Scope-Based Estimation Is Fiction — Why Agencies Lose Margin Before the Project Starts

Rachel Jacobs, a leading authority on scaling ecommerce agencies, stated a fundamental truth: “Most agencies price based on scope. They estimate hours, add a margin, and send the proposal. The scope estimate is fiction the moment the project starts.” ¹

This isn’t a complaint about scope creep. It’s a fundamental critique of how agencies price their work. The entire pricing model is built on an estimate that everyone knows will be wrong — but nobody admits it.

The Fiction

Here’s what happens in every agency:

1. Sales receives a project brief
2. A senior developer estimates hours based on similar past projects
3. A margin is added — typically 20-30%
4. The proposal is sent with a fixed price

Then the project starts, and within two weeks, the estimate is obsolete. Requirements that seemed clear during the sales process reveal ambiguity during implementation. Third-party APIs behave unexpectedly. The client’s “small change” cascades through the architecture.

The agency works more hours than estimated. The margin erodes. The project that was supposed to generate profit instead covers costs.

And the root cause isn’t poor estimation — it’s the assumption that scope can be known in advance.

Why Estimation Is Inherently Flawed

Software development is not manufacturing. It’s research and development disguised as production. Every project discovers new information during implementation — about the technology, the domain, the users, and the feasibility of the original requirements.

A fixed scope estimate assumes none of this discovery will happen. It assumes:

• All requirements are known before coding starts
• No new information will surface during implementation
• The team’s velocity is predictable
• Third-party systems behave as documented

Every experienced developer knows these assumptions are false. Yet agencies continue to estimate based on them.

The Cost of the Fiction

Margin Erosion

A project estimated at 200 hours that takes 280 hours doesn’t just lose margin — it destroys it. At $100/hour billable rate and $70/hour effective cost (including overhead):

• Estimated profit: 200 × ($100 - $70) = $6,000
• Actual profit: (200 × $100) - (280 × $70) = $20,000 - $19,600 = $400

The project that should have generated $6,000 in profit generates $400. A 93% margin loss.

Damaged Client Relationships

When the estimate proves wrong, the agency faces an impossible choice: absorb the loss or go back to the client for more money. Either option damages trust.

Inaccurate Future Estimates

Without historical data on estimate vs. actuals, agencies repeat the same mistakes. Each project is a new guess, not an iteration on past performance.

How to Price Projects Honestly

1. Use Range Estimates, Not Single Numbers

Replace “This project costs $50,000” with:

Conservative estimate: $42,000 (if everything goes perfectly)
Realistic estimate:   $50,000 (based on similar projects)
Pessimistic estimate: $65,000 (if unexpected complexity arises)

The client sees the uncertainty and can make an informed decision.

2. Separate Core Scope from Risk Buffer

Present the price as:

• Core scope: $40,000 (what you’re certain about)
• Risk buffer: $10,000 (25% for uncertainty)
• Total: $50,000

When the project comes in under budget, the unused buffer belongs to the client (or is converted to additional scope). This aligns incentives: the agency isn’t penalized for accurate estimates.

3. Track Estimate vs. Actuals

After each project, compare the original estimate to actual hours. This data is the single most valuable input for improving future estimates. Without it, every new estimate is a guess.

How Apropo.io Helps Agencies

Market Benchmarks: Before sending a proposal, compare your estimate against market data. If your estimate for a SaaS MVP is $30,000 but the median is $60,000 ², you’re likely underestimating. The benchmark helps calibrate before you commit.

Quote Sanity Check: Verify that your proposal includes risk buffers and that the components align with industry standards.

Estimate vs. Delivery Tracking: After project completion, compare estimated vs. actual hours. This data feeds directly into improving your next estimate.

Interactive Quotes: Present range estimates and risk buffers so clients see the uncertainty behind the number. Transparency builds trust.

Summary

Scope-based estimation is fiction. Every software project discovers new information during implementation, and that information changes the scope. Agencies that pretend otherwise build their business model on a flawed assumption.

The fix isn’t to estimate more accurately — it’s to change the pricing model. Use range estimates, separate risk buffers, and track actuals against estimates. The goal isn’t perfect estimation. It’s honest communication about uncertainty, backed by data.