Offshore engineering can add capacity faster and at lower cost than local hiring. It does not do so in every situation. In some conditions the savings disappear into coordination effort, rework or legal risk; in others the work simply cannot leave the country. A vendor who never says no to a deal is not a reliable adviser. This field note lists eight conditions under which DACH companies should not offshore a given piece of work, or should do so only after fixing the underlying issue, and suggests what to do instead.
1. Nobody on your side can own the product
An external team can deliver against goals, but it cannot invent them. If there is no product owner or technical lead on your side who can decide priorities, write acceptance criteria and accept results, an offshore team will either wait or build the wrong thing. Both are expensive.
Instead: appoint or hire the internal owner first. If that is impossible in the short term, a fractional CTO or interim product lead may be a better first investment than delivery capacity.
2. You want people, not results
If what you really need is an engineer who sits in your stand-up, takes tasks from your team lead and works exactly like an employee, then you are describing employment or Arbeitnehmerüberlassung, not a service contract. Running an external team this way under a Werkvertrag or Dienstvertrag creates the risk that the arrangement is reclassified under §1 Abs. 1 Satz 2 AÜG, with the consequences in §§9 and 10 AÜG.
Instead: hire directly, use a licensed provider of Arbeitnehmerüberlassung, or restructure the work into deliverables a provider can own.
3. The work requires daily physical presence
Some engineering work is tied to hardware in a specific place: commissioning a production line, calibrating sensors on a vehicle, debugging a robot cell on the shop floor. Remote teams can contribute to software, simulation and tooling around this work, but not to the physical part.
Instead: split the work. Keep on-site tasks local and give remote teams the parts that can be reproduced in simulation, hardware-in-the-loop setups or remote-access test benches.
4. The information is classified or cannot be exported
Work involving classified government information (Verschlusssachen) is subject to the German rules on Geheimschutz, which generally require security-cleared personnel and approved environments. Technology listed under the EU dual-use regulation 2021/821 may require an export licence before it is shared with a team outside the EU, and some licences will not be granted.
Instead: clarify classification and export status with your export control officer before any vendor discussion. Offshore only the parts that are demonstrably not controlled, or keep the work within cleared domestic teams.
5. The engagement is very short
Every new team, local or remote, needs time to understand the codebase, the domain and the conventions. If the total engagement is only a few weeks, onboarding may consume most of it.
Instead: for short, highly specialised tasks, an experienced local freelancer with a clearly scoped, independent assignment is often more efficient.
6. Requirements change hourly
Early-stage discovery, where the founders are still figuring out what the product is, relies on dense, fast, informal communication. A remote team across time zones can participate, but the cost of constant context transfer is high.
Instead: keep discovery with a small co-located team. Bring in external capacity when the first stable product slice exists and sprint goals can be written down.
7. The work is your core differentiating IP and you cannot protect it
Some algorithms, models or know-how are the company's core value. Offshoring them is not automatically wrong, but it requires strong contractual rights of use, access controls and a plan for knowledge retention. If your organisation cannot put these in place, the risk may outweigh the benefit.
Instead: keep the core in-house and use external teams for surrounding work such as integrations, tooling, test automation and platform tasks. Revisit the decision once IP protection measures are mature.
8. Cost is the only reason
If the only argument is a lower rate, the business case is fragile. Hidden costs, management effort and transition costs can erase a rate advantage. Companies that offshore purely to cut cost, without a clear capacity or capability need, have little to fall back on when those costs appear.
Instead: first quantify what the work is worth and what the bottleneck costs today. If the case still holds when management time and transition costs are included, proceed. If not, reconsider.
Check the shortage before you assume it
One more check is easy to skip: whether your local market is really as tight for your specific role as the headlines suggest. The data is more mixed than many sales pitches admit. The Bundesagentur für Arbeit identified 157 shortage occupations in its 2025 analysis, but software development was not among them. IW Köln found that open IT positions fell 26.2 percent in 2024 to about 46,400, even though around 70 percent of them could statistically not be filled with suitably qualified jobseekers. The KfW-ifo Fachkräftebarometer for the second quarter of 2026 shows 21.1 percent of German companies reporting that skilled-labour shortages hamper their business, less than half the level of three years earlier. Bitkom, by contrast, still counted about 109,000 unfilled IT positions in 2025. The shortage is real but uneven: if your role is a generalist one in a large city, a well-run local search may close it faster than offshoring.
A quick self-test
Run the following questions with the people who will actually work with the external team, not only with procurement. Engineering leads usually know where ownership is unclear or where requirements still move daily.
| Question | If the answer is no |
|---|---|
| Do we have an internal owner who can accept results? | Fix ownership first |
| Can the work be described as deliverables? | Hire or use licensed Arbeitnehmerüberlassung |
| Can the work be done without daily physical presence? | Split on-site and remote tasks |
| Is the information free of classification and export restrictions, or licensable? | Keep the work domestic |
| Does the engagement last long enough to amortise onboarding? | Use a local specialist |
| Are requirements stable at least per sprint? | Finish discovery first |
| Can we protect the IP involved? | Keep the core in-house |
| Is there a reason beyond the rate? | Re-examine the business case |
A negative answer does not end the discussion, but it marks something to resolve before signing. In our view it is better for a provider to say so early than to discover it after the first invoice.
This field note is general information, not legal advice.
Where the answers are yes, the remaining work is to structure the engagement so it holds up. niogod™ runs it as a three-person pod with a German-speaking Projektleiter, a sprint-based Leistungsbeschreibung and Abnahme, and would rather tell you in the first conversation if one of these eight conditions applies. If you are unsure which side of the line your project falls on, that first conversation is a reasonable place to find out.
Talk to niogod™
Planning engineering capacity in DACH? Book a 30-minute consultation with the niogod™ team.
Book consultation