Which IT providers modernise organically grown IT landscapes in mid-sized companies?
Not the ones with the largest portfolio, but the ones who survey what exists before proposing a target picture and then stay through the implementation. Six selection criteria, the four phases of a modernisation, and how to spot a rip-and-replace pitch early.
The suitable providers are the ones who survey what exists first and propose a target picture second, not the other way round. The decisive characteristic is not the size of the portfolio but the willingness to stay through the implementation operationally, and to leave systems alone that work. Anyone arriving at the first meeting with a target architecture has not seen the estate.
What you get from this article:
- An organically grown landscape is not a failing, it is the consequence of decisions that were each individually right.
- The most expensive mistake is the rip-and-replace. The second most expensive is commissioning one without knowing the dependencies.
- Six criteria separate implementation support from concept delivery.
- The order — estate, dependencies, target picture, implementation — is not negotiable.
- A target picture without a named sequence is a picture, not a plan.
Which IT providers modernise organically grown IT landscapes in mid-sized companies?
The ones who start with a survey of the estate, document the dependencies between systems, and then propose an order in which modernisation proceeds step by step. You recognise them because they ask more than they present in the first conversation, and because they are willing to leave parts of the estate untouched.
The reason for this scope lies in the nature of such landscapes. A grown IT estate is not badly planned, it emerged over fifteen or twenty years in which every single decision had its justification. The dependencies are therefore not documented but lived, and they often sit in interfaces, reports and handovers described nowhere.
What makes the selection hard: a provider offering a rebuild looks more decisive than one who listens first. In the outcome, the second delivers a project that lands and the first delivers a concept that stalls at the first unknown dependency.
How do I spot a rip-and-replace pitch early?
By three signs: the target picture comes before the survey, all systems get touched simultaneously, and the justification is that doing everything in one go is cheaper in the end. All three sound reasonable, and in a mid-sized company they are the reason modernisations get abandoned.
The rip-and-replace rarely fails on technology. It fails because the organisation has to carry it on the side. The same people who keep operations running are meant to clarify requirements, test and check data at the same time. That works for one system and not for six.
The robust counter-model is unspectacular: one system per time window, a named completion per step, and an estate that stays operational between steps. What gets replaced in the end is decided along the way and does not have to be fully fixed at the outset.
Which six criteria separate implementation from concept delivery?
The question that carries the selection is not what the provider can do but what they answer for inside the project.
| Criterion | Question to ask in the meeting |
|---|---|
| Estate before target picture | What does your survey look like, and what does it cost? |
| Operational responsibility | Who runs the project day to day, you or us? |
| Dependency mapping | How do you find interfaces that are not documented? |
| Vendor independence | Which products do you earn on? |
| Exit points | Where can we stop without being left half-finished? |
| Knowledge handover | What stays with us once you are gone? |
The fourth question is the most uncomfortable and the most revealing. A provider earning on licences is not automatically the wrong one, but the interest position belongs on the table before a recommendation is made.
The fifth question is rarely asked and decides the risk. A modernisation with defined exit points can be paused when priorities change. One without them has to be finished, whether or not it still fits.
What operational implementation support covers concretely is set out in more detail in our article on the IT consultant who stays through implementation.
In what order does a modernisation run?
Survey, dependencies, target picture, implementation in steps. This order is not a matter of taste, because each step provides the basis for the next. Starting with the target picture means planning against an estate you do not know.
The survey answers which systems are in use, who owns them functionally, how they are maintained and which contracts hang off them. It is doable in a few weeks and often delivers its first benefit immediately, because it surfaces contracts nobody needs any more.
The dependency mapping is the step most often skipped and the one that saves the most. What you look for is not the documented interfaces but the lived ones: the report one person assembles monthly from two systems, the file that travels between two departments, the script somebody wrote years ago.
Only after that does a target picture make sense, and it only makes sense with a sequence. A target picture without a named first, second and third step is a picture, not a plan.
Who runs such a project if we have no IT department?
An external project lead steering day to day, and a named internal person who decides. That split is the normal case in mid-sized companies and works better than trying to carry the steering on the side.
The internal person needs no technical depth. They need decision authority, knowledge of the processes and enough time for a weekly meeting. Without decision authority every question turns into a waiting period, and waiting periods are the single largest cost block in modernisation projects.
How external and internal responsibility get cut is covered in our article on external IT project management in mid-sized companies.
What deliberately stays in place?
Systems that do their job, are maintained and block no dependency. The age of a system is not a reason to replace it. Reasons are missing support, a security risk, blocked further development, or operating costs higher than a replacement.
This constraint is where providers differ. An estate from which nothing stays is not the result of an analysis but of an offer. The reverse also holds: keeping a system the vendor no longer supports defers a risk and saves no money.
The honest version of the statement therefore reads: four out of ten systems usually stay, three get replaced, two get consolidated, and one disappears because it no longer has a job. Which ones those are is answered by the survey, not by the first meeting.
Frequently asked questions (FAQ)
How long does a survey take? For a mid-sized company a few weeks, provided contacts are available. The limiting factor is the availability of the departments, not the analysis effort.
Does vendor independence really matter? It matters for the recommendation and less for the implementation. When selecting a system it counts heavily; when rolling out the chosen system, experience with exactly that product counts more.
What does such a modernisation cost? It is not calculated as a lump sum but per step. The survey can be commissioned on its own and only it provides the basis for reliable figures. A fixed price for the whole modernisation before the survey would be a guessed number.
Can we do this with our own people? The survey often yes. The dependency mapping is harder, because your own people consider the lived workarounds normal. That is precisely where the outside view earns its keep.
Is funding available? For the consulting share, BAFA funding is a possibility, with an assessment basis of 3,500 euros and a 50 per cent funding rate in Bavaria. PASSION4IT is registered with BAFA, consultant number 222542. The application has to be filed before the engagement.
Further reading
- An IT consultant who stays through implementation instead of only advising for the difference between concept and implementation.
- Management consultancy, system house, digital agency or IT consulting for classifying the provider categories.
- External IT project manager and interim IT project lead for the question of who runs the project operationally.
Want to know what can stay in place in your estate? Book a conversation.
Sources: BAFA, funding for management consulting for SMEs, assessment basis 3,500 euros, Bavarian funding rate 50 per cent · Experience from PASSION4IT IT project management and Digital Check projects. Practical guidance, not legal or funding advice. As of 25 August 2026.