A wire products manufacturer in Rajkot, two plants, about 220 people, ₹60 crore. The founder's son had been running operations for two years and had spent four months getting to a shortlist. Then he presented it, and it died in eleven minutes. His father wasn't hostile. He asked one question the proposal had no answer to, and that was that.
The proposal was good. It had the right product, a fair price and a sensible timeline. What it didn't have was a number anybody in that room already believed.
How do you justify an ERP investment to the board?
Four numbers, all of them already yours
The mistake in most ERP proposals is that the numbers come from the vendor. Every figure a vendor supplies gets discounted in the room, and rightly, because it's marketing until proven otherwise. Numbers that come out of your own ledger don't get discounted. They get argued about, which is different and much more useful.
Working capital sitting still. Pull stock valuation and find what hasn't moved in twelve months. In most plants of this size the figure is larger than anyone expects and nobody has said it out loud. In Rajkot it was ₹31 lakh across two stores, and about a third of it was bought-out items for a machine configuration that had changed.
Overtime and expediting. Add last year's overtime to the premium freight you paid to rescue late dispatches. That whole number is a proxy for planning you couldn't do, and it's an operating cost your finance team already tracks.
What late and short deliveries cost. Debit notes, penalty clauses and the discount you gave to keep a customer who'd been let down twice. Every plant has these and almost nobody totals them.
Reconciliation hours. Count the people who spend a week each month making reports agree, then multiply by twelve. This one is uncomfortable because the answer is usually a person and a half.
What to leave out
Three things weaken a proposal and they're the three most people put in first.
Industry statistics about digital transformation. Everybody in the room knows they're generic, and using one signals that you couldn't find a number in your own business. Efficiency percentages from a vendor deck have the same problem. And a five-year projection invites an argument about year four that has nothing to do with the decision in front of you.
Cut all three. A one-page case built on four of your own numbers beats twelve pages of anything else, because the discussion it starts is about your plant.
Where the payback actually comes from
People expect the return to come from labour savings. It rarely does, and leading with that invites a conversation about headcount that nobody wants to have. Most of the return is working capital released and rework avoided.
The objection that ends most proposals
"We tried this before and it didn't work."
In a family business this is almost never really about software. It's about a previous attempt where money was spent, the floor never used the thing, and somebody was quietly blamed. Argue with the objection and you lose. Concede it and you can move.
The version that worked in Rajkot was short. He agreed the first attempt had failed, said plainly why, which was that it went live across five modules at once and the supervisors were never asked, and then proposed one module, one plant, ninety days, with a named supervisor who could veto it. His father approved that in the same meeting. The scope was smaller than the original ask and it got signed, which is the trade worth making.
"He stopped trying to convince me the software was good. He told me what went wrong last time and asked for a tenth of the money to prove it wouldn't happen again. That I could say yes to."Founder, wire products manufacturer, Rajkot
The specific ways these rollouts fail are worth knowing before you write the proposal, because naming them yourself is what makes the concession credible. We've set out the common ones in the five ERP mistakes manufacturers make. For the cost side of the page, the three-year total cost of ownership model gives you a figure that holds up under questioning better than a licence price does.
Facto goes live one module at a time on purpose, deployed by our own engineers, which makes the ninety-day proposal above an ordinary thing to ask for rather than a concession. If you're writing this page for your own board, talk to our team and we'll help you put the four numbers together from your data before you present it.
