Publicité
ERP IMPLEMENTATION
🇫🇷 Lire en français →

What Does One Month of ERP Project Delay Cost? A Calculation Model for the Steering Committee

Put a price on one month of ERP delay across 6 cost lines, compare postponement, scope cuts and extra resources, and present it to your steering committee.

What Does One Month of ERP Project Delay Cost? A Calculation Model for the Steering Committee

When an ERP project slips, the steering committee almost always asks the same question: “How many more weeks?” The question that should be asked is a different one: “What does each additional month cost us, and what could we buy with that money?” As long as delay is expressed in weeks, it stays abstract. Expressed in euros per month, it becomes a trade-off parameter on a par with scope and budget.

This article offers a calculation model you can fill in within an hour with your project manager and your financial controller, then present to the steering committee (SteerCo). It does not claim to give “the” cost of a delay: that cost depends on your contract, your organisation and your business case. It gives you a method for leaving nothing out and for comparing your options honestly.

Why Delay Is Almost Always Under-Costed in Steering Committees

In most committees, the cost of delay boils down to a single line: “the integrator is asking for a change order of X days.” It is the only item that is easy to read, because it arrives as a quote. Everything else stays diffuse.

Three mechanisms explain this bias.

Internal cost is invisible. The business and IT staff assigned to the project are already on the payroll. Their time generates no additional invoice, so nobody counts it. Yet every extra month keeps them on the project instead of their normal work, or forces them to do both at once.

Deferred benefits are not in the project budget. The business case promised gains (lower inventory, faster close, fewer re-keying tasks). A one-month delay pushes those gains back by a month. That loss appears in no project dashboard, because it is money you have not yet earned.

Coexistence costs add up silently. As long as the old system lives on, its licences, hosting, maintenance and double entry keep running. These lines are spread across different budgets (infrastructure, finance, business units) and nobody consolidates them.

The result: the committee compares a “small delay” with a “big cost of cutting scope” when it is really comparing two sets of costs, only one of which has been priced. That imbalance systematically pushes toward accepting the postponement, then a second, then a third. For the ritual that frames these decisions, see our weekly steering committee template (in French).

The 6 Cost Lines of One Month of Delay

Here are the six lines to review. For each one, start from a monthly cost and ask: “Does this cost continue if go-live slips by a month?“

1. The Internal Project Team

This is the fully loaded cost of the people assigned to the project: project manager, key users, architects, IT, finance, procurement. The calculation: number of full-time equivalents (FTEs) assigned multiplied by the average monthly loaded cost.

Two clarifications. First, count only the time actually spent on the project, not the headcount nominally “involved.” Second, add the cost of backfill or overload: if key users are covered by temps or overtime, those costs are real.

2. The Integrator: Time and Materials or Fixed Price

Treatment depends on the contract. Under time and materials (T&M), every month of extension is billed: number of consultants multiplied by working days multiplied by daily rate. Under fixed price, the direct cost is not automatically yours, but two risks remain: the change order (if the delay is attributable to changes on your side) and quality degradation (an integrator losing money reassigns its best people). To go deeper on the difference, read our article on fixed price, time and materials and outcome commitments (in French).

3. Dual Running and Legacy Licences

While the old system keeps running, you keep paying: licences and maintenance, hosting, support and, above all, double entry or manual reconciliations if part of the flows has already moved over. Add renegotiation costs: some vendors charge more for a last-minute extension than for a planned annual contract. Check your notice and contract end dates: a go-live that slips past the end of maintenance for the old ERP changes the nature of the delay entirely.

4. Deferred Benefits

This is the most politically sensitive line, because it rests on the business case. Method: take the recurring annual gains the project is meant to deliver (re-keying savings, lower inventory, shorter close cycles, avoided legacy costs), divide by twelve, and apply only what is actually unlocked by go-live.

Be careful: a benefit that was not seriously quantified before the project must not be invented at the moment you need to justify a trade-off. If the business case is fragile, say so and show this line separately, with its confidence level. The CFO will accept a model more readily if it distinguishes cash cost (lines 1 to 3) from opportunity cost (line 4).

5. Fixed-Deadline or Regulatory Risk

Some projects have a non-negotiable date: end of vendor support, an e-invoicing mandate, a fiscal year-end, a merger, a site opening. If go-live overshoots that date, you pay either for a workaround (manual entry, a temporary tool, a compliance provider) or for the risk of a penalty or business disruption.

Do not price this line with a theoretical fine you cannot source. Price the cost of plan B: what does the interim solution cost, and for how many months? That is a cost you can have validated by a quote.

6. Team Fatigue and Turnover

This is the line nobody dares to price, yet it weighs heavily: a project that drags on exhausts key users, who are also your best business experts. The departure of just one of them can delay an entire work package.

A simple method, to be shown as a provision: estimate the probability that a key person leaves the project or the company if the delay exceeds a threshold (for example one quarter), and multiply by the cost of replacement and ramp-up. This is an estimate; present it as one, with the assumption written down.

Step-by-Step Calculation Model

Step 1: Set the Scope of the Calculation

Define the reference period (one month of delay) and the current go-live date. State whether the delay is assumed to be linear: integrator and dual-running costs often are, regulatory risk is not (it flips all at once at a threshold date).

Step 2: Fill in the Table

LineFormulaValue to enterData source
1. Internal teamFTEs assigned × monthly loaded cost… €HR / management control
2. IntegratorConsultants × days × daily rate (T&M) or likely change order (fixed price)… €Contract, schedule
3. Dual runningLicences + hosting + maintenance + double-entry FTEs… €Procurement, IT, finance
4. Deferred benefitsAnnual business-case gains ÷ 12 (share unlocked at go-live)… €Approved business case
5. Fixed deadlineMonthly cost of plan B or workaround… €Quotes, IT
6. Fatigue and turnoverProbability × replacement cost… €HR, estimate
Monthly totalSum of lines 1 to 6… €

Add two useful columns: “cash / opportunity” and “confidence level” (high, medium, low). The committee must see what is certain and what is estimated.

Step 3: An Illustrative Example

Note that the example below is entirely fictitious. The amounts are assumptions chosen to illustrate the method, not market benchmarks. They describe a 400-person mid-sized company replacing its ERP.

LineFictitious assumptionsMonthly cost
1. Internal team12 FTEs × €7,000 loaded€84,000
2. Integrator (T&M)8 consultants × 20 days × €850€136,000
3. Dual runningLegacy licences €18,000 + hosting and maintenance €7,000 + 3 double-entry FTEs × €5,500€41,500
4. Deferred benefits€1.2M expected annual gains ÷ 12€100,000
5. Fixed deadlineInterim tool and manual entry for a compliance obligation€25,000
6. Fatigue and turnover10% probability of losing a key person × €150,000€15,000
Monthly total€401,500

Reading: in this fictitious scenario, one month of delay costs about €400,000, of which €261,500 is cash (lines 1 to 3) and €140,000 is opportunity cost and risk (lines 4 to 6). A two-month delay, all else equal, approaches €800,000.

The message to the committee is not “the delay is serious.” It is: “each month of slippage equals X euros, and any option that costs less than X per month saved deserves examination.” That sentence changes the nature of the discussion.

Step 4: Test the Sensitivity

Rerun the calculation with a low and a high assumption on the two most uncertain lines (often 4 and 6). If the result stays within a range where the decision does not change, the uncertainty is not an obstacle. If the decision flips, you know which assumption to firm up before the SteerCo.

Delay vs Scope Reduction vs Extra Resources: The Decision Tree

Once the monthly cost is known, the committee compares three levers, knowing that the time, scope and budget triangle does not let you hold all three.

The Three Options

Option A: accept the postponement. You pay the monthly cost of delay for each month bought for quality. Relevant when the cause of the delay is insufficient preparation (data, testing) that threatens cutover.

Option B: reduce scope at go-live. You push some modules or processes to a phase 2. You save project time, but you pay for interim interfaces, a longer partial dual run, and deferred benefits on the postponed processes.

Option C: reinforce the team. You add consultants or key users to hold the date. The cost of reinforcement is direct, and its effectiveness is limited: adding people to a late team often slows the project down in the short term (onboarding, coordination), especially on work packages where the bottleneck is business decisions rather than production capacity.

The Decision Tree

Ask these questions in this order:

  1. Is go-live blocked by a quality criterion (testing, data, cutover)? If so, no extra staffing replaces the time needed for fixes: favour A, or B if the defect is localised to a separable scope.
  2. Is the bottleneck capacity (volume of development, testing, migration)? If so, C is relevant, provided the reinforcements are operational within a few weeks and the tasks can be parallelised.
  3. Is the bottleneck decisions (business trade-offs not made, design not signed off)? In that case neither A nor C fixes the problem: you need a sponsor decision, which means a governance trade-off.
  4. Is there a block of scope with low value at go-live? If so, B is probably the cheapest option: postpone what pays back least.
  5. Is there a non-negotiable date (line 5)? If so, compare the cost of the compliance plan B with the cost of B or C to hold the date.

Comparing Options in Euros

For each option, estimate three quantities: the direct cost of the option, the number of months of delay avoided or incurred, and the knock-on costs. Back to the fictitious example:

OptionFictitious assumptionsEffect on cost
A: 2-month postponement2 × €401,500about €803,000
B: 3 modules moved to phase 2, go-live held€30,000 interim development + €30,000/month deferred benefits over 6 months + €10,000/month partial dual run over 6 monthsabout €270,000
C: 4 extra consultants for 2 months to gain 1 month4 × 20 days × €850 × 2 months, plus 1 remaining month of delayabout €136,000 + €401,500 = €537,500

In this fictitious example, scope reduction is the cheapest option because the postponed scope carries little value. In another situation, where the postponed scope holds most of the benefits, the ranking would flip. That is the point of the model: the right answer depends on your project’s numbers, not on a general rule.

When Postponing Go-Live Is the Right Decision

The cost model must never become an argument for forcing a premature go-live. A delay is expensive, but a failed cutover generally costs more: invoicing stoppages, delivery disruptions, a compromised accounting close, lost user confidence. Postponement is the right call when quality criteria have not been met.

Go/No-Go Criteria to Formalise

Define them before the critical period, not the night before go-live:

  • Testing: end-to-end scenarios covering critical processes (order-to-cash, procure-to-pay, period close) have passed, and no blocking or major defect is open.
  • Data: the migration has been rehearsed at least once as a dry run, and reconciliation gaps (inventory, ledger balances, outstanding items) are explained and accepted by the data owners.
  • Users: key users and critical users are trained and have signed off their scenarios in user acceptance testing.
  • Cutover: the cutover plan has been rehearsed, with measured timing, a rollback plan and an identified support team. The detail is in our operational cutover checklist from D-30 to D-3 (in French).
  • Operations: hypercare is staffed and escalation paths are known.

How to Use the Cost Model in This Decision

The calculation exists to make the postponement explicit and bounded, not to justify it after the fact. A good go/no-go committee looks at two columns: the cost of postponing by one month (your monthly total) and the estimated cost of a cutover incident (business stoppage, manual catch-up, production fixes). If you cannot price the second, ask each process owner: “If this flow is unavailable for three days, what happens?” The answers, even qualitative ones, put the postponement in perspective.

One point of caution: a postponement must always come with an exit condition. “We postpone by a month” with no clear criterion amounts to postponing indefinitely. Write it down: “we cut over if criteria X, Y and Z are met on date D; otherwise we reconvene the SteerCo.”

If warning signals are piling up and you are hesitating, reread our article on the warning signs of an ERP project in trouble.

Contracting: Penalties, Milestones, Postponement Clauses

The cost of delay must also be dealt with in the contract with the integrator and the vendor, ideally before signing.

Payment milestones tied to deliverables. Link payments to verifiable deliverables (approved specifications, successful acceptance testing, validated dry-run migration) rather than to dates. A milestone paid on a date, without an accepted deliverable, protects nothing.

Late-delivery penalties. They are common but often badly designed: a cap too low to deter, triggers subject to exclusions so broad that they rarely apply. Negotiate the definition of attributable delay, the cap and the calculation method. Keep in mind that a penalty almost never compensates the real cost: it mainly serves to align priorities.

Postponement clauses. Provide the mechanism: who can request a postponement, with what notice, what the consequences are for invoicing, and who bears the costs if the customer requests the postponement because its own teams were not prepared.

Change governance clause. Every change request must have its schedule and cost impact visible before approval. This prevents scope from swelling without the delay being attributed.

Reversibility and assistance clauses. They protect your ability to carry on if the relationship deteriorates.

These points are detailed in our guide to the clauses to check before signing an ERP contract. For any contract drafting, have the text reviewed by your legal counsel: the clauses above are negotiation guideposts, not a contract template.

A “Cost of Delay” Slide Template for the SteerCo

A single slide is enough if it is well structured. Here is the template.

Title: one sentence that contains the decision. Example: “Each month of delay costs about €400k: we are asking for a decision on three options.”

Block 1 (left): the monthly cost of delay. The six-line table, with two subtotals (cash, opportunity and risk) and the confidence level of each line.

Block 2 (centre): the three options. Postponement, scope reduction, extra resources: direct cost, effect on the go-live date, main risk.

Block 3 (right): the recommendation. One option, the exit condition (go/no-go criteria and review date) and the next decision point.

Slide footer: the assumptions. Three or four key assumptions spelled out: the daily rate used, the share of benefits unlocked, the turnover probability. A committee that sees the assumptions will discuss the assumptions rather than dispute the total.

Practical tips: send the table to management control before the meeting so they can validate the figures, always show a range when an assumption is fragile, and record the decision in the minutes with its estimated cost. This follow-up lets you compare planned and actual cost at the next committee.

Going Further

The cost of delay is only one side of the equation: the other is the initial budget. To understand how it is built, read our analysis of ERP implementation cost and what to budget. To structure your vendor and integrator evaluations upfront, download our ERP evaluation scorecard to compare offers on quantified criteria.