All Articles

Your ERP Will Go Live. The Payback Is Decided Before It Starts.

Rapid Acceleration PartnersNovember 17, 2026
An engineering team reviews CAD models and simulation results on a bank of monitors in an office overlooking a factory floor

The ERP Reality Check · Part 1 of 4

When an ERP program doesn't pay back, the cause is usually the approach, not the software: old processes copied into a new system.

Finance feels it first, and the companies that redesign approvals, reconciliations and controls before go-live pull further ahead every quarter.

Your ERP program will almost certainly go live. Whether it pays back is a different question, and most of the answer gets written before anyone signs the statement of work. It sits in the ERP business case: what you promised, how you'll measure it, and who in the business owns each number.

The research points the same way. McKinsey's analysis found nearly 70% of ERP programs fall short of their full potential, and it traces the gap mostly to how companies run the program rather than to the software. Gartner predicts that by 2027, more than 70% of recently implemented ERP initiatives won't fully meet their original business-case goals.

For a CFO, that second number is the uncomfortable one. The business case is the document your team wrote to get the money, and it's the yardstick the board will hold up in year three.

The system usually works. The value leaks out somewhere between the approval memo and the second year of live operations. We see it leak in four places, and each one has a fix a finance leader can put in place.

Leak 1: an ERP business case nobody can measure

Plenty of business cases promise "better visibility", "a single source of truth" and "efficiency gains". None of those can be checked. There's no baseline, so nobody can say later whether anything moved. And the benefits often belong to the program team or the integrator, both of whom leave once the system is stable.

A business case like that gets the budget approved. It just can't tell you whether the money came back.

There's a second gap: costs that never make it into the case at all. The manual work around the ERP, like building quotes by hand from spreadsheets and price lists, rarely gets a baseline. We put numbers on one example in the true cost of manual quote generation.

Decide what success looks like before kickoff

The move that works is a scorecard, set before you sign the SOW. Pick five or six KPIs that the program is supposed to change, baseline them today, and give each one an owner who runs that part of the business.

Three details make this hold up.

Baseline now. Once the program starts, teams stop tracking the old measures carefully, and a year later nobody can prove the delta.

Put an owner in the business, not on the program. Program teams disband. A head of FP&A who signed up for forecast accuracy is still there in month eighteen, and still answers for it.

Keep the list short. Twenty KPIs means nobody watches any of them. Six that tie straight to operating profit and cash get attention at every meeting.

Here is what that can look like. The figures are illustrative; yours come from your own ledger.

KPI Baseline today Target, 12 months after go-live Owner
Days to close 9 working days 5 working days Group controller
Cost per supplier invoice processed $11.40 $6.00 Head of shared services
Manual journals per month 1,400 500 Head of financial reporting
Forecast accuracy, next quarter revenue ±8% ±3% Head of FP&A
DSO 58 days 49 days Head of credit and collections
Working capital tied up in inventory 74 days 65 days Supply chain director

The programs that pay back are the ones measured from day one. In one RAP deployment, a $1.5B credit union hit 8x ROI in year one, with 70% of its targeted use cases live in 60 days and no headcount added.

This is the core of ERP value realization. Everything after kickoff either protects these numbers or quietly trades them away.

Leak 2: the old process, moved into a new system

The second leak is design. If the future state is today's process with new screens, the scorecard won't move. You've paid for a new system to run the same approvals, the same reconciliations and the same workarounds.

You don't have to redesign everything. Redesign the few processes that carry the value on your scorecard, usually the close, procure-to-pay and collections, and keep the rest close to standard. For a manufacturer or distributor, that often means three-way match on supplier invoices, customer deductions and intercompany across plants.

Leak 3: nobody tracks value after go-live

Go-live is where most programs stop counting. Hypercare ends, the integrator rolls off, the program team moves to the next project. Yet most benefits show up, or don't, somewhere between month three and month eighteen.

So keep measuring. ERP benefits tracking should run for at least 12 months after go-live, with the scorecard reported to the CFO every quarter by the same owners who signed up for it. If days to close stalls at seven, you want to know in month four, while there's still budget and attention to fix it.

The same discipline belongs earlier, too. Review benefits at every stage gate, next to budget and dates. When someone proposes cutting a workflow to protect the timeline, price the cut in scorecard terms. "We save six weeks and give up two days of close" is a decision a CFO can make. "We descoped some automation" isn't.

Leak 4: governance that tracks dates instead of value

Look at a typical steering committee pack. It opens with milestone status, defect counts and budget burn. Those matter. But if value shows up on slide fourteen, it gets discussed for four minutes at the end.

McKinsey's numbers on large tech programs, not only ERP, show where that leads: 65% to 80% run over budget or timeline, and just 25% to 35% hit their operating profit and cash-flow targets. Running late is common. Losing the payback is the expensive part.

Two changes fix most of this. Lead the steering committee agenda with the scorecard, before dates and defects. And have the program lead report to the CFO. Reporting lines decide which questions get asked.

A short check before you sign

  • ☐ Five or six KPIs, each baselined with today's numbers
  • ☐ A named business owner for every KPI, outside the program team
  • ☐ Target values agreed with those owners, not just the program sponsor
  • ☐ The processes you'll redesign, tied to specific KPIs
  • ☐ Benefits reviewed at each stage gate, next to budget and dates
  • ☐ A measurement plan that runs 12+ months past go-live
  • ☐ Program lead reporting to the CFO, steering pack led by the scorecard

If you can tick all seven, your ERP ROI is something you can actually report on. If you can't, the business case is still a wish list, and that's fixable before kickoff more cheaply than at any point after.

Bring your ERP business case. In 30 minutes, we'll show which of those numbers an AI-run process can move, and how you'd measure it. Book a 30-minute session.

FAQ

What should an ERP business case include?

Beyond cost and timeline, it should carry five or six measurable KPIs with today's baseline, a target for each, and a named business owner. Typical choices are days to close, cost per invoice, manual journals per month, forecast accuracy and DSO. Add which processes you'll redesign to move those numbers and how long you'll keep measuring.

How do you measure ERP ROI?

Compare the scorecard KPIs against the baseline you recorded before kickoff, then convert the changes into operating profit and cash. Fewer days to close and lower cost per invoice show up in finance cost. Lower DSO and inventory days release working capital. Without a pre-kickoff baseline, any ROI figure is an estimate.

When should ERP benefits be tracked?

Start before kickoff, with the baseline. Review benefits at every stage gate next to budget and dates, so scope cuts get priced in value terms. Then keep tracking for at least 12 months after go-live, reported quarterly, because most benefits appear between month three and month eighteen of live operations.

Who should own ERP benefits?

The business leaders who run the affected processes, not the program team or the integrator. The group controller owns days to close, the head of FP&A owns forecast accuracy, collections owns DSO. Those people stay after go-live. The program lead should report to the CFO, who holds the overall scorecard.

How many KPIs should an ERP scorecard have?

Five or six is enough for most programs. Each should tie directly to operating profit, cash or finance cost, and each needs a baseline and an owner. Longer lists spread attention thin, and steering committees stop discussing them. Pick the measures the program is most likely to move, and the board most likely to ask about.

Take the next step

Ready to Move From Exploration to Execution?

Leading manufacturers are closing the AI execution gap right now. Book a demo and see your operation transformed in 90 days.