Custom Software Pricing Models

Custom Software Pricing Models: Fixed Price vs T&M vs Outcome-Based (2026 Comparison)

“How much does custom software cost?” is the wrong question.

The right one is:

“Which pricing model should I use for this project?”

Because the same project, run under different models, can end up costing 2 to 3 times more.

An $80K fixed-price project can balloon to $240K total over 18 months once you add scope creep, change requests, bug-fix retainers and partial rewrites.

The same project run on T&M, or time and material, can end at around $110K because there are no surprises, and ownership stays on your side.

The same project on outcome-based pricing aligns the vendor’s incentives with yours. But if you structure it wrong, the vendor walks away mid-project and you are left holding half a system.

In other words:

The pricing model is not a contract detail.

It is the first decision that determines the fate of the entire project.

This article compares the four major models: Fixed Price, T&M, Outcome-Based and Hybrid.

We look at when they work, when they fail, real cost ranges and contract red flags.

This framework is distilled from 16 years of custom software work, including both projects we built and contracts we were brought in to rescue.

The 4 Pricing Models: Quick Overview

Fixed Price

Risk holder: Vendor on paper, but in reality the risk often shifts back to you through change requests.

Flexibility: Low.

Best project fit: Defined scope, small project.

Typical final cost compared with quote: 150-200%.

Time & Material, or T&M

Risk holder: You.

Flexibility: High.

Best project fit: Post-discovery, main build.

Typical final cost compared with quote: 95-110%.

Outcome-Based

Risk holder: Shared, if structured correctly.

Flexibility: Medium.

Best project fit: Projects with a measurable KPI.

Typical final cost compared with quote: 90-130%, depending on KPI structure.

Hybrid, or Phased

Risk holder: Distributed by phase.

Flexibility: High.

Best project fit: Most enterprise projects.

Typical final cost compared with quote: 100-120%.

Each model is explained below.

Model 1: Fixed Price

How It Works

The vendor commits to a defined scope at a single price and delivers against milestones.

The upfront payment is usually 20-50%, and the remaining payment is tied to milestones.

When It Works

Fixed Price can work when discovery and design are complete with full scope clarity.

This means not just a 1-2 page user flow, but an actual data model, API contract and UI specification.

It can also work when:

The project is small, usually 3-6 months and under $200K.

The vendor has done 80% or more of it before.

The project follows a repeatable pattern.

You are not expecting change in regulation, market conditions or product direction.

When It Fails

Fixed Price usually fails when discovery is skipped or shallow.

In that case, the vendor adds a 40-50% risk premium and then bills change requests on top.

It also fails when:

The project is longer than 6 months.

Market, regulation or user feedback is likely to change the scope.

Every small shift becomes a separate contract.

The project involves innovative or new technology.

Your internal team is too junior to evaluate change request impact.

Vendor pricing becomes unchecked.

Real Cost

Industry average across Türkiye and global markets, based on 2024-2026 project data:

72% of fixed-price projects exceed their final budget by 30-150%.

The single biggest contributor is change request billing.

Anything the vendor marks as “out of scope” is usually premium-priced.

Red Flags in Vendor Conversations

“We can give you a single price without discovery.”

This usually means either the buffer is bloated or they will bill out-of-scope items separately.

“Our change request process is very flexible.”

Flexible often means expensive.

Upfront payment above 50%.

This means the vendor is collecting risk from you.

Contract Watch List

Make sure the scope document is attached to the contract and every page is signed off.

Change request pricing must be written in the contract, including hourly rate and admin fee.

Acceptance criteria per milestone must be measurable.

“User-friendly” does not count.

“Lighthouse score above 90” does.

Delay penalty should be contractually defined.

A reasonable range is 1-3% per week.

Model 2: Time & Material, or T&M

How It Works

Billing is based on hours or story points.

A team is allocated to the project, for example:

1 project manager

3 developers

1 quality assurance specialist

The project is usually invoiced monthly or bi-weekly.

Contracts often include a ceiling, or budget cap.

When It Works

T&M is the default model for most modern enterprise projects.

It works well after discovery, especially for MVP development and iteration.

It is also suitable for:

Projects where scope will certainly shift.

Research and development projects.

New products.

AI pilots.

Projects where the internal team can run project management.

Situations where transparency matters.

Every hour can be visible, every sprint can be demoed and every pull request can be reviewed.

When It Fails

T&M fails when your internal team cannot run project management.

If you cannot control vendor velocity, billing compounds quickly.

It also fails when:

There is no ceiling in the contract.

The vendor falls into “just two more weeks” loops.

Sprint reviews are not attended.

Quality control drops.

Rework grows.

The vendor does not show the senior, mid-level and junior split.

You end up paying senior rates for junior work.

Real Cost

T&M projects typically finish at 85-110% of the estimate.

Large budget overruns are rare.

Because risk sits largely with you, the vendor’s incentive is not necessarily speed.

It is quality.

That usually plays out well when the project is managed properly.

Red Flags

The senior, mid-level and junior mix is hidden.

If there is only one “team rate,” you may be paying senior rates for juniors.

There is no time tracking.

There is no Jira or Linear access.

You cannot see what you are paying for.

The vendor does not report velocity.

“Going a bit slow” can stay vague for weeks.

Contract Watch List

Senior, mid-level and junior hourly rates should be listed separately.

There should be a monthly ceiling that requires approval to exceed.

You should have access to the time tracking tool from day one.

You should have the right to attend sprint reviews and see every sprint demo.

Monthly velocity, burndown and bug rate reports should be included.

Model 3: Outcome-Based or Success Fee

How It Works

The vendor takes a base fee, usually 50-70%.

The rest is tied to a measurable outcome.

Example:

“If daily active users exceed 1,000 within 90 days of MVP launch, the remaining 30% is paid.”

When It Works

Outcome-based pricing works when there is a measurable KPI the vendor can actually influence.

Examples include:

Daily active users

Conversion rate

Latency

Processing time

It also works when:

The vendor knows the domain deeply.

The vendor is comfortable taking outcome-based risk.

The KPI dependency is fair.

For example, if the vendor does not control the marketing budget, daily active users may not be a fair outcome metric for them.

When It Fails

Outcome-based pricing fails when the KPI is not in the vendor’s control.

The vendor complains, and the contract turns into a dispute.

It also fails when:

The base fee is too low.

The vendor rushes the project to chase the outcome bonus.

Quality suffers.

The KPI is a single number.

The vendor may “hack” the metric by counting the wrong users or measuring latency on one edge case.

Real Cost

When set up correctly, outcome-based projects land at 90-100% of the original budget.

Vendor incentives are aligned, so quality is usually high.

When set up poorly, the vendor may exit early and the contract can turn into a dispute.

In that case, total cost can reach 1.5 to 2 times the original budget.

Red Flags

The vendor accepts an outcome model without demanding extensive discovery.

This means they cannot actually calculate the risk.

The base fee is under 40%.

This raises questions about the vendor’s financial sustainability.

They may drop the project mid-way.

The KPI measurement methodology is not detailed in the contract.

This means you may lose any later dispute.

Contract Watch List

The KPI measurement methodology must be fully defined in the contract.

It should state:

Who measures the KPI

Which tool is used

When the snapshot is taken

Which variables are excluded

Variables outside the vendor’s control, such as marketing budget, should be excluded.

The outcome measurement window should usually be 30-90 days.

The base fee should usually be 50% or more to protect vendor financial stability.

Model 4: Hybrid, or Phased

How It Works

The project is split into phases, and each phase uses a different pricing model.

Phase 1: Discovery

Pricing model: Fixed-fee.

Typical duration: 2-4 weeks.

Typical cost: $5K-$25K.

Deliverables:

Technical specification

Risk registry

Milestone plan

Phase 2: MVP or Initial Build

Pricing model: T&M.

Typical duration: 3-6 months.

Structure: T&M with a ceiling.

Phase 3: Iterate or Scale

Pricing model: Continued T&M or transition to outcome-based pricing.

Phase 4: Maintenance

Pricing model: Monthly retainer or SLA-based structure.

When It Works

Hybrid works for most mid-to-large enterprise projects.

It is especially suitable when:

The project is 6+ months.

The budget is $200K+.

You want to keep vendor-switching optionality open.

You want a phase-gate after each phase.

Scope is expected to evolve.

Control needs to stay with you.

When It Fails

Hybrid fails when phase transitions are not well-defined in the contract.

The vendor may start Phase 2 as if Phase 1 is not done.

It also fails when:

Discovery deliverables are weak.

The first T&M month of Phase 2 turns into scope chaos.

The maintenance retainer is set too low.

The vendor neglects maintenance.

Production issues get slow responses.

Real Cost

Hybrid usually lands in the 100-120% range.

It is the most stable model.

The extra time and cost spent on discovery materially improves the predictability of the main build.

Contract Watch List

Use a phase-gate contract.

You should have continuation rights at each phase boundary.

You should have penalty-free exit rights.

Discovery deliverables should be explicitly listed.

These should include:

Technical specification

User flows

Data model

API contract

Risk registry

Milestone plan

Phase 2 T&M ceiling should be set after discovery, not before.

Maintenance SLA and hourly rates should be listed separately.

Which Model for Which Project?

Fully Defined Scope, 3-6 Months, Repeatable Pattern

Best fit: Fixed Price.

Why: The vendor can manage risk, and the surprise factor is low.

Research and Development, New Product or Evolving Scope

Best fit: T&M.

Why: Flexibility is essential, and ownership stays internal.

Clear Measurable KPI and Vendor Can Influence It

Best fit: Outcome-Based.

Why: Incentives are aligned, but setup quality matters.

6+ Months, $200K+ Budget and Evolving Scope

Best fit: Hybrid.

Why: Discovery is fixed, MVP is T&M, and scale can continue as T&M or outcome-based.

Default for Most Enterprise Customers

Best fit: Hybrid.

Why: It manages risk phase by phase, does not surrender control and sets up a healthy long-term vendor relationship.

Upfront, Milestones and Payment Terms: Practical Notes

Upfront Payment

20-30% is ideal.

It is enough for vendor cash flow and low risk for you.

50% or more means the vendor is collecting risk from you.

Be cautious.

0% may signal vendor stability concerns.

Look elsewhere if the vendor cannot sustain basic delivery without any upfront.

Milestone Count

For a 3-month project, there should be at least 3 milestones.

For a 6-month project, there should be at least 5 milestones.

No single milestone should exceed 25% of the total project budget.

Net 30, Net 60, Net 90

Net 30 is standard and acceptable.

Net 60 or longer can create vendor cash flow concerns.

This increases mid-project exit risk.

If you demand Net 60 or Net 90, the vendor usually adds a premium of 5-8%.

IP and Ownership

All code and documentation should belong to you after final payment.

If the vendor uses internal libraries, they should either be MIT/Apache-licensed or transferred to your ownership.

This is covered in more detail in Category 5 of our Vendor Selection Checklist.

7 Pricing-Specific Red Flags in Contracts

1. Single hourly rate

There is no senior, mid-level or junior split.

2. Fixed-price quote without discovery

The vendor is pricing risk blindly.

3. Upfront payment above 50%

The vendor is moving risk to you.

4. Change request pricing is not in the contract

If it is open to negotiation later, you usually lose.

5. No penalty for delay

The cost of slipping sits on you, not the vendor.

6. No phase-gate

Exiting mid-project becomes expensive or impossible.

7. Maintenance retainer is mandatory with the build

This may be a vendor lock-in signal.

If 3 or more of these apply, find another vendor.

Next Step

Pricing model and vendor selection should be decided together.