Insight · 8 min

Why fixed-price outsourcing fails — and what to negotiate instead

The structural reason most fixed-price software projects miss scope, and three commercial structures that don't have the same failure mode.

Procurement teams love fixed-price contracts. They're predictable, they're easy to budget, and they shift risk to the vendor. The problem is that they almost always fail in software — and the failure mode is the same every time.

The structural reason fixed-price fails

Software fixed-price contracts assume scope is knowable up front. In about 5% of engagements, it actually is — small features, well-defined integrations, replatforms with no business-rule changes. In the other 95%, scope discovery happens during build. The vendor has two options:

  • Absorb the scope. Margins evaporate. Quality suffers. The team gets pressured to cut corners. The relationship turns adversarial.
  • Bill change requests aggressively. The CR process becomes the project. Every meeting is a negotiation. The vendor wins on paper and loses the renewal.
Fixed-price contracts don't reduce risk. They concentrate it in the change-request process.

Three structures that work

1. Fixed-price with a defined-scope wedge

Use fixed-price only for the part of the work where scope is genuinely knowable. A typical wedge: discovery, architecture, and integration audit. Everything downstream is T&M with a hard cap and a transparent burn-rate dashboard. This works because the parties have already aligned on scope before the variable-cost work begins.

2. T&M with a budget cap and named team

The vendor cannot exceed a defined dollar cap. The team is named in the SOW with CVs attached and cannot be substituted without your approval. Daily/weekly burn is published in a shared dashboard. You get the predictability of fixed-price (cap) with the scope adaptability of T&M.

The catch: this requires real engineering managers on your side to monitor. If you're going to set this and forget it, fixed-price might be safer.

3. Outcome-based with a milestone schedule

Payment tied to validated outcomes — not deliverables. “Order orchestration handles all six channels with <200ms p99 latency” rather than “feature X delivered.” This requires more upfront work to define what good looks like, but once defined, it aligns vendor incentives with yours better than any other structure.

Outcome-based works best when there are 2–4 clear success criteria. It works poorly when there are 25 small features — the overhead of defining each one as an outcome consumes the savings.

The questions to ask before signing

  • If scope expands by 20%, what happens to price and timeline? (Get this in writing.)
  • Who owns the integration risk if a third party causes a delay?
  • What's the change-request turnaround time? Anything over 5 business days kills momentum.
  • If the named team gets reassigned, what's the SLA on replacement?
  • Is there an out clause — and what does it cost?

What we do

For most engagements we offer all three structures and recommend based on scope clarity. Greenfield platforms get phased fixed-price-then-T&M. Modernization gets dedicated team retainer. Replacement projects get outcome-based. We almost never quote fixed-price for the whole engagement — and when a client insists on it, it's usually a sign that the scoping work hasn't been done yet.


If you're evaluating quotes from us or from competitors, we're happy to walk through the commercial structure on a 30-minute call. Sometimes the answer is “our quote is the wrong structure for this work” — and we'll say so.

More from the field notes.

If this resonated, we'd like to work together.

Most of what we publish comes from real engagements. If you're building in this space, the next step is a conversation.