🇩🇪Germany
Germany is the largest economy in the European Union and structurally the hardest market in Europe for a foreign agentic accounting tool to enter on its own. Three facts set the shape: DATEV — a cooperative owned by tax advisors, with tens of thousands of Steuerberater members — is the standard interface between a German SMB and its external accountant; the SKR03 and SKR04 charts of accounts, both published by DATEV, are near-universal; and the phased B2B e-invoicing mandate under the Wachstumschancengesetz requires every German business to be able to *receive* structured e-invoices from 1 January 2025, with issuing obligations phasing in through 2027–2028. The cloud cohort — sevDesk, Lexware Office, Buchhaltungsbutler, Candis — is where agentic features are shipping fastest, but none of them can route around DATEV.
Germany is the market where "agentic" has to be translated before it can be sold. The dominant bookkeeping workflow — SMB captures receipts, monthly DATEV-format export to the Steuerberater, Steuerberater reviews and files — was designed around a very specific division of labour that DATEV has protected for fifty years, and it is not obviously improved by handing more judgement to an AI agent. The German buyer's first question is rarely "can this automate my journal entries" but "is this GoBD-compliant, does it export to DATEV, and can my Steuerberater sign off on the audit trail." That filter is what sevDesk, Lexware Office, and Candis have had to clear before they could sell a single agentic feature. The 2025–2028 e-invoicing rollout is the forcing function that changes this — every German business now has to handle structured XRechnung / ZUGFeRD inputs and outputs, and that creates a natural surface area for agentic processing (validation, GL mapping under SKR03 / SKR04, tax-code assignment, archiving). The tools that will matter in Germany through 2027 are the ones that treat the e-invoicing mandate as infrastructure and let agentic features compound on top. Xero, Intuit, and the US agentic-native cohort are essentially absent; this is DATEV's market to lose and sevDesk's to earn.