Most German CTOs who hesitate about engineering capacity in India are not prejudiced. They are pattern-matching on experiences from a decade ago, on stories from peers, or on engagements that were structured badly from the start. Some of those concerns remain valid. Others no longer describe the market. A few identify a real problem but attribute it to the wrong cause, which leads to the wrong fix. This analysis takes seven common assumptions, separates what is accurate from what is not, and gives a concrete check for each.
The market in figures
Several of the assumptions below were formed when India's technology sector was smaller and narrower. Current figures describe a different market:
- Nasscom's February 2026 strategic review estimates India's technology industry revenue at USD 315 billion for FY26, with a direct workforce of 5.95 million.
- The same review reports that more than 2 million of those employees have been upskilled in AI, including up to 300,000 in advanced AI skills.
- A Nasscom and Zinnov report from May 2026 counts 2,117 global capability centres in India with 2.36 million professionals and USD 98.4 billion in FY26 revenue.
- On the German side, Bitkom's 2026 labour market study counts 79,000 missing IT specialists, and 47 percent of companies cite insufficient German skills as a hiring hurdle. Language is already a constraint at home, not only offshore.
None of this proves that a given vendor is good. It does mean that the questions should be about structure and evidence rather than about the country.
1. "Indian teams do maintenance and testing, not product engineering"
What is true: a large part of the Indian IT services industry was built on application maintenance and testing, and some vendors still sell mainly that.
What has changed: India hosts roughly 1,700 global capability centres according to a 2025 Vestian estimate, about 53 percent of the global total. Many of them run core product development, research and platform engineering for their parent companies. The engineers who do that work are part of the same labour market your vendor hires from.
Check: ask the vendor for anonymised descriptions of greenfield work in your domain, and interview the actual engineers about design decisions they made, not tasks they completed.
2. "The time zone makes collaboration impossible"
What is true: India is 4.5 hours ahead of CET in winter and 3.5 hours ahead during European summer time. Late-afternoon German meetings fall into the Indian evening.
What is overlooked: a standard German working day still overlaps with Indian working hours for roughly four to five hours. That is enough for daily coordination if the process is designed for it.
Check: map your current recurring meetings. If most of them require more than four hours of daily synchronous time, the issue is your process, and it will affect any distributed team, including one in Berlin and Munich.
3. "Our data cannot legally go to India"
What is true: India has no EU adequacy decision, so transfers of personal data are restricted under DSGVO Art. 44 ff.
What is incomplete: restricted does not mean prohibited. Transfers can rely on Standardvertragsklauseln under DSGVO Art. 46 together with a transfer impact assessment and supplementary measures. India has its own Digital Personal Data Protection Act 2023, with implementing Rules notified in November 2025. In many engineering engagements, the team does not need production personal data at all.
Check: classify what data the team actually needs. Synthetic or pseudonymised test data often removes most of the question. Where personal data is necessary, involve your data protection officer early and agree a processing agreement under DSGVO Art. 28.
4. "Offshore is always cheaper"
What is true: engineering salaries in India are lower than in Germany at comparable seniority.
What is misleading: the invoice is not the cost. If your senior engineers spend a third of their time writing detailed tickets, reviewing every change and chasing status, the real cost includes that time and the roadmap work they no longer do.
Check: calculate cost per accepted deliverable including your internal management time, not cost per engineer-month.
5. "Attrition makes continuity impossible"
What is true: engineers in the largest Indian hubs receive frequent offers, and some vendors have high turnover.
What is misattributed: continuity problems usually come from undocumented knowledge, single points of failure and no overlap when someone leaves. Those are structural issues that also hit German teams.
Check: ask how knowledge is documented, how many people understand each component, and what the contractual handover period is when a team member changes.
6. "A team abroad cannot fall under AÜG"
What is true: the classic AÜG scenario involves workers placed physically in the client's premises.
What is risky: the decisive criteria under AÜG §1 are integration into the client's organisation and subjection to its instructions. Integration can happen through daily stand-ups led by your managers, tickets assigned directly to individuals and performance feedback from your side, wherever the engineer sits. Whether and how German rules apply to a purely foreign team is a question for legal review, but the operating pattern should be designed as if it matters.
Check: who assigns work to individual engineers, and who could dismiss them? If the answer is your staff, review the structure.
7. "Indian engineers will not say no"
What is true: communication norms differ, and in some hierarchical environments raising problems with a client is discouraged.
What is misattributed: this is primarily a function of the vendor's leadership culture and your own process. Engineers who are punished for bad news anywhere will hide it.
Check: ask the vendor how a missed estimate was escalated in a past project, and build explicit risk reporting into every sprint review.
How to use these checks
The checks are most useful before a contract, not after. Run them as part of vendor selection: one interview round focused on design ownership, one review of your own meeting load, one session with your data protection officer on data classification, and one cost model that includes internal management time. If a vendor cannot support these checks with concrete answers, that is informative in itself.
Summary table
| Assumption | Verdict | Real lever |
|---|---|---|
| Maintenance only | Outdated | Interview for design ownership |
| Time zone blocks collaboration | Overstated | Process design |
| Data cannot go to India | Incomplete | Data minimisation, SCCs, Art. 28 agreement |
| Always cheaper | Misleading | Cost per accepted deliverable |
| Attrition kills continuity | Misattributed | Documentation and overlap |
| No AÜG exposure abroad | Risky | Who instructs whom |
| No one says no | Misattributed | Vendor leadership and escalation design |
Several of these levers are structural rather than geographic: who steers the work, who reviews it, and who carries the result. The legal points above are general information, not legal advice.
Most of the risks that remain concern how an engagement is built rather than where the engineers sit. In the niogod™ pod, a German-speaking Projektleiter and a Senior Consultant on the vendor side stand between the client and the engineers and carry responsibility for steering and quality (see the pod structure). If one of these assumptions is holding back a decision for next year, testing it against a concrete structure is a sensible next step.
Talk to niogod™
Planning engineering capacity in DACH? Book a 30-minute consultation with the niogod™ team.
Book consultation