Home
Industries
Administration & Back Office

Cross-industry · Administration · Back Office & Process Consulting

A bad process with new software is still a bad process.

Only more expensive. That is the sentence almost every digitalization project here begins with – and the reason we draw up a deletion list before a feature list. Administration and back office are the area where companies and public bodies have exactly the same problems: cases that run on paper, sign-offs by email, Excel as the unofficial line-of-business system. We analyse what can go – and build something lean for the rest.

See the approach Request a current-state analysis

Consulting and delivery

We do not write a recommendation that then has to go out to tender. We analyse, decide with you and build – which shortens the path from insight to effect considerably.

Deleting is the cheapest win

The cheapest project success is a process that disappears without replacement. We look for those actively, even when it shrinks our own scope of work.

Strategy, not just process

We do not stop at the question of how a process can be made faster; we go as far as whether a company should still be doing it itself in five years.

In short

Who this is for: cross-industry – companies of any size and public bodies. Anywhere administrative work ties up time without creating value.
Typical trigger: a form that is filled in, signed, scanned and typed up again. A sign-off that travels by email. An Excel file that has become a line-of-business system.
What we do: a current-state analysis on the real process, a deletion list before the feature list, AI-assisted capture and document extraction, lean web applications for the rest – plus process and strategy consulting.
The difference from classic consulting: we deliver it ourselves. A good analysis that vanishes into a backlog has helped nobody.

01 — Starting point

The paperwork everyone knows about and nobody touches.

These processes appear in no project portfolio, because individually they look too small. Together they cost more than most digitalization programmes.

The form that gets handled four times

Filled in, printed, signed, scanned, typed up. Five steps for one piece of information – and every one of them is a place where a case can stall.

Sign-offs by email

An application travels through inboxes as an attachment. Nobody knows whose turn it is. Who has already approved is somewhere in the quoted history.

Excel as the unofficial line-of-business system

For everything no system covers, a spreadsheet appears. It grows, it gets passed along, and eventually a whole process hangs on a file that one person understands.

Double entry between systems

Data from PDFs, email attachments and portals is transferred into the next system by hand. One person spends half a working day on it – every day.

Processes without readers

A report is produced monthly because it is produced monthly. Nobody can say who reads it. When you ask, it often turns out: nobody.

Digitalization as an end in itself

“We are digitizing process X now” – without first checking whether process X is still needed. The result: the redundant process is digital, faster and more expensive than before, maintenance contract included.

02 — Approach

Current-state analysis, deletion list, delivery.

In that order. The first two steps decide whether it succeeds – the third is craft.

Observe rather than ask

We look at how a process actually runs – not at how the process description says it runs. In almost every company those two differ considerably, and the truth is in the practice, not in the document.

The four questions

For every process: what was it introduced for? Who reads the output? What happens if we drop it without replacement? And if it has to stay – which field in it is actually evaluated? Some forms do not survive those questions.

Deletion list before feature list

What falls away is not built, not maintained and not trained. It is the cheapest and fastest win in the entire project – and the only one almost nobody actively looks for, because it brings nobody any revenue. Not us either; we do it anyway.

Prioritize by pain, not by visibility

The process that eats the most time is rarely the one that gets talked about in meetings. We prioritize by effort and frequency, not by attention.

Build the smallest sensible slice

One process, one department, real operation. What holds there gets extended – what does not hold has cost little. Concepts never tested in real use rarely survive.

Use AI where it genuinely carries

Digitizing legacy records, extracting documents, structuring free text, drafting: AI is the right tool for that today. For a clean dataset and a well-designed form it is not – that stays craft.

03 — User stories

What comes out of it in practice.

Use cases we meet across industries – in companies as much as in public administration.

“As an employee, I want to submit a request without printing anything.”

A guided online journey with a status display and automatic routing to the right approval level. The case stays visible instead of disappearing into inboxes.

“As a manager, I want to approve without reading an email chain.”

An approval view with everything needed at a glance: what, why, from whom, and what has already been agreed. One click, documented and traceable.

“As back office, I do not want to retype invoices.”

AI-assisted document extraction with a human check: the process gets faster without control disappearing. This is the standard case in which AI genuinely saves money.

AI automation

“As an administration, I want to take on legacy files without months of typing.”

Paper records and old forms are read with AI assistance, checked and imported – so a new system starts with its history rather than from zero.

Public sector

“As a team, we want out of the shared Excel file.”

A lean web application instead of a spreadsheet travelling by email: several users at once, versioning, permissions, traceability.

Excel modernization

“As management, I want to know what we actually all do.”

An inventory of the processes with an effort estimate and a recommendation: keep, change, delete, automate. A basis for decisions instead of for gut feel.

04 — Consulting

Why we go all the way into strategy.

Process consulting stops at the question of how a process can run better. The more interesting question comes after it – and it cannot be answered without talking about the business model.

Process consulting: how does it run better? – Sequence, responsibilities, media breaks, waiting times. That is the level most digitalization projects operate on – and it matters, but it is not sufficient.
Strategy consulting: should we be doing this at all? – Some processes exist because a decision was taken that way ten years ago. Whether it is still right today is something nobody checks – because that is not a process question but a question about the business model.
A current-state analysis without deference – What is done today, how, and which part of it is honestly nonsense? Internally that question rarely gets asked, because it is uncomfortable. From outside it is easier to ask – and that is a good part of our value.
The software is the consequence, not the starting point – Once the process question is settled, the software question is usually easy. It does not work the other way round: a tool that meets an unresolved process only sets it in concrete.
We deliver what we recommend – That is what separates us from classic consulting. An analysis that then has to go out to tender loses half its substance on the way to delivery – and in the end nobody is accountable for it.
We say when there is nothing to do – It happens regularly that a process already runs well and a project would bring no benefit. Then we say so. A project sold without effect damages us more than a project not sold.

Service 1

Current-state analysis and an inventory of the processes, with effort estimates and a recommendation

Service 2

Process and strategy consulting – up to the question about the business model

Service 3

Delivery: lean web applications, apps and AI automation

Service 4

Steering when several parties are involved – see IT project management

05 — FAQ

Common questions about process digitalization.

What does a current-state analysis cost?

It depends on the scope – a single process is a different matter from an inventory across several departments. After a short first conversation we say which cut makes sense and what it costs. What we do not do is a months-long analysis phase with a slide deck at the end; the benefit arises in the first few weeks, and after that it costs more than it returns.

You delete processes. Does that not shrink your own order?

Yes, regularly. We do it anyway, for two reasons. First: a project that digitizes a redundant process has no effect – and clients notice, after a year at the latest. Second: the credibility to be able to say “you do not need this” is exactly why clients call us again on the next, larger question.

Where do you use AI and where not?

AI is strong with unstructured input: extracting documents, structuring free text, taking on legacy paper records, drafting, evaluating large volumes of text. It is weak where accuracy and traceability count and the process is structured anyway – a clean form with plausibility checks beats an AI solution there. We say per process which category it falls into, instead of writing AI into everything.

Does our data go into third-party AI models?

No, not without explicit agreement. Which data flows where is part of the design and is settled in advance – including the question of which processing has to stay within your own area of responsibility. How we use AI in our own development process we describe openly under AI automation and on the About us page.

We have already engaged a process consultancy. Are you competing with that?

Not necessarily. What is usually missing in that kind of setup is not another analysis, but somebody who actually implements the recommendations and steers between the business side, IT and suppliers. We are happy to take that role alongside an existing consultancy – and to work with it rather than against it.

How long until something is noticeable?

With well-scoped processes: weeks, not quarters. That is less about us than about the cut – a single process with two user groups is quickly built. What makes projects long is unresolved special cases and coordination across several departments. So we start small and extend, rather than the other way round.

Do you work with public bodies as well?

Yes. The patterns are practically identical; the conditions differ – procurement, accessibility, data protection. We describe a delivered project on our page about the public sector and municipalities, including an honest assessment of when we are not the right partner as a small team.

Which process would nobody miss?

If something comes to mind immediately, you have already made the start. Describe it briefly – we will say whether deleting, changing or building is the right answer.

Request a current-state analysis

06 — Related