The ERP business case for manufacturers: building an ROI your board will approve
Most ERP proposals that reach an Indian manufacturer’s board fail for the same reason: they list features and a vendor quote, then ask for approval. Directors approve cash out against cash back, with the risks named. A credible ERP business case starts from what your current way of working costs today, claims only the benefits you can defend line by line, and puts the full cost of ownership over three to five years next to them. This article gives you the method, a worked example for an ₹80 crore manufacturer, and the way we suggest presenting it.

Why most ERP business cases get sent back
We have sat in enough review meetings to know the pattern. The CFO asks where the “20% inventory reduction” came from. Nobody can say. The MD asks what happens if go-live slips. Nobody has modelled it. The ERP business case goes back for rework, and the project loses a quarter.
Three mistakes cause most of this. Benefits are borrowed from vendor brochures instead of measured in your own plant. Costs stop at the implementation quote and ignore hosting, support and your own people’s time. And the same rupee gets counted twice, typically as both freed-up working capital and the interest saved on it. Fix those three and the board conversation changes.
Step 1: baseline what the current system really costs
Every ERP business case stands on its baseline. Before you estimate a single benefit, measure the present: two to three weeks of honest digging by finance, with help from stores, planning and sales. Collect these numbers for the last twelve months:
- Manual effort: hours per week spent re-keying data between Tally, Excel and WhatsApp, reconciling stock, chasing job work challans and preparing MIS. Multiply hours by a loaded hourly cost (salary plus PF, ESI and overheads).
- Inventory: average stock value, slow-moving and dead stock, and the value written off or sold as scrap because it expired, was damaged or was simply forgotten.
- Working capital: debtor days and inventory days, and the rate you pay on your cash credit or working capital limit.
- Month-end close: how many working days after month end before management sees a reliable P&L.
- Billing errors: credit notes raised because of wrong rates, missed freight or duplicate invoices, and the margin lost on them.
- GST mismatch effort: hours spent matching purchase registers with GSTR-2B, and input tax credit deferred or lost because supplier invoices did not match.
Picture a Coimbatore foundry where pattern-shop and fettling data still lives in paper registers, so production cost per casting is a guess. That guess has a cost, but leave it unpriced unless you can measure it.
Step 2: quantify benefits conservatively
Every benefit line in the ERP business case needs three things: the baseline number, the improvement claimed, and the person who owns delivering it. If the stores manager will not sign up to a 5% stock reduction, do not claim it.
Be deliberately cautious. Claim a fraction of manual hours saved, not all of them, because people get redeployed rather than leaving payroll. Value working capital released at your cost of funds, once. Assume the first year delivers only part of the benefit, because users take months to trust the new numbers. And keep softer gains (faster decisions, audit trail, cleaner e-invoicing and GST compliance, a system that can scale to a second plant) as a separate, unpriced list. Boards respect that more than an inflated total.
Step 3: total cost of ownership over 3 to 5 years
An ERP business case that shows only the implementation fee is incomplete. Total cost of ownership starts with implementation, which covers configuration, data migration, training and any custom development. For small, standard ERPNext projects, Frappe’s Starter Pack pricing gives a reference point: Core at ₹80,000 for 40 hours, Manufacturing at ₹1,40,000 for 70 hours, All in One at ₹2,00,000 for 100 hours, and Frappe HR at ₹60,000. Larger or more complex projects are scoped separately. Our ERPNext implementation cost guide and estimator will give you a realistic range for your size.
Then add hosting, which depends on users, data volume and whether you choose a managed cloud or your own server; see our ERPNext hosting cost breakdown for India. Annual support or AMC. Hardware such as barcode scanners and label printers. And the cost everyone forgets: your own team’s time. Key users spend weeks on process workshops, data cleansing and testing. Put a number on it, because the board will ask.
ERPNext is open source, so there is no per-user licence fee. That changes the cost curve; it does not make the project free.
A worked ERP business case for an ₹80 crore manufacturer
The numbers below are illustrative. They describe a hypothetical engineering manufacturer with ₹80 crore turnover, one plant and about 40 ERP users. They are not drawn from any client and are not benchmarks. Use the structure, then replace every number with your own baseline.
| Benefit line (illustrative) | Baseline assumption | Conservative claim | Annual value |
|---|---|---|---|
| Manual effort | 5,400 hours a year at ₹400 an hour (₹21.6 lakh) | 40% recovered | ₹8.6 lakh |
| Inventory carrying cost | Average stock ₹16 crore, cost of funds 10% | 5% lower stock, ₹80 lakh released | ₹8.0 lakh |
| Stock write-offs | ₹40 lakh a year | 20% lower | ₹8.0 lakh |
| Receivables | Debtor days cut by 3, about ₹66 lakh released | Valued at 10% cost of funds | ₹6.6 lakh |
| Month-end close | 12 days to 6 days | Senior finance time freed | ₹2.0 lakh |
| Billing errors | ₹10 lakh margin lost a year | Half recovered | ₹5.0 lakh |
| GST input tax credit | ₹6 lakh deferred or lost a year | Half recovered | ₹3.0 lakh |
| Total annual benefit at full run rate | ₹41.2 lakh |
Note what we did with working capital. The ₹1.46 crore released from stock and debtors is real cash, and it belongs in the board pack as a balance sheet effect. But in the cash flow model we count only the interest saved on it. Counting both is the classic double count.
On the cost side, again illustrative: one-time costs of ₹31 lakh (implementation, customisation and data migration ₹15 lakh, internal team time ₹8 lakh, training and backfill ₹2 lakh, scanners and printers ₹2 lakh, contingency ₹4 lakh). A project of this size would be scoped, not priced off the Starter Pack. Recurring costs of ₹6.5 lakh a year (hosting ₹1.5 lakh, AMC ₹3 lakh, part of an internal ERP owner’s time ₹2 lakh).
Assume year one delivers half the benefit. The illustrative net cash flows are then minus ₹31 lakh upfront, ₹14.1 lakh in year one and ₹34.7 lakh in each of years two to five. Cumulative net benefit over five years is about ₹1.22 crore.
Payback and NPV in plain terms
Payback is how long before cumulative savings cover what you spent. Here it is about 18 months. NPV (net present value) asks a stricter question: after discounting future savings, because a rupee in year four is worth less than a rupee today, is the project still ahead? Use a discount rate a little above your borrowing cost to reflect risk. At 12%, this illustrative case gives an NPV of roughly ₹76 lakh. Positive NPV means the project earns more than your hurdle rate.
Risks and sensitivities: test the ERP business case before the board does
A single-scenario ERP business case invites suspicion. Show the board what happens when things go wrong, using the same illustrative example:
| Scenario (illustrative) | Five-year NPV at 12% | Payback |
|---|---|---|
| Base case | About ₹76 lakh | About 1.5 years |
| One-time cost 50% over budget | About ₹60 lakh | About 1.9 years |
| No benefit at all in year one | About ₹57 lakh | About 2.1 years |
| Benefits only half of plan | About ₹11 lakh | About 2.9 years |
| Half the benefits and 50% cost overrun | About minus ₹5 lakh | About 4 years |
The last row matters most. It tells the board the project only fails if both the benefits and the budget go badly wrong at once. That is a far more honest message than “ROI of 300%”, and it shows where management attention belongs: benefits delivery, not vendor rate negotiation.
The real risks are familiar: poor master data, key users pulled back to production at go-live, customisation that copies old habits. Our ERP implementation readiness framework helps you score these before you commit.
How to present ERP ROI for manufacturers to your board
Keep the ERP business case to one page with an appendix. In our experience boards respond to this order:
- The decision you want: approve ₹X one-time and ₹Y a year, with a go-live month.
- The problem in the company’s own numbers: the baseline from Step 1, not industry averages.
- Benefits by line, each with an owner, plus the unpriced list kept separate.
- Total cost of ownership over five years, including internal time.
- Payback, NPV and the sensitivity table.
- Top five risks and who mitigates each one.
- How you will measure: the same baseline metrics, reported at three, six and twelve months after go-live.
Then use a simple decision rule. If the base case NPV is positive and the downside case is close to break-even, approve. If the base case only works with aggressive benefit claims, run a smaller first phase (finance, stock and GST) and prove the numbers. If you cannot produce a baseline at all, the first project is fixing your data, not buying software.
A Tiruppur knitwear exporter and a Pune auto-component maker will have different benefit lines; job work tracking might dominate one, scrap and rework the other. The method stays the same.
Once the board approves, timing matters. Our ERPNext implementation timeline shows what a realistic plan to a 1 April go-live looks like, and the multi-company ERPNext guide covers groups with more than one legal entity.
Want a second pair of eyes on your numbers? We will help you build the baseline and stress-test the case before it goes to the board. Book a free 30-minute ERPNext consultation, or start with our cost estimator.
Frequently asked questions
How do you calculate ROI on an ERP implementation?
Measure the current cost of manual work, inventory, working capital, billing errors and GST mismatches, claim a conservative improvement on each, then compare those annual savings with the total cost of ownership over three to five years using payback and NPV.
What is a reasonable payback period for ERP in manufacturing?
There is no universal figure. It depends on your baseline and project cost. Build your own model, and show the board both a base case and a downside case rather than quoting an industry number.
What costs should an ERP business case include?
Implementation, data migration, customisation, hosting, annual support or AMC, hardware such as scanners, training, contingency and the time your own staff spend on the project.
Should freed working capital be counted as a benefit?
Count it once. Show the cash released as a balance sheet effect, and in the cash flow model count only the interest saved on it at your cost of funds.
Does ERPNext have licence fees?
ERPNext is open source, so there is no per-user licence fee. You still pay for implementation, hosting and support.
What discount rate should I use for ERP NPV?
Use your company’s cost of capital or a rate slightly above your working capital borrowing rate to reflect project risk. Your CFO or auditor can confirm the right figure.
ERP business case: key takeaways
A board-ready ERP business case starts from your own measured baseline, claims conservative benefits with named owners, counts working capital once, includes internal time and recurring costs over three to five years, and shows payback, NPV and a downside scenario. Base GST benefits on real reconciliation data; the GST portal explains how GSTR-2B is generated from supplier filings, which is where most input tax credit leakage shows up.