A hiring manager sees a €95,000 salary and a €500 day rate and concludes the contractor costs roughly 30% more. The arithmetic behind that conclusion is wrong in three places, and correcting it usually reverses the answer for engagements under twelve months.
The calculation people run
Salary divided by working days. €95,000 over 220 days is €432 a day, against €500 for the contractor. The contractor looks 16% more expensive.
The calculation to run instead
Total annual cost of employment, divided by productive days actually delivered.
Cost side, on a €95,000 base in a Western European market:
- Employer social contributions and payroll taxes: €15,000 to €28,000 depending on jurisdiction
- Recruitment fee, amortised over expected tenure: €12,000 to €19,000 at a 15% agency fee over an 18-month average tenure
- Equipment, software licences, workspace: €3,000 to €6,000
- Bonus, pension, insurance and other benefits: €5,000 to €15,000
Total loaded cost lands between €130,000 and €163,000.
Days side:
- 260 weekday total
- Less 25 days statutory and contractual holiday
- Less 8 to 12 public holidays
- Less 5 to 10 days sickness at European averages
- Less 15 to 25 days of ramp in year one, training, internal meetings and non-delivery time
Productive delivery days land between 190 and 205, not 220.
At €145,000 loaded across 197 productive days, the effective cost is €736 a day. The contractor at €500 is 32% cheaper, not 16% more expensive.
Where the comparison flips back
This is not an argument that contracting always wins. It stops winning when:
- Tenure runs long. Recruitment cost amortises. At three years the employee’s effective day cost falls by roughly €70, and the gap narrows sharply.
- The role is core and permanent. Institutional knowledge accrues to employees in a way it does not to contractors on rolling terms.
- Utilisation is genuinely full. The employee model assumes you can keep one person busy for 200 days on work only they can do. Many teams cannot, and pay for the idle time anyway.
- You need equity to compete. Some engineers will not take a contract at any rate.
The variable nobody prices
Time to productivity. Hiring a senior AI engineer in Western Europe takes 10 to 16 weeks from approved requisition to first commit, and that is before notice periods. An augmented engineer starts in one to three weeks.
If the work has a deadline, those 9 to 13 missing weeks are not a cost line, they are the whole decision. A team that ships a quarter earlier usually justifies the rate difference several times over, and the comparison never appears in the spreadsheet because it is not a cost, it is foregone revenue.
A cleaner way to frame the choice
Stop asking which is cheaper and ask which risk you would rather hold:
| You hold this risk | Employment | Contract or augmentation |
|---|---|---|
| Wrong hire | High, slow to unwind | Low, 30-day exit |
| Demand falls | You keep paying | You stop |
| Demand rises | Slow to add | Fast to add |
| Knowledge walks out | Moderate | Higher, unless you manage it |
| Cost per productive day | Higher under 2 years | Lower under 2 years |
The last row is the only one that appears in a budget, and it is the least important of the five.
What to do with this
Three practical steps:
- Rebuild your own number. Take one real engineer’s total cost last year and divide by the days they actually delivered. Most teams have never done this and are surprised by the result.
- Set the horizon before you compare. Under twelve months, augmentation wins on cost almost regardless of rate. Over three years, employment wins. Between the two it depends on utilisation.
- Price the delay. Put a number on shipping a quarter later. If you cannot, the comparison is incomplete.
Current European day rates by role, market and seniority sit on the daily rate benchmark. If you want the loaded comparison run against a specific team shape, the project cost calculator does it directly.