In early summer 2026, 36 per cent of the companies surveyed by the Association of German Chambers of Industry and Commerce named the shortage of skilled workers as a risk to their business development. A year earlier the figure was 45 per cent. Anyone who concludes from this that vacancies are now filled more quickly is confusing two different measurements — and planning their staffing needs on the wrong one.
The DIHK risk list in early summer 2026
The chambers of industry and commerce regularly ask companies where they see the greatest risks to their own economic development. For the early summer 2026 evaluation, the DIHK received around 23,000 responses via 79 chambers. Each risk is assessed separately, multiple answers are possible, and the order follows from the share of companies naming a given item.
For the items put to all companies, the DIHK reports the following values: energy and raw material prices 70 per cent, economic policy conditions 58 per cent, labour costs 57 per cent, domestic demand 56 per cent, shortage of skilled workers 36 per cent, financing 15 per cent.
Two further items in the survey — foreign demand and exchange rates — are left out here. The report explicitly marks them as values collected only in the export industry. They therefore rest on a different population and cannot be sorted into the same ranking. Anyone who counts them arrives at a different placement, and that placement would be wrong.
What is interesting about this list is less the position than the movement behind it. In early summer 2025, 45 per cent of the companies surveyed by the DIHK named the shortage of skilled workers; at the start of 2026 it was 40 per cent; in early summer 2026 it is 36 per cent. The position in the list has not changed: in all three waves the shortage of skilled workers ranks fifth. What has changed is the share of companies naming it at all.
What a share measures — and what it does not
A risk survey with multiple answers records separately, for each item, how many companies tick it. The result is a statement about perception among the companies surveyed, not about the availability of staff. The two can diverge without either figure being wrong.
Why the share has fallen by nine points in twelve months is not something the survey reports. It records mentions, not their reasons. It therefore cannot be read from the decline that an individual vacancy is filled faster today than a year ago.
A second survey by the same institution asks a different question. For the DIHK Skilled Labour Report 2025/2026, more than 21,000 companies were asked in autumn 2025 whether they are unable to fill vacancies, at least in part. Among health and social service providers, 61 per cent say so according to the Skilled Labour Report, in the construction sector 49 per cent after 53 per cent the previous year, and in hospitality 40 per cent.
This survey dates from autumn 2025 and therefore sits between the economic waves of early summer 2025 and the start of 2026. It does not explain their course but describes a point of its own. For hospitality and for health and social service providers the report shows no prior-year value; no direction can be derived for those two sectors.
These values, too, count companies and not positions. The 61 per cent in the Skilled Labour Report mean that roughly three in five health and social service providers surveyed were unable to fill at least one position. How many positions were affected, and how long they stayed open, the figure does not say.
183 days: a duration instead of an assessment
There is one indicator that works without self-reporting. The statistics department of the Federal Employment Agency (BA) reports the vacancy duration for registered job openings. According to its glossary, this is, at the point a position leaves the register, the difference between the date of that departure and the earliest possible starting date. The clock therefore starts on the day the company would first have needed the person.
For the period from April 2025 to March 2026 the BA reports the following values: at skilled-worker level 183 days, across all requirement levels 165 days, for specialists 157 days, for unskilled and semi-skilled work 156 days, and at expert level 109 days.
The four requirement levels describe the complexity of an activity, not the qualification of the person performing it. The helper level covers unskilled and semi-skilled work; the skilled-worker level covers occupationally focused activities for which vocational training of at least two years is normally expected. Above them sit the specialist level, with a master craftsman, technician or bachelor’s qualification, and the expert level, with a degree course of at least four years.
The longest duration therefore does not fall on activities requiring a university degree but on those requiring vocational training. Why the order comes out this way is not something the statistics report. They measure durations, not causes.
Two things belong to the interpretation. The first is the period: the value is a rolling twelve-month average and therefore largely covers the time in which the DIHK risk share fell from 45 per cent in early summer 2025 to 40 per cent at the start of 2026. The two periods are not congruent: the measurement window begins before the first of those waves and ends after the second.
The second is the population. Measurement takes place when a position leaves the register — that is, when it is filled or withdrawn. Positions that remain open do not enter the average. In which direction this shifts the average cannot be derived from the definition alone. It is a planning figure with variation upwards and downwards, not a commitment for the individual case.
Six months — and when they start counting
The 183 days reported by the BA are around six months. They start at the earliest possible starting date. The time spent before that on the decision, the job description and the advertisement is not included — it comes on top.
Against this stands the route via international recruitment. According to our own calculation it takes approximately four months, counted from the completed staffing requirement analysis and the signed placement agreement.
Three stages carry the largest part of this time: candidate search and pre-qualification at about two to four weeks, the authority procedure at around six to eight weeks for the work permit, and the visa procedure at about four weeks. Added to this are the candidate presentation, for which we ask for feedback within 14 calendar days, the collection and legalisation of documents, and arrival, registration and the start of work. These stage durations are values from our own placement process and not figures issued by an authority. Information provided for orientation, not legal advice.
The two periods start at different points and are therefore not added together. The vacancy duration runs from the earliest possible starting date of a registered position; the four months run from the agreement and the completed requirement analysis. For occupations whose foreign qualification must be recognised in Germany, the recognition procedure comes on top. The honest sentence is therefore not that the international route is faster. It is: both routes take months.
Anyone who takes both routes one after the other does add the periods together. The position stays open first because the domestic market is to be served; only after months without a result does the placement begin, and its four months only start running then.
The difference between the two routes therefore lies less in speed than in predictability. How long a search on the domestic market will take, you only know once the position is filled. The placement process, by contrast, consists of seven named steps whose status you can ask about.
What this means for your staffing plan
The figures do not lead to a recommendation for a particular route, but to a recommendation about timing. The need does not arise at the moment a resignation lands on your desk, but at the moment you can foresee it. Three steps turn that insight into a plan.
First: determine the requirement level. Does the position call for completed vocational training, or is it unskilled or semi-skilled work? This distinction decides the legal route, the duration, and whether a recognition procedure becomes necessary.
Second: clarify the legal route in advance. For skilled workers with vocational training, Section 18a of the Residence Act applies. It is tied to the general requirements of Section 18 (2): a concrete job offer, the approval of the Federal Employment Agency where required, the established equivalence of the qualification acquired abroad, and, for regulated professions, the licence to practise. For unskilled and semi-skilled work, the Western Balkans regulation under Section 26 (2) of the Employment Regulation may apply, likewise tied to a concrete job offer on standard market terms, to the approval of the Federal Employment Agency, and to an annual quota.
Third: count backwards. Set the date on which the person is to start work in your company and subtract the process duration from there. The result is the day on which the requirement analysis has to be in place — not the day on which you start thinking about it. If the person is to be fully trained in by a particular date, add the induction period in front of that as well.
This backward calculation is the difference between staff planning and a staff search. Structure beats chance.
Free initial consultation — we call you back, when suits you? In 15 minutes we clarify the requirement level, the legal route and a realistic schedule for your vacancy.
Further reading
- Filling helper roles: 156 days on average — the vacancy duration for unskilled and semi-skilled work in detail
- Duration of international recruitment: realistic timelines — the seven steps with their individual durations
- Time to fill by federal state: 72 days apart — how far the duration varies by region
- Building a candidate pipeline: making recruiting predictable — how recurring demand turns into lead time
Sources
- DIHK: DIHK-Konjunkturumfrage Frühsommer 2026. 2026. https://www.dihk.de/resource/blob/178144/33116d7ac9ff39cc499320bacef6a246/konjunktur-dihk-konjunkturumfrage-fruehsommer-2026-data.pdf (retrieved 10 September 2026).
- DIHK: DIHK-Fachkräftereport 2025/2026 — Fachkräfteengpässe bleiben Herausforderung. 2026. https://www.dihk.de/resource/blob/159714/0ae2cf7cc1789688edf318a7655ad5d4/dihk-fachkraeftereport-2025-2026-data.pdf (retrieved 10 September 2026).
- Bundesagentur für Arbeit: Vakanzzeit von Arbeitsstellen nach Anforderungsniveau in Deutschland, April 2025 bis März 2026, compiled by Statista. 2026. https://de.statista.com/statistik/daten/studie/1318407/umfrage/vakanzzeit-gemeldeter-arbeitsstellen-nach-anforderungsniveau (retrieved 10 September 2026).
- Statistik der Bundesagentur für Arbeit: Glossar — Vakanzzeit von Arbeitsstellen. No year given. https://statistik.arbeitsagentur.de/DE/Navigation/Grundlagen/Definitionen/Glossar/Glossar-Nav.html (retrieved 11 September 2026).
- Statistik der Bundesagentur für Arbeit: Methodische Hinweise — Anforderungsniveau eines Berufes. No year given. https://statistik.arbeitsagentur.de/DE/Statischer-Content/Grundlagen/Methodik-Qualitaet/Methodische-Hinweise/uebergreifend-MethHinweise/Anforderungsniveau-Berufe.html (retrieved 11 September 2026).
