Let's talk






Reporting obligations and regulatory reporting
For fund accounting, tax and operations at KVGs, depositaries and custodian banks – and for vendors of reporting modules. We design the workflows behind MiKaDiv, tax reporting and submissions to the BZSt so that expert users can find, check, correct and release records on time. The result: reports that go out punctually, traceably and without a parallel spreadsheet.
Beabstracto is a UX studio in Munich for complex expert systems. For regulatory reporting – MiKaDiv under sections 45b and 45c of the German Income Tax Act (EStG), tax reporting, submissions to the Federal Central Tax Office (BZSt) – we design the workflows in which expert users find, validate, correct, release and track reports. We add implementation support on the client side: as-is analysis, functional concept, test concept and UAT, key-user training, go-live and hypercare. Our clients are KVGs (German asset management companies), depositaries and custodian banks, as well as vendors of reporting modules. An as-is analysis delivers the first reliable results within a few weeks.
We work as UX consultant and business analyst for regulatory reporting at the interface of business department, development and vendor.
The format is known, the interface is in place – and it still gets tight on the deadline. These are the situations we hear most often.
MiKaDiv reports have fixed deadlines in the following year: 31 March, 30 April, 31 July. The path there still runs through Excel, e-mail and word of mouth. Who released what, and when, is recorded nowhere.
Data arrives from depositaries and sub-custodians at different depths and with different histories. The chain is supposed to be traceable per holding and per income event – often it exists only in the heads of a few individuals. An error here hits both the report and the reclaim.
The system returns error codes that nobody in the business department can interpret without a developer. Corrections end up in tickets instead of in the workflow. Under SLA pressure, people guess instead of checking.
The internal policy requires a release, but the tool has no checkpoint. So the file goes out by e-mail and someone replies ‘ok’. For an audit, that is not an audit trail.
The vendor delivers the reporting module and the business department is expected to use it productively from January. Functional concept, test cases, key users and training are missing – and nobody translates between business department and development.
Regulatory reporting has no moving targets. These dates are fixed.
BZSt / section 52(44b) EStG
BZSt, MiKaDiv FAQ / sections 45b, 45c EStG
Directive (EU) 2025/50, OJ L of 10 January 2025
Case: MiKaDiv reporting module (/cases/raquest)
Three services that can be combined – depending on whether you are introducing a reporting module, building one, or making your existing reporting process robust.
We design the workflow in which expert users process large volumes of reports under SLA pressure: an explicit status model per record, filters and bulk actions for high-volume operation, an error mapping that translates validation errors into business language, and a four-eyes checkpoint before submission to the BZSt. Every decision remains traceable in the audit trail. For vendors of reporting modules we work as an external UX consultancy on the product; for KVGs and depositaries, on process and configuration.
We design the workflow in which expert users process large volumes of reports under SLA pressure: an explicit status model per record, filters and bulk actions for high-volume operation, an error mapping that translates validation errors into business language, and a four-eyes checkpoint before submission to the BZSt. Every decision remains traceable in the audit trail. For vendors of reporting modules we work as an external UX consultancy on the product; for KVGs and depositaries, on process and configuration.
As business analyst and requirements engineer, we translate the reporting obligation, the communication handbook and internal policies into requirements that development and vendors can implement. We model processes end to end (BPMN), write user stories with acceptance criteria and derive test cases from them. The regulatory mapping – which requirement forces which system behaviour – is documented, not assumed.
We support the introduction of a reporting module on the client side – from the test concept through UAT to go-live and hypercare. Key users are involved and trained early so that the knowledge stays in-house. The measures are simple: reports released on time, fewer queries to the vendor, no more parallel spreadsheet.
These terms shape everyday reporting work. We speak them so that workshops do not have to start with the basics.
Six steps that recur in every reporting project – on the client side as on the vendor side.
We observe how the report actually comes into being today: sources, Excel lists, word of mouth, releases. Together with fund accounting, tax and operations we document process, roles and pain points. Output: process picture, risk points, prioritised open questions.
a few weeks
Glossary, taxonomy and the assignment of requirement to system behaviour – sections 45b and 45c EStG, communication handbook, internal policies. Business department, development and vendor agree on the same terms before any screens exist.
End-to-end processes instead of feature lists: Manage Reports, report detail, correction run, release. From these come the status model, error mapping, information architecture and user stories with acceptance criteria.
A clickable prototype with realistic data, tested with the people who will later work against the deadline. We check whether status, errors and release are understood without explanation – and correct before implementation.
Test cases from the functional concept, UAT with key users, training materials for the business department. Acceptance means: a report runs from data delivery to release without a detour through Excel.
Cutover on the effective date, support through the first reporting cycles, sharpening of statuses and error texts from real operations. Handover to regular operations with work instructions – and proof that the parallel spreadsheet is gone.
From our UX consultancy for the vendor of a withholding-tax reclaim software – and from reporting processes where the format was long settled but the data was not.
Domain knowledge was spread across developers, product owner and management; personas were implicit and unprioritised. A glossary, a taxonomy and the regulatory mapping (BZSt, ELMA, MiKaDiv, SLAs) brought everyone to the same table. Only then could interface behaviour be argued about productively. Start with screens and you debate opinions; start with terms and you debate decisions.
Reporting was initially discussed as a list of functions. When we drew the flow end to end – data delivery, validation, correction, release, submission, feedback – failure points and SLA risks became visible that no single feature would have covered. Failure points became UX tensions, tensions became design principles, principles became interface behaviour.
A validation error that appears only as a code creates a ticket. An error that says which field at which depositary is affected and what the correction path looks like creates a correction. The error mapping was one of the foundations that let developers model filters, validation, bulk actions and the report-detail structure concretely.
A fully automated submission reports even when the data would actually say ‘no’. A four-eyes checkpoint before submission costs minutes and prevents a systematic error in the custody chain from carrying into the report. Anyone who has once seen a report corrected before submission trusts the next one more – and does not build a parallel spreadsheet.
A high-fidelity prototype of the Manage Reports journey was treated by stakeholders at large banks as production software. It clarified the MVP scope, gave developers a concrete model for report detail and audit trail, and made every UI decision traceable to persona, risk scenario and principle. Concepts rarely convince experts; experienceable flows do.
Public cases and articles from the withholding-tax and reporting environment. No names of banks, KVGs or depositaries.
Withholding-tax reclaim software
As an external UX consultancy for the vendor of a withholding-tax reclaim software, we developed the UX foundations for a new MiKaDiv reporting module: shared language, personas, process and error mapping, information architecture and a prototype of the Manage Reports journey. Eleven reusable foundations, from glossary to KPI framework.
→Article
What is operationally hard about MiKaDiv, why granularity and quality of the custody chain are the sensitive point, and which checklist helps in the months before the effective date.
→Article
The parallel Excel list arises where deadline, missing document and case status are not in the same place – usually at the interface between back-office and tax system.
→We orient our work on these sources; we claim no certification and no institutional relationship.
Federal Central Tax Office (BZSt) · 2026
Procedure description, FAQ and technical handbook (XML format, message structure, deadlines).
German Bundestag, Federal Law Gazette I p. 1259 · 2021
Introduces the legal basis of MiKaDiv; after postponement by the Annual Tax Act 2024, applicable from 1 January 2027.
Council of the European Union · 2025
eTRC, relief at source, quick refund within 60 days and a register of certified financial intermediaries with reporting obligations; applies from 1 January 2030.
ISO/TC 159/SC 4 · 2020
Seven interaction principles – including suitability for the task, self-descriptiveness, robustness against use errors – against which we check status model and error texts.
ISO/TC 159/SC 4 · 2019, confirmed 2025
Human-centred design across the lifecycle; our framework for as-is analysis, prototype and evaluation with expert users.
Kate Kaplan, Nielsen Norman Group · 2020
Five layers of complexity; regulatory reporting hits above all the integrative, the information and the institutional layer.
Kim Flaherty, Nielsen Norman Group · 2024
Three mechanisms for status feedback; the basis for how validation errors and report status are communicated in the interface.
Jakob Nielsen, Nielsen Norman Group · 1994, last reviewed 2024
Visibility of system status, error prevention and help with error recovery – the heuristics most often violated in reporting processes.
An initial conversation without a sales pitch. We listen, look at your process and tell you honestly whether and where we can help.
Book an initial conversation




