Real order intake in main construction rose 23.0 percent in August 2026 compared with July, adjusted for calendar and seasonal effects — the largest month-on-month rise since the series began in 1991. The Federal Statistical Office names the reason in the same sentence: several large orders.

What else is in the same release

The split is uneven: building construction +12.4 percent, civil engineering +31.6 percent against the previous month. In the calmer three-month comparison — June to August 2026 against the three months before — 7.6 percent remains (building +0.6; civil +13.8). Against August 2025, real calendar-adjusted order intake is 17.3 percent higher, nominal 23.8 percent; turnover rose 4.9 percent in real terms.

Four figures, four different windows. Read only the first and you have a record. Read all four and you have a distribution.

Why this matters for the equipment fleet

A large order is paper, not an excavator. It is worked off over months, and it rarely lands where someone is looking for a distribution board right now. For planning that means: the monthly figure works as a signal, not as a planning quantity. The more robust number is in the same release — the three-month comparison, which smooths the large orders out.

The direction is notable: civil engineering is growing much faster than building construction, and year on year the two move in opposite directions (civil +36.4, building −1.7 percent). Civil engineering means line construction sites: little space, changing locations, often no grid connection at all. That is exactly the setup in which mobile power moves along with the job instead of hanging on one box.

What is not here

No forecast. A month with large orders does not define a year, and we do not convert these figures into unit volumes — the assumptions required are not in the release. Anyone turning +23.0 percent into a sales plan has skipped the second figure.

---

*Source: Destatis, order intake in main construction, August 2026 (press release no. 375, 23 October 2026). All figures quoted appear there; the three-month comparison is Destatis's own smoothing.*