A CFO comparing an in-house hire against an outsourced development team almost always starts with the same spreadsheet: gross salary on one side, monthly vendor rate on the other. That comparison is not wrong, exactly – it is incomplete in a way that consistently favors the in-house column, because gross salary is the most visible cost of an employee and also the smallest fraction of what that employee actually costs the business over a full year. A properly built total-cost-of-ownership model closes this gap, and the CFOs who build one are frequently surprised by how much the real number moves once every line item is accounted for.
This is not an argument that in-house hiring is always the wrong call – for some roles, particularly ones close to core IP or requiring deep institutional context, it clearly is not. It is an argument that the comparison most finance teams run today is systematically biased by omission, and a biased comparison produces a biased board decision. This article walks through every cost category a fair comparison needs to include, with typical ranges, so the next build-versus-outsource decision that crosses your desk is evaluated on real numbers rather than a gross-salary illusion.
The Line Items That Never Make It Onto the First Spreadsheet
Gross salary is the anchor figure everyone starts with, and it is also the smallest true cost driver in a complete model. Layer in employer social contributions – typically 20-30% of gross salary across most Western European jurisdictions depending on the country and benefit structure – and the number is already meaningfully larger before you have added a single additional line. Recruiter or agency fees for a specialist technical hire commonly run 18-25% of first-year salary, payable whether or not the hire ultimately works out, and rarely tracked as a cost attributable to that specific headcount decision once it disappears into a general recruiting budget line.
Benefits, equipment, software licensing, and a proportional share of office overhead typically add another 10-15% on top of salary and employer contributions combined. None of these figures are secret or hard to find – they simply live in different budget categories (HR, facilities, IT procurement) than the engineering headcount line, which is precisely why they rarely get pulled together into a single per-engineer cost figure when a build-versus-outsource comparison is being made under time pressure.
The Cost Nobody Puts a Number On: Ramp Time
A new in-house engineering hire is not fully productive on day one, and treating them as though they are – which is exactly what a simple salary comparison implicitly does – understates the true cost significantly. Depending on role seniority and codebase complexity, meaningful productivity ramp typically takes eight to sixteen weeks, during which the company is paying full salary and benefits for output well below what the same engineer will eventually produce. Valued at even a conservative 40% productivity discount across a twelve-week ramp, this represents roughly one additional month of fully-loaded cost that almost never appears as a distinct line item anywhere in a standard hiring budget.
This cost compounds for specialist and senior roles specifically, which is where in-house hiring struggles most against outsourcing in the first place. A senior engineer joining a complex, poorly documented legacy codebase can take considerably longer than twelve weeks to reach full context and productivity – and that ramp cost is paid in full regardless of whether the hire eventually works out, which is the segue into the next, and often largest, hidden cost.
Attrition: The Line Item That Erases Prior Investment
Median tenure for software engineers at companies without strong retention programs sits meaningfully below the multi-year horizon most hiring budgets implicitly assume when justifying the upfront cost of a hire. When an engineer departs after roughly fourteen to eighteen months – a commonly observed pattern in competitive EU tech labor markets – the company has typically not yet recouped the full recruiting, onboarding, and ramp-time investment made in that hire, and now bears the cost of repeating the entire cycle: a new recruiter fee, a new ramp period, and a productivity gap while the departed engineer’s institutional knowledge walks out the door with them.
A complete total-cost-of-ownership model should apply an attrition-adjusted multiplier to the base hiring cost, not just the headline recruiting fee. Modeling a realistic attrition rate for the relevant role and market – often in the 15-22% annual range for competitive engineering roles in Western Europe – and amortizing the full re-hiring cycle cost across the actual average tenure, rather than an assumed multi-year tenure, typically pushes true annualized in-house cost to 1.4x-1.8x the gross salary figure most spreadsheets start with. This single adjustment is usually the largest contributor to the gap between the naive comparison and the real one.
Management Overhead: A Cost That Scales With Headcount, Not With Output
Every in-house engineer added to a team also adds a fraction of engineering management overhead – a manager’s time spent on one-on-ones, performance reviews, hiring panels for future openings, and internal coordination that scales with headcount more directly than it scales with delivered output. This cost is real and budgeted, typically as part of a management salary line, but it is rarely allocated back to the specific hiring decision under evaluation, which means the true marginal cost of each additional in-house engineer is understated by whatever fraction of a manager’s time that engineer consumes.
An outsourced dedicated team model folds this overhead into the vendor’s own internal management structure – team leads and delivery managers on the vendor’s side handle the coordination and performance management that would otherwise consume your internal managers’ time, at a cost already reflected in the monthly rate rather than appearing as a separate, uncounted internal cost.
What the Outsourced Side of the Comparison Should Include Too
A fair comparison cuts both directions, and CFOs building this model should hold the outsourced side to the same rigor. A dedicated team’s monthly rate typically already includes the vendor’s own recruiting, benefits, management overhead, and equipment costs – which is precisely why it should be compared against the fully-loaded in-house figure, not the gross salary figure, to avoid an equally biased comparison running in the opposite direction. Onboarding time for an outsourced team is real too, typically shorter than an in-house hire’s ramp but not zero, and should be modeled explicitly rather than assumed away.
Contract transition costs also belong in a complete model: if switching vendors or bringing work back in-house eventually becomes necessary, there is a knowledge-transfer cost on that side of the ledger as well. A rigorous CFO does not build a model that flatters one option by omission – the goal is a comparison honest enough to survive a board member asking a pointed follow-up question about any single line item.
Building the Actual Calculator
A defensible total-cost-of-ownership model for an in-house hire, per engineer per year, should sum: gross salary, employer social contributions (20-30% of salary), recruiting and agency fees amortized over expected tenure (18-25% of first-year salary, divided by expected tenure in years), benefits and overhead (10-15% of salary), ramp-time cost (roughly one month of fully-loaded cost, non-recurring in year one), and an attrition-adjusted re-hiring reserve based on your realistic tenure assumption for the role and market. Run this same structure against the outsourced dedicated team’s fully-loaded monthly rate, including its own onboarding period, and you have a comparison built on comparable, complete figures rather than a gross-salary illusion on one side and an all-inclusive rate on the other.
The in-house cost calculator, line by line:
Two lines in that table deserve a note on method. The management overhead line is the one most finance teams leave blank because it requires an estimate rather than a pulled figure; a reasonable starting point is to ask the engineering manager how many hours per week go to one-on-ones, reviews and hiring panels per direct report, and price those hours at the manager’s own fully-loaded rate. The attrition reserve line is the one that changes most between markets, so use your own HR data for the specific role family rather than a generic industry figure whenever that data exists.
Manifera builds this exact comparison for CFOs evaluating a dedicated team engagement, using real client data from 160+ delivered projects to show where the fully-loaded numbers typically land for comparable roles and seniority levels. You can request a detailed cost breakdown specific to your team composition and see how the model applies to your own hiring plan through our dedicated development team service, or review how engagements scale over time in our portfolio.
Making the Case to Your Board With Real Numbers
The purpose of building this calculator is not to conclude, in every case, that outsourcing wins – some roles genuinely warrant the higher fully-loaded cost of in-house ownership, particularly where institutional context and long-term IP control outweigh the cost differential. The purpose is ensuring the comparison your board sees reflects the real economics of both options rather than a gross-salary figure that consistently and silently favors whichever option looks cheapest at first glance without capturing ramp time, attrition, and management overhead.
A CFO who brings a complete, defensible total-cost-of-ownership model into a build-versus-outsource decision earns more credibility with the board than one who brings a gross-salary comparison, regardless of which option the numbers ultimately favor – because the board can see the full reasoning rather than taking a headline figure on faith.
Request a fully-loaded cost comparison tailored to your team’s specific roles and seniority mix from our Amsterdam team, built to hold up under your own board’s scrutiny.
Worked Example: One Senior Engineer, Netherlands, Year One
Put real numbers against the framework. A senior backend engineer hired in the Netherlands at €80,000 gross salary: add employer social contributions at 25% (€20,000), recruiting fees at 22% of first-year salary amortized over an assumed 16-month tenure (€17,600 ÷ 1.33 years ≈ €13,200 annualized), benefits and overhead at 12% (€9,600), and a ramp-time cost equivalent to one month fully-loaded (roughly €10,600 including contributions and overhead). Sum: approximately €133,400 in year one against an €80,000 headline salary – a 1.67x multiplier, squarely inside the 1.4x-1.8x range this model predicts.
Run the same engineer as part of a five-person dedicated offshore team at a blended $4,300/month fully-loaded rate: roughly $258,000/year for the team of five (€237,000), or €47,400 per engineer – already inclusive of the vendor’s own recruiting, benefits, and management overhead, with no separate attrition reserve needed since backfill is contractually the vendor’s responsibility. Set side by side, the €133,400 in-house figure and the €47,400 per-head offshore figure are the numbers a CFO should actually be comparing, not €80,000 against a headline monthly rate – the gap the naive comparison misses is nearly €86,000 per engineer per year.
Worked example: one senior engineer, Netherlands, year one:
Year Two: Why the Gap Does Not Close on Its Own
A common objection from finance teams is that the in-house figure is inflated by one-off year-one costs, and that the comparison will look very different once the hire is settled. Part of that is true: the ramp-time cost for the original hire does not repeat in year two. But with a realistic 16-month tenure, the attrition reserve replaces it almost exactly. Steady-state cost for the same engineer is roughly €80,000 salary, €20,000 contributions and €9,600 overhead, plus a re-hiring reserve of about €21,000 (roughly 0.75 replacement cycles per year, each costing a new recruiting fee and a new ramp period). That lands near €130,800, or about 1.64x gross salary – barely lower than year one.
The offshore side changes less, because the vendor carries recruiting, ramp and backfill inside its own rate. The figures to watch there are contractual rather than operational: any annual rate indexation clause, the notice period for scaling the team down, and what the contract says about replacing an engineer who leaves. A CFO presenting a multi-year view should model both columns for at least three years, so the board sees that the cost gap is structural rather than a year-one artifact.
FAQs
(Scenario: CFO reviewing a first-draft hiring budget) What is the real total cost of an in-house software engineer versus gross salary?
When employer contributions, recruiting fees, benefits, ramp-time, and an attrition-adjusted re-hiring reserve are all included, the real annualized cost typically runs 1.4x to 1.8x the gross salary figure most spreadsheets start with.
(Scenario: CFO whose last two engineering hires left within 18 months) Why does attrition matter so much in a cost comparison?
Because when an engineer departs after a typical 14-18 month tenure in competitive EU markets, the company has usually not yet recouped the recruiting, onboarding, and ramp-time investment, and must repeat the full cycle – a cost rarely modeled explicitly in a standard hiring budget.
(Scenario: CFO planning a senior hire into a legacy codebase) How long does it take a new in-house hire to become fully productive?
Meaningful productivity ramp typically takes eight to sixteen weeks depending on role seniority and codebase complexity, representing roughly one additional month of fully-loaded cost that is almost never itemized separately in a hiring budget.
(Scenario: CFO comparing a vendor quote against internal salary bands) Does an outsourced team’s monthly rate already include overhead costs?
Typically yes – a dedicated team’s rate usually already folds in the vendor’s own recruiting, benefits, and management overhead, which is why it should be compared against the fully-loaded in-house figure rather than gross salary alone for a fair comparison.
(Scenario: CTO arguing to keep core IP work in-house) Is in-house hiring ever the right call despite the higher total cost?
Yes, particularly for roles close to core IP or requiring deep, hard-to-transfer institutional context, where long-term ownership and control outweigh the cost differential. The goal of a complete cost model is an honest comparison, not a predetermined conclusion favoring either option.
(Scenario: CFO building a board deck comparing five in-house hires against a dedicated offshore team) How do I scale this per-engineer model up to a full team comparison for a board presentation?
Multiply the fully-loaded per-engineer figure by headcount on both sides rather than presenting a single-engineer example and asking the board to extrapolate, since ramp-time and attrition costs do not scale perfectly linearly once a team is large enough to have overlapping onboarding cohorts. Present the total year-one and steady-state year-two figures separately, since ramp and initial recruiting costs are front-loaded and understate the ongoing steady-state gap if shown only once.
(Scenario: CFO whose finance system already tracks recruiting fees separately from salary) My finance system already tracks recruiting fees as a separate line – doesn’t that mean I’ve already captured this cost?
Tracking the fee is not the same as attributing it to the specific hiring decision under evaluation – most finance systems log recruiting spend to a general HR cost center, not amortized against the specific engineer’s expected tenure and folded into that role’s true annual cost. Pull the actual fee paid for that specific hire and divide it by realistic expected tenure in years to get the annualized figure this model requires.
(Scenario: CFO asked why the offshore comparison rate differs from a public rate card) Why might the offshore dedicated team rate I’m quoted differ from the blended rate used in this worked example?
Rates vary by seniority mix, tech stack specialization, and team size – a team weighted toward senior engineers or a niche stack like blockchain or embedded systems will quote above the blended $4,000-$4,500 range used here, while a junior-weighted QA or support team will quote below it. Always request a rate breakdown by seniority level for your specific team composition rather than assuming a single blended figure applies universally.
(Scenario: CFO wants to model a mid-year hire rather than a full year) How do I adjust this model for a hire who joins mid-year rather than at the start of the fiscal year?
Prorate the ongoing costs – salary, contributions, benefits – to the actual months employed, but do not prorate the recruiting fee or ramp-time cost, since both are largely fixed regardless of when in the year the hire starts. This means a mid-year hire’s effective cost multiplier against their prorated salary will look higher than the full-year 1.4x-1.8x range, since the fixed costs are concentrated into fewer months of output.
Suggested Posts
How useful was this post?
Share it with your friends
Get our latest articles here!
Do you have any questions?
Help us improve the content of this Insightful blog by asking us questions. Manifera's team of experts will help you answer these questions as soon as possible.








