A founder sets a $130,000 salary against a $180,000 project fee and concludes the hire is cheaper. Both numbers in that comparison are wrong.
What the seat actually costs
Total cost of an in-house engineer runs about 1.3 to 1.5 times base salary once employer taxes, insurance, equipment and leave are counted. Making the hire adds $9,000 to $25,000, and the median time to fill is 44 days before onboarding starts. Salary is the smallest line in the calculation.
- Loaded cost. A $130,000 engineer lands somewhere near $180,000 to $210,000 a year all in, per KORE1's 2026 hiring cost guide.
- Cost to hire. Sourcing, recruiter time and interview hours run $9,000 to $25,000 per hire when counted honestly.
- Time to useful output. Median 44 days to signature, then onboarding, then ramp. First real delivery is realistically five to seven months out.
- Coverage. One engineer is not a team. Design, infrastructure and product judgment are either separate seats or open gaps.
What the external engagement actually costs
More per hour, and nothing per month. That second half is the part the salary comparison misses.
| In-house hire | Agency or studio | |
|---|---|---|
| Year one cost | Loaded salary plus cost to hire | Project fee, then nothing |
| Time to first delivery | Five to seven months | Weeks |
| Disciplines covered | One | Design, engineering, infrastructure |
| Cost when the work pauses | Unchanged | Zero |
| Knowledge when it ends | Stays | Leaves, unless you plan for it |
That last row is the honest cost of going external, and it is the one worth negotiating rather than ignoring. Knowledge walking out at handover is a real loss. It is also what documentation and a retainer exist to prevent, so ask for both in the contract rather than hoping.
The line where it flips
External delivery wins when demand is variable, short, or needs several disciplines at once. In-house wins when demand is steady, high volume, and concentrated in one skill you will use every month for years. The expensive mistake is not picking wrong. It is picking before you know which situation you are in.
Most companies building a first product do not have steady demand yet. They have one hard question and a deadline. That is the variable case, which is why first builds are so often external and second ones so often are not.
A test that takes five minutes
- Will this skill be needed every month for the next two years? If not, do not buy a seat.
- Can you technically interview for it? If not, you cannot hire for it yet, and hiring blind is the most expensive option here.
- Is the specification finished? If not, hourly capacity will build the wrong thing very accurately.
- What does a three month delay cost you? Put a real number on it, then set that against the gap between the quotes.
Run the comparison on total year-one cost against time to market, and it tends to point somewhere the salary-versus-fee version never did.