-time setup cost and a monthly retainer for ongoing support, updates, and refinement as your rates evolve. Most firms see the time saving cover the cost in the first month.

The Broader Picture: AI in Construction Operations

Estimating is where most construction firms see the clearest, fastest return from AI. But it is one part of a broader opportunity.

The same principles apply to procurement, using AI to match scopes to supplier history and generate RFQ packages faster. To project reporting, pulling data from multiple sources and producing client-ready updates without manual assembly. To supplier communication, document handling, and programme management.

Construction is a sector that has been slower than most to adopt these tools. Firms that move early gain a competitive edge, not just in speed but in the ability to take on more work without proportionally increasing overhead.

The technology exists now. The question is which firms choose to use it.

Find Out What Is Possible for Your Business

If your firm is spending too much time on estimating, losing quotes due to slow turnaround, or relying on one person's knowledge to price work, a free AI audit is the best place to start.

We will review your current process, look at what data you have available, and tell you clearly whether an AI estimating tool would deliver a meaningful return for your business. No jargon, no sales pitch. Just an honest assessment.

Book a Free AI Audit

15 minutes. No cost. A clear picture of where AI can save your team time and improve your margins.