An ERP for small and mid-size manufacturers in India: BOMs, production planning, stores and inventory, purchase, job-work in both directions, batch costing and dispatch — deployed on your floor, shaped to your process, GST-native. Status: deployed and demonstrable.
What does the Tally-Excel-register stack actually cost a factory?
Walk any mid-size plant — Pune’s auto belt, Ludhiana, Rajkot, Coimbatore — and the same stack appears: Tally for accounts, Excel for planning, carbon-copy registers in stores, and memory for job-work. It works, kasa kay, the way a bridge with a known crack works: daily, until the day it doesn’t.
The costs hide in the gaps between the tools. Material truth diverges from book truth — “stores mein hai, par kahan hai?” is a sentence that costs real money at audit and real hours every day. Job-work challans age unreconciled until the vendor’s version of events is the only version anyone can produce. Purchase rates reset with every order because rate history lives in a drawer. And the real product cost — material plus labour plus power plus rejection plus wastage — is discovered annually, approximately, in an exercise everyone dreads. A factory making precision parts runs its commercial side on approximations; the machines deserve better paperwork, barobar?
What does the Manufacturing ERP include?
- BOMs and routing, multi-level and versioned. Engineering changes tracked with effective dates, so “which drawing was this batch built to?” has an answer years later.
- Production planning that respects capacity. Sales orders to work orders to floor schedules, with honest capacity maths — the plan the floor can actually run, not the one the spreadsheet permits.
- Stores with bin truth. GRN, issues-to-production, rejections, returns — stock that matches the floor because every movement is an event, not a month-end adjustment. The ledger discipline pointed at material: same architecture, steel instead of rupees.
- Purchase with memory. Indents, POs, rate history per item per supplier, GRN-versus-invoice matching that catches the quiet over-billing generic processes miss.
- Job-work, both directions. Challans out and in, material accountability, wastage norms, ageing alarms — the trade’s most trust-dependent flow, given receipts.
- Batch costing while it’s fixable. Per-batch actuals against standards — material, labour, power, rejection — variance named the week it happens, not the year after. The first honest cost sheet is usually a revelation and occasionally an argument; both are progress.
- Dispatch done right. Packing lists, e-way bills, GST e-invoicing (mandatory above the ₹5-crore turnover threshold — cbic.gov.in), and the dispatch register your customers’ gates stop arguing with.
- The Tally bridge. Clean vouchers flow to statutory accounting; the CA’s world is undisturbed and better-fed.
How does a factory adopt this without stopping?
In the sequence that pays back fastest and risks least — proven in the ERP migration engagement. Stores and purchase first: GRNs and issues go digital in weeks, with the paper register running parallel until physical counts agree three times running. Production second, once material truth is stable: work orders mirror the floor’s actual flow, supervisor tablets replace the whiteboard-photo ritual. Costing last, when real data flows — because costing built on bad stores data is fiction with decimals.
Two adoption truths most vendors skip. The floor is the customer: capture designed for supervisors’ hands, in their language, at their pace — an ERP the floor routes around is an expensive report generator. The CA is the veto: statutory accounting stays in Tally via the bridge, the CA blessed early — because month-three CA vetoes have killed more factory ERPs than every technical failure combined.
What’s the honest pitch?
This ERP will not make your machines faster. It makes the commercial half of the factory as precise as the production half: material that reconciles, job-work that settles on evidence, costs known while they’re correctable, dispatches that bill complete. The margin recovered from those four — material, job-work, purchase and costing — is the product’s real price justification, visible in the first quarter’s own reports — most factories find it in the first quarter, usually in the job-work ageing and the purchase rate drift, chhota-chhota amounts that compound exactly like the interest nobody charged on them.
The engineering behind it is auditable before you buy: the event-sourced architecture that makes stock movements tamper-evident, the staged migration method, and fourteen published teardowns showing how I reason about systems. The trade context lives at /industries/automotive and /industries/retail-systems; the distribution sibling at /products/distribution-erp.
What should the first quarter’s numbers look like?
Concrete expectations, because “efficiency” is not a deliverable. Stores accuracy: physical-versus-system variance falls to low single digits within two cycles of counts, because every movement now leaves a trail — and shrinkage that was invisible becomes an exception report with names on it. Job-work recovery: the ageing report surfaces material sitting at vendors beyond norms in week one; most factories recover more in the first quarter than the deployment cost, purely from challans nobody was chasing. Purchase drift: rate history flags the items whose prices crept between orders — the quiet over-billing that thrives on missing memory. Costing truth: the first honest per-batch cost sheet lands within the quarter, and at least one product every factory believed was profitable turns out not to be. That last discovery is uncomfortable and worth the entire project — pricing built on real costs is the most underrated margin lever in manufacturing.
Related reading
/work/erp-migration — the case study · /industries/automotive — the workshop-side context · /products/distribution-erp — the distribution sibling · /blog/zero-downtime-migration — the method
The month-end test
If month-end costing at your plant is a detective episode — bring last month’s cost sheet to the demo and we’ll rebuild it live from real flows. Factory chaalu, hisaab barobar: /contact.
Questions I actually get
Is shop-floor data entry actually realistic with our workforce?
Yes, because the capture is designed for hands that run machines, not keyboards: supervisor tablets, barcode scans where they help, forgiving forms with big targets, and Hindi or Marathi labels where the floor wants them. If a supervisor can use WhatsApp, the floor terminal will not slow them down — that is the design bar, tested on real floors.
We run multiple units. Does it consolidate?
Per-plant stores, planning and costing with a consolidated owner view — transfers between units tracked like the inter-company transactions they are. Each plant sees its own floor; the owner sees the group truth without phone calls. The multi-entity pattern is the same one every product on this site uses.
We are on Tally plus Excel. How scary is migration?
Staged and boring, by design: stores and purchase go live first because they pay back fastest and risk least, production follows once material truth is stable, and costing arrives last when real data flows. Registers run parallel until counts agree repeatedly. The factory never stops — the full method is written up in the ERP migration case study.
Does our CA keep Tally?
Yes — the Tally bridge sends clean vouchers to statutory accounting, so your CA's world is undisturbed and fed better data than manual entry ever gave it. Winning the CA's blessing early is a deliberate step in every deployment, because a veto from that desk in month three has sunk more ERPs than any technical failure.
What about job-work — we send material out and take work in?
Both directions are first-class: your material at vendors tracked challan by challan with ageing alarms, and inbound job-work with material accountability to the principal. The 57F4-style reconciliation that usually lives in one person's memory becomes a report — and the disputes at settlement shrink to arithmetic.
What does it cost, and how is it priced?
Deployment plus annual support, sized by plant count and modules — not per-seat rent that punishes you for training more staff. As a benchmark it prices at a fraction of the enterprise ERP tier, because you are buying the system a mid-size factory needs, not eleven brochure modules. The scoped quote follows the first call, fixed.