In freelancermap's Freelancer-Kompass 2026, 43% of freelancers said they had no secured workload for the coming months, and about half said their order situation was worse than a year before. That is the headline figure. The same first wave, published on 9 December 2025, also asked where running projects get stuck: 55% of freelancers named unclear requirements, 47% delayed feedback and 42% missing decisions. freelancermap restated those answers in a release on 1 September 2026, where the 42% reads as a lack of decision authority on the client side. In my view those three figures explain more of the 2026 market than the headline.
I work as an independent firewall and network-security contractor in German enterprises. Firewall change is the purest test of the decision problem, because no rule goes live until someone with authority accepts the risk. When that person is missing, the engineer waits, and the client pays the full rate for the waiting. This post reads the Freelancer-Kompass from that seat. It does not cover rates, which I wrote about in what twenty years of experience is worth per hour, or the consultancy market, covered in the German consulting market is sorting, not shrinking.
What does the Freelancer-Kompass 2026 actually say?
The Freelancer-Kompass 2026 is freelancermap's annual survey of freelancers in the German-speaking market, run in several waves. The headline numbers come from the first wave, which ran from 17 November to 3 December 2025 with more than 1,300 participants and was published on 9 December 2025. The full study, according to freelancermap's study page, covers around 5,400 freelancers surveyed between 17 November 2025 and 8 February 2026. The figures are almost a year old now, and the September release reuses them, so treat them as the state of the market going into 2026.
| Finding | Value | Source and basis |
|---|---|---|
| No secured workload for the coming months | 43% | First wave, 1,300+ freelancers, Nov–Dec 2025 |
| Order situation worse than the year before | about 50% | First wave |
| Freelancers in IT and software reporting falling demand | 23% | First wave; automotive: 32% |
| Unclear requirements and briefings as the top challenge | 55% | First wave; restated in the release of 1 Sep 2026 |
| Delayed feedback | 47% | Same as above |
| Missing decisions, worded in Sep 2026 as lack of decision authority at the client | 42% | Same as above |
| At least one project lost over sham self-employment concerns | 25% | Release of 1 Sep 2026 |
freelancermap's CEO, Thomas Maas, framed the first-wave result as structural. Paraphrased: when orders fall across this many sectors and nearly half of freelancers have no planning security, the risk is no longer an individual one. I agree with the diagnosis and read the cause differently.
Is the problem a lack of demand for security skills?
No. The broader labour-market data says security skills are still the most sought-after in German IT, even after a sharp fall. The Hays Fachkräfte-Index IT for Q2 2026 put IT security specialists at an index value of 373, the highest of any IT role, down 136 points from the previous quarter. The overall IT index fell 37 points to 59. t3n reported the same figures on 29 August 2026. One caveat matters here: the Hays index measures job advertisements against a 2015 baseline, so it mainly describes hiring of employees, not the freelance project market.
The freelance project market tells a narrower story. freelance.de counted about 14,700 projects posted in Q2 2026, up from about 13,400 a year earlier. IT stayed the largest category, but its share dropped from 33.4% in Q1 2026 to 27.9% in Q2, which freelance.de puts at about 21% fewer IT projects in absolute terms than in Q1. Read together, my interpretation is this: companies still want security people, more of them now try to hire them as staff, and fewer projects reach the open market at all. A skill that is in demand can still sit idle if the projects that need it never get signed.
Why do firewall projects stall on the client side?
In my experience, firewall projects stall on the client side because every meaningful change needs a risk owner, and many organisations have not named one. A migration, a rule clean-up or a new segment touches applications owned by other teams. Each of those teams must confirm which flows they still need, and someone above them must accept that a flow will be cut. The 42% of freelancers in the freelancermap data who report a lack of decision authority at the client describe, in my reading, this gap. In firewall work that gap sits on the critical path.
I see the pattern in the same three places on most engagements. The briefing describes the target platform in detail and the decision process not at all. The application owners answer late, because reviewing old rules is nobody's objective for the quarter. And the change advisory board, which should decide, forwards the question back to the technical team because nobody in the meeting can accept the risk of switching off an unknown flow. This is the same structural failure I described in security controls that fail silently: the control exists on paper, and the owner who should exercise it is absent.
| Survey finding | How it shows up in a firewall project |
|---|---|
| Unclear requirements (55%) | Scope says “migrate the rule base”, without saying whether unused rules may be removed |
| Delayed feedback (47%) | Application owners take weeks to confirm which flows are still needed |
| No decision authority (42%) | The change board cannot approve removing a flow nobody can explain |
| Sham self-employment fear (25% lost a project) | The contractor is kept out of the meetings where the decisions are made |
How does the sham self-employment fear change a project?
The fear of Scheinselbstständigkeit, sham self-employment, already costs freelancers work directly. According to the same September release, a quarter of the freelancers surveyed had at least one project fall through for that reason, and 11% had already been affected by an audit. Among the companies surveyed, 59% named sham self-employment as the biggest legal weakness in working with freelancers and 31% named temporary agency work rules. The release draws on both the 2025 and the 2026 editions of the study.
My read is that the fear does a second kind of damage that the survey does not measure. To keep a contractor looking independent, some clients keep the contractor away from internal decision meetings, team channels and direct reporting lines. That is a sensible legal instinct. In a firewall project it also removes the one person who understands the rule base from the room where the risk gets accepted. The decision then takes longer, and the contractor's independence is protected at the cost of the project. The reform debate I covered in Statusfeststellung reform: certainty at 16.74% of the fee would not fix this, because the problem is less the legal test than the way clients organise around it.
What should a client settle before bringing in a firewall contractor?
A client should settle who decides before deciding who to hire. freelancermap's own advice in the release points the same way: clear deliverables, a timeline and agreed communication paths before the project starts. For firewall work I would make it more specific, because every week spent waiting for a decision lands on the client's invoice.
- Name the risk owner in the statement of work. One person who can approve removing a flow, with a deputy for holidays.
- Agree a default for silence. For example: a flow with no confirmed owner after a set period is logged, then disabled during a change window with a tested rollback.
- Separate the decision from the engineering. The contractor prepares the evidence and the options, the client decides. That split also supports the contractor's independent status, because the client keeps the decision and the contractor delivers a defined result.
- Budget the application owners' time. Rule reviews fail when they are an extra task on top of a full quarter.
None of this is new in governance terms. The pattern is the same one that kills AI projects, which I covered in why enterprise AI pilots fail on governance: the technology works, the ownership is missing. NIS2 raised the documentation bar for many firms, as I argued in the Mittelstand NIS2 piece, and a documented decision path is part of that evidence.
What does this mean for a senior firewall contractor?
For a senior firewall contractor, my reading of the 2026 data is that the scarce resource is no longer engineering hours but client decisions. In my view, the work that keeps moving in this market is work with an external deadline that forces a decision: an end-of-support date, an audit finding, a regulator's request. Projects without one drift. Reading the Freelancer-Kompass, I see a market that still needs security engineers. My interpretation of its numbers: 43% of freelancers lacked secured work while, in many organisations that need them, nobody could agree internally on who signs off.
That is an uncomfortable reading for both sides. Clients can fix most of it before the first day of a project. Contractors can name it in the first week instead of absorbing it for six months. The freelancermap numbers give both sides a shared vocabulary for a problem that, in firewall projects, has always been there.