# How to Build the Business Case for Preconstruction Software

> A preconstruction software business case: compare the license to estimator labor and scope risk, and find the hours or scope gaps that pay for it each year.

- URL: https://consight.ai/articles/preconstruction-software-business-case
- Published: 2026-11-11
- Updated: 2026-11-11
- Topic: [Preconstruction & Estimating](https://consight.ai/articles/topics/preconstruction)
- Author: Stewart Toubia

A business case for preconstruction software compares the license to two larger costs: estimator time and missed scope. Take a general contractor with $100 million in revenue and 23 estimators. Estimating labor costs that contractor about $2.9 million a year. A $90,000 platform is 3.1% of that.

The platform pays for itself if each estimator gets back about 1 hour 15 minutes a week. It also pays for itself if it prevents scope losses equal to 0.09% of revenue. Either result covers the cost alone.

This page walks through that example. The calculator above opens with the same figures, so you can swap in your own company.

## Software is the smallest cost in preconstruction

Most software debates compare one license to another. That is the wrong comparison. Estimator labor is many times larger than any tool, and a single missed scope item can cost more than the license.

Here is the example for a team of 23 estimators:

| **Annual cost**   | **How it is figured**              | **Amount** | **Share of labor** |
| ----------------- | ---------------------------------- | ---------: | -----------------: |
| Estimating labor  | 23 estimators × $90,000 × 1.4 load | $2,898,000 |               100% |
| Proposed platform | Quote, unlimited users, AI budget  |    $90,000 |               3.1% |

Bid distribution tools usually cost less than AI preconstruction software. The gap between the two is a few percent of what the estimating team costs. The real money sits in the estimators' hours and in the scope they might miss.

The quote includes a monthly AI budget. AI usage is the processing that reads plans, writes scopes, and levels proposals. A bid distribution tool has no such cost, because it does not do that work.

## How to figure the loaded cost of an estimator

Start with the average estimator salary, then multiply by a load factor for benefits and payroll taxes. The example uses 1.4. The Bureau of Labor Statistics reports that wages were 70.0% of private employers' compensation costs in June 2026 ([BLS Employer Costs for Employee Compensation](https://www.bls.gov/news.release/ecec.nr0.htm)). Salary divided by 0.70 is a factor of about 1.43, so 1.4 is a fair round figure.

Divide the loaded cost by 2,080 work hours a year. At $90,000 and 1.4, one estimator costs $126,000 a year, or about $60.58 an hour. Use your own payroll numbers if you have them.

## Where the estimators' time goes

Scope writing and bid leveling take the most manual hours. A typical commercial project has 35 to 50 bid packages. By hand, each package takes 2 to 6 hours to scope and level. That is 70 to 300 estimator hours a project. These hours per package are Consight's estimates, so check them against your own team.

| **Task, per bid package** | **By hand** | **In Consight** |
| ------------------------- | ----------: | --------------: |
| Scope writing             |   1–2 hours |    3.73 minutes |
| Bid leveling              |   1–4 hours |    3.42 minutes |

The Consight times are averages of 246 scope runs and 214 bid leveling runs from October 30 to November 5, 2026. They cover the runs alone, not the estimator's review. The [performance benchmarks](https://consight.ai/articles/preconstruction-ai-performance-benchmarks) explain the method and its limits.

Every hour you save per estimator per week is worth real money. On the example team, one hour a week across 23 estimators is worth $72,450 a year:

| **Hours saved per estimator per week** | **Annual value** | **Covers $90,000?** |
| -------------------------------------: | ---------------: | ------------------: |
|                                    1.0 |          $72,450 |                  No |
|                                    1.5 |         $108,675 |                 Yes |
|                                    2.0 |         $144,900 |                 Yes |

The break-even point is 1.24 hours a week, or about 1 hour 15 minutes. That is about 15 minutes a day, less than one bid package a week.

## Where the scope risk hides

The second path to break-even is scope risk. A missed scope item usually turns into a change order after buyout, and the contractor pays for it. Every dollar of that loss comes straight out of profit.

Measure scope risk against profit, not against the project. The example contractor has $100 million in annual revenue. At a 5% net margin, that is $5 million of profit. The 5% margin is an assumption for a typical general contractor; enter your own in the calculator.

| **Scope risk, example contractor**            |   **Amount** |
| --------------------------------------------- | -----------: |
| Annual revenue                                | $100,000,000 |
| Annual net profit at a 5% margin              |   $5,000,000 |
| One missed scope item, 1.5% of a $20M project |     $300,000 |
| Share of a year's profit that one miss erases |           6% |
| Years of a $90,000 platform one miss pays for |         3.33 |

To break even on scope risk alone, the software must prevent $90,000 of scope losses a year. That is 0.09% of revenue, less than a third of one missed item.

Alston Construction's Orlando office reported 33% fewer change orders after it started using Consight's risk review. That figure is customer-reported. Use it as a signal, not a promise, and measure your own change orders. For the review itself, see [how to find scope gaps in subcontractor bids](https://consight.ai/articles/how-to-find-scope-gaps-in-subcontractor-bids).

## The two paths to break-even

Write the case as two separate paths, so the executive team needs to believe only one of them:

1. **Path A, estimator time.** Each estimator gets back about 1 hour 15 minutes a week. The team's time savings then equal the license.
2. **Path B, scope risk.** The software prevents scope losses equal to 0.09% of revenue, or $90,000 a year, across every project the team bids.

Neither path depends on the other. Any mix of the two clears the bar too. A team that saves one hour a week per estimator covers $72,450 of the cost from time alone. That team then needs only $17,550 of prevented scope losses to break even.

## What changes for the estimating team

Bid distribution tools send invitations and track bidders. They do not write scopes, read proposals, or check for gaps. The business case rests on the steps that are still manual:

| **Step**             | **Today**                                          | **With AI preconstruction software**                       |
| -------------------- | -------------------------------------------------- | ---------------------------------------------------------- |
| Scopes and packages  | Written by hand, 1 to 2 hours a package            | Drafted from plans and specs, with source references       |
| Collecting proposals | Pulled from inboxes and portals by hand            | Emailed PDF and Excel proposals filed to the right package |
| Bid leveling         | Values keyed in by hand, 1 to 4 hours a package    | Proposals read and mapped to scope lines                   |
| Scope risk           | Caught only if a reviewer spots it before deadline | Exclusions, scope gaps, and spec conflicts flagged         |
| Award                | Briefings and revision requests assembled by email | Written leveling summary and revision requests             |

## Questions your executive team will ask

**We already pay for bid management. Why spend more?** The two tools do different jobs. Bid management distributes bids and qualifies subcontractors. Preconstruction software cuts estimator hours and catches scope gaps, and part of its price pays for the AI usage that does that work. Price it against labor and risk, not against another license. [AI bid leveling vs bid management software](https://consight.ai/articles/ai-bid-leveling-vs-bid-management-software) sets out the difference.

**How do we know the AI got it right?** Ask every vendor to link each leveled figure to the proposal text behind it. Then estimators check a source instead of retyping a number.

**Will it fit how our estimators work?** Check that it uses your package names and imports your bid forms. Check that it exports to Excel and connects to the systems you already use.

**What does the annual price include?** One annual price covers the whole team within the contracted business unit or region. Adding estimators, project managers, or executives does not add license cost. AI usage counts against a monthly AI budget, so ask how the AI budget fits your bid volume.

## How to measure the return after rollout

Agree on a short scorecard before day one. Then measure against your own baseline, not a vendor's.

| **Metric**                   | **How to measure it**                                           | **Break-even signal**                  |
| ---------------------------- | --------------------------------------------------------------- | -------------------------------------- |
| Hours per package leveled    | Log time on a few packages before rollout, then on similar ones | 1 hour 15 minutes per estimator a week |
| Scope writing time           | Time a scope written by hand vs a reviewed draft                | Counts toward the same weekly target   |
| Gaps flagged before award    | Flags the estimators confirm were real gaps                     | $90,000 of confirmed gaps a year       |
| Scope-related change orders  | Change orders traced to missed scope, vs similar past jobs      | Fewer scope-gap change orders          |
| Packages run through the app | Share of active packages scoped and leveled in the software     | Most active packages                   |

A structured pilot gives you the before numbers. [How to pilot AI bid leveling software](https://consight.ai/articles/how-to-pilot-ai-bid-leveling-software) lays one out over 60 to 90 days.

## How the calculator works

You enter eight numbers. Four describe the team and the tool: team size, average salary, load factor, and annual software cost. Two describe the company: annual revenue and net profit margin. Two describe scope risk: a typical project value and the cost of one missed scope item as a percent of that value. Enter percents as whole numbers (5, not 0.05).

The calculator returns eleven figures:

1. **Loaded hourly rate** is salary × load factor ÷ 2,080 hours.
2. **Annual estimating labor** is estimators × salary × load factor.
3. **Software share of labor** is software cost ÷ annual estimating labor × 100.
4. **Value of one hour a week** is annual estimating labor ÷ 40, the hours in a work week.
5. **Break-even hours a week** is software cost ÷ the value of one hour a week.
6. **Break-even minutes a day** is break-even hours a week × 60 ÷ 5 workdays.
7. **Annual net profit** is revenue × net margin ÷ 100.
8. **Cost of one missed item** is project value × the miss percent ÷ 100.
9. **Share of a year's profit one miss erases** is the cost of one missed item ÷ annual net profit × 100.
10. **Break-even scope losses as a share of revenue** is software cost ÷ revenue × 100.
11. **Years one miss pays for** is the cost of one missed item ÷ software cost.

## Worked example: 23 estimators and a $90,000 platform

The calculator opens with this example.

- Loaded hourly rate: $90,000 × 1.4 ÷ 2,080 = $60.58.
- Annual estimating labor: 23 × $90,000 × 1.4 = $2,898,000.
- Software share of labor: $90,000 ÷ $2,898,000 = 3.11%.
- Value of one hour a week: $2,898,000 ÷ 40 = $72,450.
- Break-even hours a week: $90,000 ÷ $72,450 = 1.24 hours, or about 1 hour 15 minutes.
- Break-even minutes a day: 1.2422 × 60 ÷ 5 = 14.91 minutes.
- Annual net profit: $100,000,000 × 5 ÷ 100 = $5,000,000.
- Cost of one missed item: $20,000,000 × 1.5 ÷ 100 = $300,000.
- Share of a year's profit one miss erases: $300,000 ÷ $5,000,000 = 6%.
- Break-even scope losses as a share of revenue: $90,000 ÷ $100,000,000 = 0.09%.
- Years one miss pays for: $300,000 ÷ $90,000 = 3.33 years.

Consight is AI preconstruction software for commercial general contractors. It drafts scopes from plans and specs, levels subcontractor proposals, and flags exclusions and scope gaps, with every figure linked to its source. [Talk to sales](https://consight.ai/talk-to-sales) to build this case with your own numbers.
