Agency Margins on Development Work: The Math Nobody Shows You

Agencies typically mark up white-label development 1.5–2x. At a $12,000/month dedicated-team retainer billed to the client at $20,000–$24,000, the agency keeps $8,000–$12,000/month in margin without hiring an engineer. Over 12 months, that's $96,000–$144,000 in gross profit from a single client relationship.

Most agencies discover this math by accident. They take on a project too large for their team, subcontract the development, and realise the margin is better than anything they make on in-house work. Then they do it again. And again. Until white-label development becomes a revenue line, not a stopgap.

This resource breaks down the actual numbers — what agencies pay, what they charge, what margin they keep, and where the model fails.

What do agencies actually pay for white-label development?

A dedicated offshore development team — two mid-to-senior engineers working full-time on your client's project — runs $10,000–$14,000/month from a structured Indian engineering partner. The range depends on seniority, stack complexity, and whether project management is included.

That number is the agency's cost of goods sold. Everything above it is margin.

The rate covers engineers who are employed full-time by the development partner — not contractors assembled for the project. This matters because contractor-based teams have hidden costs: onboarding churn, inconsistent availability, and the overhead of managing people who don't know each other's code. A dedicated team from an established partner eliminates those costs. You're paying for a functioning engineering unit, not a collection of resumes. For a detailed breakdown of what drives these rates, see the offshore software development rates resource.

At the $12,000/month midpoint for two engineers, the per-engineer cost is $6,000/month — roughly $36/hour at 167 billable hours. Compare that to a single mid-level US developer at $65–$85/hour, or a senior at $100–$150/hour. The cost gap is the entire business model.

What markup do agencies charge on development work?

The standard range is 1.5x to 2.5x on partner cost, depending on how the agency positions the work to clients. Agencies that sell "our development team" at domestic rates command the highest markup. Agencies that disclose the offshore model and compete on price sit at the lower end.

Here is what three common scenarios look like:

The 2x markup is the most common among US agencies billing development work as a managed service. At that rate, the client pays less than they would for a single senior US engineer — and gets two. The agency keeps the difference.

How does the margin compare to hiring in-house?

The comparison that matters is not offshore vs. in-house for the agency. It is the total cost and risk to the agency of delivering engineering capacity through each model.

The utilisation risk row is where the model gets interesting. An agency with two in-house engineers paying $30K/month has that cost whether or not a client project is running. An agency using a dedicated offshore partner pays $12K/month only when a client engagement exists. Between engagements, the cost is zero.

For agencies running 3–5 concurrent client projects, this difference compounds. In-house teams require bench capacity — engineers between projects who still draw salary. Offshore dedicated teams scale with demand. When a project ends, the cost ends with it.

What does the margin look like over 12 months?

A realistic 12-month projection accounts for the ramp. No agency starts at full capacity on day one. Here is what a typical first-year engagement looks like when it goes well:

Months 1–2: Pilot phase. One engineer, $6,000/month to the partner, billed to the client at $10,000/month. Agency margin: $4,000/month. Total for period: $8,000. The pilot proves the delivery model works and gives the agency confidence to expand.

Months 3–6: Scale phase. Two engineers, $12,000/month to the partner, billed to the client at $21,000/month (1.75x). Agency margin: $9,000/month. Total for period: $36,000. The team is established. The agency has validated the process with their client.

Months 7–12: Steady state. Two to three engineers averaging $15,000/month to the partner, billed at $26,000/month. Agency margin: $11,000/month. Total for period: $66,000. The partnership is now producing consistent, predictable revenue.

Year 1 total margin: approximately $110,000 from a single client relationship.

That number assumes no additional clients. Agencies running two or three concurrent partnerships on the same offshore team infrastructure multiply the margin without multiplying their own overhead.

When does the math break?

The white-label development model is not universally profitable. There are four specific situations where the margin disappears or turns negative.

Hourly billing instead of retainer. When agencies bill clients by the hour and pay partners by the hour, margin becomes unpredictable. A slow sprint where the team only logs 60% of available hours means the agency's margin drops to near zero — while the partner's cost stays fixed (salaried engineers still get paid). Retainers eliminate this problem. The agency pays a fixed monthly amount and bills a fixed monthly amount. The margin is the same every month regardless of sprint velocity.

Single-project engagements under three months. Every new team engagement has onboarding cost — the first two to three weeks are slower as the team learns the codebase, client expectations, and agency processes. On a 12-month engagement, that onboarding cost is amortised across 50 productive weeks. On a 6-week project, onboarding consumes 30–40% of the engagement. The agency pays full rate during ramp but can only bill partial productivity.

Under-scoped projects where the agency eats overruns. If the agency sells a fixed-price project to the client but pays the development partner on a time-and-materials basis, any scope creep comes directly out of margin. A project scoped at 800 hours that actually requires 1,100 hours costs the agency an additional $10,000–$15,000 — which may wipe out the entire project margin. The fix is straightforward: either pass the retainer model through to the client (monthly team cost, not fixed project price) or build a 20–30% buffer into fixed-price quotes.

Client insists on direct management of the offshore team. When the client manages the development team directly — running standups, assigning tasks, reviewing PRs — the agency becomes a staffing pass-through. The client inevitably asks why they are paying the agency a premium for engineers they manage themselves. This leads to rate pressure, margin compression, and eventually the client going direct. The agency must own the delivery relationship: project management, sprint planning, QA, and client communication all stay with the agency. The development partner provides the engineering capacity. The agency provides everything around it.

How do you protect margin when clients push back on price?

Price pressure is inevitable. Clients compare rates, talk to other agencies, and occasionally Google offshore development costs directly. The defence is positioning, not secrecy.

Never disclose your partner's rates to the client. The client is buying a managed development service from your agency. They do not need to know your cost structure any more than a restaurant customer needs to know food costs. The rate you quote is the rate for the service — engineering, project management, quality assurance, communication, and accountability — not for individual engineers.

Price the team as "your development team," not "offshore developers." The framing matters. "We've assigned a dedicated engineering team to your project" carries different weight than "we have some offshore developers who can work on this." The first is a commitment. The second is a commodity. Commodities get price-shopped. Commitments get renewed.

Bundle project management into the rate. If the client breaks down your rate into "engineer cost + PM cost + margin," the conversation moves to line-item negotiation. If the rate is a single number for a managed engineering service, there is nothing to unbundle. You are providing a team that ships working software, communicates proactively, and meets deadlines. The rate covers all of that.

Demonstrate the cost of the alternative. When a client questions pricing, the comparison is not your rate vs. a cheaper offshore shop. The comparison is your rate vs. the client hiring in-house. Two senior US engineers cost $30,000–$35,000/month in fully loaded compensation. Your managed team at $20,000–$24,000/month is already 30–40% less — with no recruiting cost, no benefits administration, no severance risk, and no bench time between projects.

What's the minimum engagement that makes financial sense?

Three months, two engineers. That is the minimum where the economics work for both the agency and the development partner.

Below three months, onboarding cost exceeds productive output. The development team spends weeks learning the codebase and processes, then the engagement ends before the team hits full velocity. The agency pays for ramp time it cannot bill at full rates.

Below two engineers, the team lacks the redundancy to maintain velocity through normal disruptions — PTO, sick days, context-switching. A single-engineer engagement means any absence creates a complete stoppage. Two engineers provide coverage and the ability to pair on complex problems.

The financial minimum: 3 months at $12,000/month = $36,000 total cost to the agency. At 1.75x markup, the agency bills $63,000 and keeps $27,000 in margin. That covers the agency's overhead for acquiring and managing the client relationship, with margin remaining.

The ideal engagement is 6–12 months. The onboarding cost is amortised over a longer productive period, the team gets faster as they learn the domain, and the agency's per-month margin increases as the team operates more efficiently. Most agency partnerships that start as 3-month pilots extend to 12+ months once the delivery model is proven.

What does this look like in practice?

A US agency partner currently runs a dedicated Madgeek engineering team. The engagement generates enough margin for the agency to cover one full-time US hire — while the actual development team costs less than half a single US senior engineer's fully loaded compensation.

The agency presents the team to their clients as their own. NDA from day one. Madgeek engineers use the agency's tools, join the agency's standups (when appropriate), and deliver under the agency's brand. The client sees a well-run engineering team. The agency sees consistent margin on every project that team touches.

That is the model working as designed. The agency adds development capacity without adding headcount. The client gets a team that ships. Madgeek provides the engineering. Everyone's margins are protected because everyone's role is clear.

For a broader look at how white-label software development partnerships work beyond the financial model, that resource covers delivery structure, IP ownership, and communication protocols.

The numbers that matter

Here is the summary math for agencies evaluating this model:

• Cost of a 2-engineer dedicated team: $10,000–$14,000/month
• Standard client billing at 1.75–2x: $17,500–$28,000/month
• Monthly margin range: $6,000–$14,000
• Annual margin from one partnership: $72,000–$168,000
• Minimum viable engagement: 3 months, 2 engineers
• Break-even risk: hourly billing, sub-3-month projects, scope overruns, loss of delivery control

The math works when the agency controls the client relationship, prices as a managed service, and commits to engagements long enough for the team to reach full productivity. It fails when the agency acts as a pass-through, competes on hourly rate, or takes on projects too short to recover onboarding cost.

If you're running an agency and want to explore what a dedicated development partnership looks like with real numbers for your specific situation, start a conversation. We will scope it to your client mix and billing model.