Reading time: 15 minutes | Published in September 2026, expanded in November 2026 | Relevant for: Manufacturing · ERP · abas · Odoo · Budget · Schedule · Project management · CFO · Digital transformation
When an ERP project is signed, everyone looks at the same figure: the amount in the contract. Realising twelve months later that this figure has long been left behind is a shared experience of manufacturers in Türkiye. The budget has doubled, nobody can say exactly where the money went, and the question “where did we spend this money?” hangs in the air in meeting rooms. The interesting thing is this: the overrun usually stems not from one big disaster but from small decisions that pile up unnoticed.
Summary — ERP budget overruns accumulate in four “silent” sources: integrations discovered late, unclear data ownership, overlooked departments and process variants, and scope creep that cannot be stopped. What these four sources have in common, therefore, is that decisions are taken late. Taking a decision at the start of the project is cheap, taking it during implementation is expensive, and taking it after go-live is the most expensive of all. In other words, what determines the cost is not the content of the decision but its timing. In this article we reveal this accumulation layer by layer, like an X-ray film, and look at the questions that should be asked before the project begins.
A budget overrun usually comes with a schedule overrun: the go-live date slips, and the team and the consultants carry the same work for longer. This expanded version of the article also covers why schedules overrun, the integration risks specific to Türkiye, how to manage scope creep and the ways to prevent both overruns.
Contents
- Why is the contract figure misleading?
- Where does cost accumulate? Four invisible sources
- The common thread: decisions are taken late
- Why do schedules overrun?
- Budget risks specific to Türkiye
- What does the X-ray show?
- Before starting the project, ask yourself this:
- 8 ways to prevent budget and schedule overruns
- Scope creep and change management
- Risk register
- Success criteria
- How do you recover a project that has overrun its budget?
- The ERP budget of a small business
- 6 common misconceptions about ERP projects, and the reality
- Frequently asked questions
Why is the contract figure misleading?
The price at the time of signing is the price of the scope known at that moment. The problem is how much of the scope is actually known at that moment.
In most projects it is assumed at the outset that “everything is clear”. The modules have been chosen, the departments have been defined, a date has been set. But this clarity is often only an appearance — less a written, approved scope than a shared optimism. As the project progresses, it turns out that the things thought to be “clear” were in fact assumptions. And every assumption, when it breaks, turns into a cost item.
The root of a budget overrun usually lies not at the end of the project but precisely in this initial illusion of clarity.
Where does cost accumulate? Four invisible sources
1. Integrations discovered late. The sentence “it also needs to integrate with this system” is heard in the middle of the project. Integrations that were not in scope at the start are later added to the bill as both labour and delay. In Türkiye, e-invoice, e-dispatch note, bank and logistics integrations are the classic source of these surprises.
2. Unclear data ownership. Where the data is to be migrated is clear, but who is responsible for cleansing it is not. Faulty data delays go-live; every week of delay means additional cost. If the question “who is preparing the data?” was not asked at the start of the project, it becomes expensive in the middle of it.
3. Overlooked departments and process variants. Three departments are addressed in the first phase, and the fourth is postponed with “we’ll look at it later”. Yet that department’s process depends on the others and cannot be left out. In the same way, it is only during implementation that people discover a single process is carried out in five different ways in the factory. Every variant requires extra effort if it has not been accounted for.
4. Scope creep. As the project progresses, requests of the “since we’re here, let’s add this too” kind pile up. Even if each one looks small, together they make up the second half of the budget. If the boundaries of the scope have not been drawn clearly, there is no limit to stop this creep either.
The common thread: decisions are taken late
Looking closely at these four sources, the same thing stands out beneath all of them. Cost arises not from badly written code but from a decision taken late.
Making a decision at the start of the project is cheap — you are still on paper, and changing it is as easy as a sentence. Making the same decision after implementation has begun is expensive, because something that has already been built now has to be taken apart and rebuilt. Making it after go-live is the most expensive of all. In other words, the cost curve is shaped not by the content of the decision but by its timing.
A situation frequently encountered in manufacturing companies in Türkiye is this: the technical team “gets down to work” straight away, while uncertainties that require decisions are postponed on the grounds that they will be “sorted out along the way”. Every uncertainty sorted out along the way costs many times more than it would have cost had it been resolved at the start.
Why do schedules overrun?
Budget overruns and schedule overruns grow from the same root, but where the timeline slips needs to be seen separately, because every week of delay also extends the cost of the team, the consultants and the old system. In our assessment, the timeline slips most often at these five points:
- Decisions taken late. While a process design waits for approval, the configuration, testing and training steps that depend on it wait too. A single pending decision can stop several teams at once.
- Data migration and data cleansing. Item records, bills of materials, customer and supplier accounts and open balances only show their real state in a trial migration. If cleansing was not planned at the start, it lands on the critical path just before go-live. That is why a data cleansing plan belongs in the first weeks of the project.
- Testing and user acceptance testing (UAT) squeezed at the end. As the timeline slips, testing is the first phase to be cut. Insufficient testing carries problems into live use, where fixing them takes longer and costs more.
- Key users who cannot be freed from daily work. The people who know the process best also run day-to-day operations. If the time they give to the project is not protected in writing, workshops and tests are postponed again and again.
- Integrations discovered late. When an interface that was not in scope surfaces in the middle of the project, it adds not only workload but also the timeline of the system and the organisation on the other side.
These five have the same thing in common: each is the postponement of a decision that should have been taken, or of a responsibility that should have been assigned, at the start.
Budget risks specific to Türkiye: e-document, bank and logistics integrations
For a company operating in Türkiye, the ERP carries not only the processes inside the company but also the mandatory, day-to-day data exchange with outside systems. These connections are among the items most often missing from a written scope, and when they are discovered late they add directly to both the budget and the timeline.
- The e-document applications of the Revenue Administration (GİB). e-invoice, e-archive invoice, e-dispatch note and e-ledger: which of them are mandatory for the company, which integration method will send them, and how document numbering, cancellation and return scenarios will work must be settled at the outset.
- Bank integrations. Automatic retrieval of account transactions, payment orders and reconciliation; every bank the company works with has its own method and file format, and that means a separate work item.
- Logistics and carrier integrations. Creating shipments and retrieving tracking numbers and delivery information; the e-dispatch note and the shipping process need to be designed together.
- Localisation. The Turkish chart of accounts, tax and withholding rules, exchange-rate differences and statutory reports; how far the standard package covers them, and where additional development is needed, must be clear before the contract.
The message here is simple: these integrations are not add-ons to be left for “later”. They belong in the inventory from the first day of the scope, each with an owner, a test scenario and a go-live step.
What does the X-ray show?
When you analyse the cost of an ERP project, the problem points always turn up in the same places:
- Scope that was thought to be clear at the start but was never approved in writing
- Data preparation for which nobody took responsibility
- Integrations that came to mind later
- A number of departments and variants that grew because of postponement
- Additional requests that piled up because there was no limit to stop them
None of these is a technical shortcoming. All of them are the result of a decision being taken later than it should have been.
Before starting the project, ask yourself this:
If you are about to start an ERP project, here are a few questions for reviewing your budget from the outset:
- Is our scope written and approved, or is it “in everyone’s head”?
- Which integrations are certain, and which “will probably be needed”?
- Do we know the name of the person responsible for data cleansing?
- Have we mapped the dependencies of the departments not included in the first phase on the others?
- Where have we drawn the boundary of the scope — at what point will we call an additional request “the next phase”?
If you cannot give a clear answer to these questions today, that is not a problem — it is normal. The real problem is looking for answers in the middle of the project, when the costs have already piled up.
8 ways to prevent budget and schedule overruns
None of the measures below requires a new tool; each one moves a decision or a responsibility to the start of the project.
- A validated scope before the contract. Processes, departments, variants and integrations should be written down and approved by the business. The price should be the price of that scope.
- A written change process. Every additional request should take the same route: definition, impact analysis, approval. Work added on a verbal approval is work that does not appear in the budget.
- One empowered sponsor. Decisions are owned by a single executive who can arbitrate between departments when needed. If decision authority is scattered, decisions are delayed too.
- Data ownership assigned at the start. For every master data area, the person responsible for cleansing and for approving the migration should be named in the first weeks of the project.
- An integration inventory drawn up front. Every connection to be built with e-document, bank, logistics, production and other systems should sit in a single list, with its owner and its test scenario.
- A phased go-live. Instead of switching everything on the same day, phases whose dependencies have been mapped split the risk and carry the lessons learned into the next phase.
- Protected key-user time. The time key users will give to the project should be set aside in writing by their managers, and their daily workload redistributed accordingly.
- A contingency reserve that is decided and documented. The allowance for uncertainty should not be left implicit “in case something comes up”; it should be decided explicitly on the basis of the risk register, with a written rule on who may use it and under what conditions.
Scope creep and change management
Scope creep grows through small individual decisions, so what stops it is not one big “no” but a simple process that every request goes through. A change request should contain at least: what is requested and why it is needed, the processes and systems affected, its impact on effort and timeline, and whether it will be done in this phase or the next.
The person who approves the request is the sponsor responsible for the project’s budget and timeline; larger requests are decided by the project steering committee. Every approved change is recorded in the scope document and the budget, and rejected or postponed requests also stay on record.
Risk register: what to track, and how?
The risk register is a living list in which the points seen on the X-ray are tracked throughout the project. Each risk is given an owner, and the register is reviewed in regular project meetings. The table below is an example to start from:
| Risk | Early sign | Measure |
|---|---|---|
| Scope creep | Verbal extra requests that never reach the log | Route every request through the change process |
| Late decisions | Process designs awaiting approval | Put decision dates in the plan, escalate to the sponsor |
| Data quality | Faulty records in a trial migration | Assign data owners, plan the cleansing |
| Integration delay | The other side’s test environment is not ready | Draw up the inventory early, start testing early |
| Key users unavailable | Postponed workshops and tests | Set project time aside in writing |
Success criteria: what to define at the start of the project
Whether a project has succeeded can only be measured if what counts as “success” was written down at the start:
- Which processes will run end to end in the new system at go-live
- The intended direction of improvement in business indicators such as month-end close, order fulfilment or stock accuracy
- Signs that users really use the system: parallel spreadsheets disappearing, fewer records keyed in by hand
- Budget and timeline assessed against a baseline that is tracked together with the approved changes
- The level to which support requests are expected to fall after go-live
How do you recover a project that has overrun its budget?
In a project whose budget or timeline has already been exceeded, the first reflex is to speed up; what is needed first, however, is to stop and make a diagnosis.
- Pause and re-baseline the scope. Set out in writing what has actually been done, what remains and what was added to the scope along the way; the new budget and timeline are built on that baseline.
- Separate the must-haves from the nice-to-haves. Keep in the first phase only what go-live really needs, and move the rest to later phases.
- Re-plan the phases. Map the dependencies again and divide go-live into smaller, manageable steps.
- Fix decision ownership. Write down again which decision lies with whom, and put the pending decisions, with their dates, on the sponsor’s desk.
How do you plan the ERP budget of a small business?
In a small business the principles do not change, only the scale does: still write the scope down, and treat staying close to standard processes as a budget decision, because every customisation creates cost both at implementation and in later version upgrades. With few key users, protecting their time matters even more here; and set the contingency reserve explicitly on the basis of the risk register, just as in larger projects. To list the cost items one by one, see the ERP project cost guide.
6 common misconceptions about ERP projects, and the reality
| Misconception | Reality |
|---|---|
| The licence fee is the cost of the project. | The licence is only one item of the total cost; implementation, data migration, integration, training and the internal team’s time also go into the budget. |
| Customisation is always inexpensive. | Every customisation is paid for again and again in testing, maintenance and version upgrades; departing from the standard process is a budget decision. |
| Data migration is a technical job. | The business decides which data is migrated and whether it is correct; the technical team only performs the transfer. |
| Training happens after go-live. | Training happens alongside testing, before go-live; an untrained user goes back to the old ways from the first day. |
| The scope becomes clear during the project. | Scope that becomes clear during the project is scope that is not in the contract; clarity is built before the contract. |
| More features mean better results. | What determines the result is not the number of features but whether the processes in use run end to end in the system. |
Frequently asked questions
How much contingency budget should an ERP project set aside?
There is no ratio that fits everyone; the size of the reserve is derived from the risks in the risk register and the uncertainty that still remains in the validated scope. What matters is that the allowance is decided explicitly, documented, and that it is written down under what conditions and with whose approval it may be used. In our assessment, the better the scope has been validated, the smaller the allowance needed.
Does the contract model affect budget overruns?
It does, but on its own it does not prevent them. Under a time-and-materials (man-day) model, the scope risk stays with the buyer: the more uncertainty, the more days are spent. A price fixed before the scope has been validated does not remove that risk either; it only moves it into change requests. What limits overruns is less the contract model than a validated scope and a written change process. We discuss the difference between the man-day model and the value-based model in the article Is Consulting That Sells Man-Days Coming to an End?.
Where do schedule overruns happen most often?
In our assessment, most often in data migration and cleansing, in tests squeezed into the end, and in process designs held up by late decisions. All three become manageable once an owner and a date are assigned to them at the start.
Who stops scope creep?
One empowered sponsor, together with a written change process. The sponsor approves no additional request without seeing its impact; larger requests go to the steering committee, and postponed requests are written onto the list for the next phase.
Related articles: Why do manufacturing companies in Türkiye digitalise the “wrong” way? and Projects without decision-making authority: the silent killer of digital transformation.
For ERP modules, selection steps and cost items, see our ERP guide.
If you would like to clarify the scope and risk points of your ERP project from the very beginning, you can get in touch with the MindDX team.
