Research

Every finding is filed under one system, and the split is the scoping answer. Six of them sit inside the three SAP modules already running and are worth ₹3.5 Cr a year: that half is configuration and process inside software the client has paid for, so it is cheaper to fix and harder to sell, because nothing new arrives. The other six sit on work no software touches at all and are worth ₹6.2 Cr, close to twice as much. That is where a build goes, and it is why the first phase is worth scoping outside the ERP rather than inside it. Both figures are counted from the findings themselves, so they move when a finding moves.

Three systems the client has never bought: Warehouse management, Invoice capture and Spend analytics. The work still happens, so a person is doing it. Warehouse management falls to paper, at all three plants. The receipt is posted whenever somebody gets to it, which sets the clock on the match failure. Invoice capture falls to nine people in AP, keying about 96,000 invoices a year from PDF and paper. Spend analytics falls to nobody at all, which is why no price on this list is measured from their own data. Three findings sit here, worth ₹2.7 Cr a year, and this is where a first phase has the most room: there is no incumbent product to displace and nothing to migrate off. It is also the weakest evidence on the page, because the one system that would measure any of it is the one nobody bought.

Three processes that exist and run, but not anywhere that leaves a record. Purchase approvals falls to email, and WhatsApp when it is urgent. The record is written after the decision, or not at all. Demand planning falls to one spreadsheet, maintained by the planning team. It is why stock cover runs at 38 days. Freight tendering falls to phone and email, one carrier per lane. Nothing compares a rate against the last one. Three findings sit here, worth ₹3.5 Cr a year. These are the hardest to raise on a call, because nobody experiences them as broken: the purchase gets approved, the lorry turns up, the plan gets made. What is missing is the trail, which is why none of it can be measured and why the fix reads as a control rather than as a saving.

SAP MM, SAP FI and SAP SD, all on SAP ECC 6.0 and all supplied by SAP. SAP MM handles purchase orders, goods receipts and the supplier master. SAP FI handles invoices, matching and the payment run. SAP SD handles orders out, distributors and rebate schemes. Six findings sit on them, worth ₹3.5 Cr a year. None of these is a problem in itself. They are the ground anything new would stand on, and the reason to read the section is that an integration touching a live module is a conversation with SAP and whoever manages the estate, not just with the client.

SAP ECC 6.0, with three modules live and six processes running outside all of them. Anything built here reads from and writes back to SAP MM, SAP FI and SAP SD, which is the answer to the first question an engineer asks and the first one a client asks after it. The work splits cleanly: ₹6.2 Cr of it needs something new that stands beside SAP and feeds it, and ₹3.5 Cr of it is configuration and process inside modules that are already paid for. Nothing here needs the ERP replaced, which is the sentence worth saying early, because it is the fear the room brings to the meeting.