Here is a scene from almost every factory I visited in the two years before we started building OEMup.
The owner’s phone buzzes during a plant walk. It is a WhatsApp message with a photo of a drawing and one line: “Sir, can you make this? 40 nos.” He glances at it, says “send it to me properly”, and puts the phone back in his pocket. Twenty minutes later there is a vendor at the gate, and by evening there are ninety other messages.
Nobody lost that enquiry on purpose. It simply never became one. It was never written down, so it was never counted, never assigned, never followed up — and when the month-end sales review happens, it does not appear as a loss. It does not appear at all.
Your conversion rate is measuring the wrong denominator
Take a ₹6 crore fabrication and machining job shop — the same shop we costed jobs for in an earlier article. Here is a month.
| Stage | Count | What happened |
|---|---|---|
| Touches that arrived | 60 | Every ping, call, form and message across all channels |
| Recorded anywhere | 23 | 37 were never written down |
| Quoted | 18 | A costed quotation went out |
| Won | 6 | Converted to an order |
Quote → order = 6 ÷ 18 = 33% — the number that gets reported
Touch → order = 6 ÷ 60 = 10% — the number that is true
Thirty-three per cent sounds like a business that knows what it is doing. And the 33% is not a lie — it is an accurate measure of the stage where the process actually functions. That is exactly the problem. The one metric every factory tracks measures the healthiest part of the funnel and is structurally incapable of seeing the sickest.
And here is the honest part: some of those 37 were junk. A trader in another state wanting two pieces, somebody collecting rates, a wrong-number enquiry for a product you do not make. The point is not that you lost 37 orders. The point is that you cannot say which they were. An unrecorded lead is not a loss you can learn from — it is a blind spot, and blind spots do not show up in reviews.
Where leads actually arrive
Break that 60 down by how it reached you, and the shape of the problem becomes obvious.
| Channel | Count | Share | Character |
|---|---|---|---|
| IndiaMART / TradeIndia | 22 | 37% | Highest volume, lowest intent. You are one of five shops sent the same RFQ. |
| WhatsApp to owner or sales | 14 | 23% | High intent, zero record. Lives and dies in a personal phone. |
| Phone call | 9 | 15% | High intent. Nothing written down unless somebody chooses to. |
| Website form | 8 | 13% | Mixed intent, and frequently unanswered for days. |
| Exhibition / referral | 7 | 12% | Highest intent of all. Business cards in a drawer. |
| Total | 60 | 100% | — |
Notice the inversion. The channel that produces the most enquiries produces the worst ones — and it is the only channel that reliably gets recorded, because it arrives in something that looks like an inbox. The three channels carrying your best enquiries are precisely the three that leave no trace.
That is not a discipline failure by the owner. It is a design failure: nobody ever gave a WhatsApp message anywhere to go.
One row is missing from that table for a good number of manufacturers. If you sell through dealers or channel partners, they are a channel too — and the least visible one of all, because their enquiries sit in somebody else’s book. You see the orders they place and never the demand they were working on, which means you cannot tell a dealer who is quietly losing your enquiries from one who simply has a slow month. Add the row, even if you have to ask for the number.
Nearly a quarter of this shop’s enquiries exist only in one person’s WhatsApp. When that person is travelling, on leave, ill, or leaves the company, the pipeline goes with them — and no one else can even establish what was lost. This is the same structural risk as running your supplier chase on a WhatsApp group, pointed at the revenue side instead of the purchase side.
Capture is cheap. The discipline is thirty seconds.
The instinct when someone says “capture every lead” is to imagine a form with twenty fields that nobody will ever fill in. That is the wrong shape of fix, and it fails within a fortnight.
Five fields is enough:
- Company and person — who is asking
- Channel — where it came from, because you cannot judge a source you do not measure
- What they want — one line, or a photo of the drawing
- Owner — the person responsible for the next move
- Next-contact date — the single field that converts a record into an action
Thirty seconds, and the message stops being private. Everything else — the quantity, the drawing revision, the target price, the delivery expectation — can be filled in at qualification, by which point you have decided the lead is worth the typing.
The next-contact date is doing more work here than it looks. A lead without one is a note; a lead with one appears on somebody’s list on a specific morning. It is the same mechanism that makes the daily call list for overdue invoices work, applied at the other end of the same revenue cycle.
Qualification: four questions a job shop actually needs
Generic sales frameworks ask about budget, authority, need and timeline. For an Indian job shop those are the wrong four questions. These are the right ones, in this order:
Process, size envelope, tolerance, certification. Not “could we figure it out” — do we already run this? A fabrication shop quoting a job that needs a machining centre it does not have is buying a problem, and half of those quotes are won by accident.
Four pieces and forty pieces are different businesses. Setup is a batch cost, so the same part at a quantity of four can cost ten times as much per piece to produce as it does at forty. If the enquiry is for a quantity you cannot make money on, the answer is a polite no, not a cheap quote.
A GST number, a traceable business, terms you would accept. The cheapest moment to have this conversation is before you have spent estimating hours — and certainly before material is committed against the order.
Is there a drawing, a target date and a named person who owns this at their end? Or is somebody collecting three rates to benchmark the supplier they already use and intend to keep? Every job-shop owner in Gujarat knows this enquiry. Almost none of them have a way to mark it.
A lead that fails any one of the four does not deserve estimating time. Which raises the question nobody in an SME factory asks out loud: what is estimating time worth?
Quoting is a bottleneck. Nobody counts it.
A serious quotation for a fabricated assembly is not a price off the top of somebody’s head. It is a drawing study, a bill of materials, a costing run and at least one follow-up call. Call it 2.5 hours, which if anything is generous.
Cost that properly, the way we costed a shop-floor hour in the job costing article — cost of employment, not wage; productive hours, not paid hours.
| The estimator | Amount |
|---|---|
| Gross monthly wage | ₹35,000 |
| Employer PF, bonus and gratuity accrual | ₹5,050 |
| Cost of employment | ₹40,050 |
| Productive hours a month | 155 |
| Real hourly rate | ₹258 |
| Cost of one 2.5-hour quotation | ₹646 |
At 40 quotations a month that is ₹25,839 a month, or ₹3.1 lakh a year spent on quoting. Worth knowing, but the rupees are not the real constraint.
If estimation is about half of that person’s job, they have roughly 78 productive hours a month for it — a hard ceiling of about 31 serious quotations. Push 40 through and you are running a 129% load on a bottleneck. Nothing breaks visibly. What happens instead is that every quote goes out three days later than it should, and in an SME market a meaningful share of orders go to whoever answered first.
This is the argument that turns lead qualification from admin into operations. You already ration machine time — that is what capacity planning is. Estimating capacity is a constraint with exactly the same arithmetic and none of the attention. Qualifying a lead is not bureaucracy; it is deciding what to load onto a bottleneck you already own.
Five stages, each defined by an event
You do not need an elaborate pipeline. Five stages carry a job shop comfortably:
| Stage | What moves it here | Who owns it |
|---|---|---|
| New | Captured. Nobody has looked at it properly yet. | Whoever received it |
| Qualified | Passed the four questions. Worth factory time. | Sales |
| Quoted | A costed quotation was sent, on a date you can name. | Sales |
| Negotiation | They responded. Price, delivery or terms are being settled. | Sales / owner |
| Won / Lost | An order, or a closed lead with a reason attached. | Sales |
Every stage must be defined by something that happened, not by how somebody feels about it. “Warm” is not a stage. “Quotation sent on the 12th” is. If you cannot name the event that moves a lead from one stage to the next, the stage is decoration — and a pipeline of decorative stages produces a forecast nobody believes, including the person who made it.
The most valuable field on the form
Most factories close a lost lead by deleting it or letting it go stale. The lost reason is the only field in the whole exercise that teaches you anything, and it works only if the list is short and fixed so the answers can be counted:
- Price — we were beaten on rate
- Delivery — we could not meet the date
- Capability — we could not make it to spec
- No response — the customer went quiet
- Lost to incumbent — they were always going to stay
- Chose not to quote — we declined, deliberately
Then read the counts properly, because the obvious reading is usually wrong.
It is usually one of two other things. Either a qualification problem — you are quoting work you were never positioned to win, and the rate-benchmarkers are counted as genuine losses. Or a costing problem: a shop that absorbs overhead as a percentage of material value will systematically over-price its quick, material-heavy jobs and under-price its slow, work-heavy ones, so it loses precisely the quotes it should have won. That mechanism is worked to the rupee here, and the lost-reason field is where it first becomes visible.
A large “no response” count says something different again — usually that quotes are going out too slowly, or that follow-up stops after one attempt. Both are fixable, and neither is visible without the field.
What one more order a month is worth
All of this has to earn its place against running the factory, so price it.
This shop wins 6 orders a month, 72 a year, on ₹6 crore of turnover — an average order of about ₹8.3 lakh.
One extra won order a month = ₹8.3 lakh × 12 = ₹1 crore a year
At a 30% gross margin, roughly ₹30 lakh of additional gross margin — from winning 7 of the same 18 quotes instead of 6. That is a move from 33% to 39%.
Six percentage points is not a heroic target. It is what you would expect from quoting the right 18 leads instead of the first 18, and answering them three days sooner.
Six ways this goes wrong
1 · The pipeline lives in one person’s phone
A quarter of enquiries arrive by WhatsApp to an owner or salesperson and stay there. No one else can see them, cover them during leave, or reconstruct them after a resignation.
Cost: the highest-intent channel in the business is also the least visible one. Fix: one destination for every channel, five fields, thirty seconds. The discipline is the destination, not the detail.2 · Everything gets quoted
Without qualification, the loudest and the earliest enquiries consume the estimator, regardless of whether they were ever winnable.
Cost: a 129% load on the one person who prices your work, and the good quotes going out late. Fix: the four questions, before any estimating time is spent. A declined enquiry is a decision, not a failure.3 · Stages describe feelings
“Hot”, “warm”, “following up”. Nobody can audit them, two people grade the same lead differently, and the pipeline total means nothing.
Cost: a forecast that cannot be trusted, so it stops being used, so the pipeline stops being maintained. Fix: name the event for each stage transition. If there is no event, there is no stage.4 · Lost leads close with no reason
The lead goes stale, gets deleted, or sits at “following up” forever. Nothing is counted, so nothing upstream ever gets fixed.
Cost: the same losses repeating for years with no mechanism that could ever surface the pattern. Fix: six fixed reasons, mandatory on close, reviewed as counts once a month.5 · Leads and customers live in different systems
The lead is in a spreadsheet, the customer is in the accounting software. On conversion somebody re-types the name, the GST number and the address — and the enquiry history does not come with it.
Cost: re-keying, transcription errors on GST details, and a customer record with no memory of how the relationship started. Fix: the lead should become the customer, carrying its history, not get copied into a second system.6 · Quote-to-order is called “the conversion rate”
It is a real metric measuring a real stage. It is just not the funnel, and treating it as the funnel means the largest leak is definitionally invisible.
Cost: 62% of arrivals outside the measurement entirely, with no way to tell junk from missed opportunity. Fix: report both numbers side by side — touch-to-order and quote-to-order. The gap between them is the size of the problem.The weekly routine
Five fields, one destination, every channel. A lead that is not written down the day it arrives is usually never written down.
Run the four questions over everything in New. Batching it takes twenty minutes and stops qualification from becoming an interruption.
Count what is Qualified against what the estimator can realistically produce. If it is over, choose — deliberately, and out loud.
Not the whole pipeline — just whatever is due today. Ten minutes, and it is the same habit that runs your payment follow-up.
Touches, qualified, quoted, won, and the six lost reasons as counts. Five numbers and a list — that is the whole review.
Where this lives in OEMup
Everything above works on a whiteboard for a month, and starting there is not a bad idea. What a system has to solve first is the plumbing problem — giving every channel somewhere to land, so that the enquiry exists in a place other than a phone. In OEMup that is four routes into one pipeline.
For everything that arrives as a human conversation — the WhatsApp message during a plant walk, the phone call, the card from an exhibition, the marketplace RFQ somebody read in an inbox. This is the unglamorous one and it closes the largest gap in this article, because in the worked example those channels are half of everything that arrives and have no other way in.
A form you put on your own website that writes straight into the pipeline. The enquiry is captured with its arrival time and owner rather than sitting in a shared mailbox someone opens on Thursday, and nobody re-types it — which removes both the delay and the transcription error in a phone number.
Leads generated from your Google Ads campaigns land in the same pipeline as everything else. That matters for more than convenience: once ad-sourced leads sit beside the rest, you can judge the spend against orders won rather than against clicks and form fills, which are the only numbers the ad platform can show you.
If you sell through dealers or channel partners, they submit enquiries into your pipeline directly. Without this, dealer-sourced demand lives in the dealer’s own book and reaches you only as a firm order or not at all — so you see the wins and never the enquiries, which is the same blind spot as the WhatsApp problem, one step further away.
What the four routes actually buy you
Two things, and the second is the one people underestimate.
The first is simply that the enquiry exists. It is owned by a named person, it can be picked up when that person is travelling, and it survives a resignation. That alone moves the recorded count in the worked example off 23.
The second is that the channel field stops being a guess. When a lead arrives through its own route, the source is recorded rather than remembered — so the comparison at the top of this article becomes a report instead of an exercise. You can finally answer the question every factory owner asks and none can evidence: is the marketplace subscription worth it, and are the dealers actually sending anything? A source you do not measure is a source you cannot judge.
Capture is the part software can fix outright. Qualification is not — no system will tell you that the buyer asking for three rates already has a supplier he intends to keep, or that a quantity of four is not worth your setup. The four questions earlier in this article stay with the person who knows the market. What the four routes guarantee is that the lead is in front of that person in the first place, which is where most Indian SME factories lose it.
Downstream, because the CRM sits inside the ERP rather than beside it, a won lead does not get re-keyed: the customer order is built as a quotation with tax, HSN and price auto-populated and its own payment terms, converts into a delivery order and then a GST tax invoice, and hands to production as a manufacturing order — the path covered in the CRM article and, at the far end, in Getting Paid.
See one enquiry go all the way through
Book a 30-minute demo and bring a real enquiry — we’ll capture it, put it in the pipeline with an owner and a next-contact date, and take it through to a quotation, a sales order and a manufacturing order, so you can see the whole path from ping to shop floor.
Book a Demo →The bottom line
Lead management in a factory is not about selling harder. It is about two decisions made earlier than they are made now: what gets written down, and what gets an estimator’s afternoon.
The first is a plumbing problem — give every channel one place to land, take thirty seconds, and the highest-intent enquiries in the business stop living in a personal phone. The second is a capacity problem, and it deserves exactly the seriousness you already give to machine loading, because the estimator is a bottleneck whether or not you have ever called them one.
If you do one thing this week, count. Count every enquiry that arrives, through every channel, for five working days — a tally on paper is fine. Then compare it with how many made it into whatever you currently call the enquiry register. On most shop floors that number lands somewhere near a third, and the gap is the argument for everything above.
Related reading: Manufacturing CRM in India for what happens after an enquiry is in the system, How to Cost a Job for the costing behind a quotation that wins, Getting Paid for the far end of the same revenue cycle, and Capacity Planning for Small Factories for rationing a bottleneck when it is machines rather than people.
FAQ
What is the difference between a lead and an enquiry in a manufacturing business?
An enquiry is something you have decided to work on. A lead is the same thing before anybody has decided. The distinction matters because the decision is the valuable part — can we make it, is the quantity worth the setup, will they pay, and is there a real drawing and date behind it. Treat every lead as an enquiry and you spend estimating hours on work you were never going to win; treat none of them as enquiries and they die in a personal phone.
Why is my quotation conversion rate misleading?
Because the denominator only contains what somebody chose to write down. A shop taking 60 enquiries a month might record 23, quote 18 and win 6 — which reads as 33%. Measured against everything that actually arrived it is 10%. The quote-to-order rate measures the part of the funnel that already works and hides the 62% that was never recorded, where you cannot even say whether the lost ones were junk or gold.
How should a factory qualify a sales lead?
Four questions, in order. Can we make it — is the process, size envelope and certification within what we actually run? Do we want to make it — does the quantity justify the setup, or is this four pieces quoted at forty-piece rates? Will they pay — GST number, traceable business, terms you would accept? And is it real — a drawing, a date and a named person, or somebody collecting rates to benchmark the supplier they intend to keep? Fail any one and it does not deserve estimating time.
Where do manufacturing leads actually come from in India?
In a typical SME job shop taking 60 enquiries a month: about 37% from marketplaces like IndiaMART or TradeIndia, 23% as WhatsApp messages straight to an owner or salesperson, 15% by phone, 13% through the website form and 12% from exhibitions and referrals. The marketplace leads are the highest volume and the lowest intent; the WhatsApp and referral leads are the highest intent and the least likely to be written down anywhere anyone else can see.
How much does it cost to quote a job?
More in capacity than in rupees. A serious quotation — drawing study, BOM, costing, follow-up — takes about 2.5 hours. At an estimator’s cost of employment near ₹40,000 a month over 155 productive hours, that is roughly ₹646 per quote, or ₹3.1 lakh a year at 40 a month. The bigger constraint is the queue: if estimating is half of one person’s role the real ceiling is about 31 quotes a month, so 40 is a 129% load and everything goes out late.
What pipeline stages should a manufacturer use?
Five are enough: New, Qualified, Quoted, Negotiation, Won/Lost. The rule that makes them useful is that each must be defined by something that happened, not by how somebody feels. “Warm” is not a stage; “quotation sent on the 12th” is. If you cannot name the event that moves a lead forward, the stage is decoration.
Why should I record a reason for every lost lead?
It is the only field that teaches you anything. Keep the list short and fixed — price, delivery, capability, no response, lost to incumbent, chose not to quote — so the reasons can be counted. And read it carefully: if price is 70% of losses, that is usually a qualification or a costing problem rather than a pricing one. A shop absorbing overhead on material value will systematically lose the quotes it should win, and this field is where that surfaces first.
What is one extra won order a month actually worth?
On a ₹6 crore job shop winning 6 orders a month, the average order is about ₹8.3 lakh. Winning 7 instead of 6 — moving from 33% to 39% on the same 18 quotes — is ₹1 crore of additional revenue a year, roughly ₹30 lakh of gross margin at 30%. That is what decides whether lead management is administration or the highest-return hour in your week.
Manufacturing cost calculator · BOM cost calculator · Fabrication cost calculator · Profit margin · Production capacity — all free, no sign-up.