Choosing Yours

Nine Lines in Green Ink

Her list came by WhatsApp at 11.40 one night, a photograph of a ruled page from a school register, slightly out of focus. Nine businesses, numbered in green ink, with a capital figure pencilled beside each. Masala grinding. Paper cups. Phenyl and floor cleaner. Agarbatti. Fly-ash bricks. Cotton buds. Detergent powder. Papad. Corrugated boxes, which I suspect was a courtesy to me.

I'll call her Meena. She had taught accounts at a girls' school in Nashik Road for eleven years and had ₹12 lakh, part savings and part a gold loan she had not yet told her brother about. She had read every entry twice. What she did not have was any way of crossing eight of them off.

I sent her back a blank grid. Four rows, one column for each business, and a box at the bottom for the total. I have used that grid for about fifteen years with people who come up the stairs to my office, and I call it the Four Constraints sheet because I could never think of a cleverer name.

Four Rows, Scored One to Five

Each row asks a single question, and each answer gets a score from 1 to 5. Five is always the good end.

• Capital. Can you set this up, and fund three months of raw material and wages, with the money you actually hold? Not the money a bank might lend you. 5 means comfortably, with a cushion. 1 means only if everything goes right.
• Raw material within 200 km. Is there a dealer or mill within a day's truck journey who will sell you your main input in the quantity you need? 5 means three or more suppliers you have phoned. 1 means it comes from another state or another country.
• A named buyer reachable in 90 days. Can you write down the name of a person, a shop or a purchase department that could place an order within three months of your first batch? 5 means you have spoken to them and they have asked for samples. 1 means "the market".
• Licence and pollution difficulty. How hard is the paperwork for this product in your state? 5 means a white or green category and a simple registration. 1 means red, an effluent plant and estates that will not have you.

Add the four. The highest possible score is 20. Anything below 12 means choose again. Not "try harder". Choose again.

Scoring as a Stranger Would

The sheet only works if you are dishonest in the opposite direction from the one you want to be. Everyone inflates the buyer row. I did.

If I had filled in this grid in March 2004, I would have given myself a 4 for buyers, because I had a cousin in a Pimpri pharma distribution firm who said he could "definitely put me in touch". The honest score was a 2. My cousin put me in touch with one storekeeper, who took a sample carton, put his tea on it and never rang back. My first real order came eleven weeks later from a man I found myself, standing outside a Bhosari gate at 8 a.m.

So I give people three rules for scoring.

First, a score needs a name or a number behind it. A 4 on raw material means you can tell me the dealer's name and today's price per kilogram. A 4 on buyers means a person and a phone number, not a sector.

Second, capital includes working capital. A detergent unit with ₹7 lakh of machines and no money for soda ash scores a 1, however cheap the mixer was.

Third, ask someone who does not love you to check it. Your spouse will score you high out of loyalty or low out of fear. A raw-material dealer, a bank officer or a man who already runs the business will score you closer to the truth, usually lower.

Meena sent her first filled-in sheet two days later. Seven of the nine businesses scored between 13 and 16. She was delighted.

I looked at the buyer row. Every box said 4. Not one of them had a name written beside it.

Meena Crosses Out Six

I rang her that evening. "Put a name next to every 4," I said. "If you can't, make it a 2."

"Even if I know the shops exist?"

"Everyone knows the shops exist," I said. "That's why it's a 2."

The second sheet arrived on a Sunday, and it looked like a battlefield. Papad fell to 10. Agarbatti fell to 9, because the only buyer she could name was the temple near her mother's house, which already bought from a man in Sinnar. Fly-ash bricks never had a chance; she had no thermal plant contact and no idea who builds in Nashik. Six businesses dropped below 12 in an afternoon.

Three survived:

Business	Capital	Raw material	Buyer	Licence	Total
Phenyl and floor cleaner	5	4	4	3	16
Masala grinding	4	5	2	4	15
Cotton buds	3	2	2	4	11

The phenyl 4 now had names beside it. Eleven years of school accounts meant she knew the bursars of three schools on the Nashik Road side, the warden of a working women's hostel and the man who held the housekeeping contract for two of those schools. She had signed their invoices. She knew what they paid for a 5-litre can of black phenyl (₹180, from a supplier in Ambad, delivered late about one month in four).

Masala looked better on paper because spices are everywhere in Nashik. But "everywhere" is a raw-material score. She still could not name a single shop that would stock an unknown brand. Phenyl it was.

Harpreet Has the Lathes Already

Harpreet's sheet came from Ludhiana with fewer businesses and more confidence. His father had run a lathe shop off Gill Road since 1991, six conventional lathes, mostly turning jobs for whoever walked in. Harpreet, 29, wanted to put in two CNC machines and make turned components properly.

He scored it 18. I made him justify each box. Capital was genuine: the shed was family-owned, the lathes were paid for, and he had ₹15 lakh plus his father's goodwill with a machine dealer. Raw material within 200 km in Ludhiana is barely a question; the city is made of steel bar. Licence was a 4.

The buyer row said 5, and here, for once, I believed it. A tractor-parts assembler in Focal Point had given him an actual drawing for a hydraulic pin, 4,000 pieces a month, with a tolerance his father's manual lathes could hold only on a good day. He had the purchase officer's mobile number. He had already sent fifty samples, turned by hand.

"So I can order the CNC?" he said.

"You can order one," I said. "When the second drawing comes, order the other."

Arun Wanted the Big Numbers

Arun was the hard one, and I liked him a great deal. Fourteen years in software in Bengaluru, ₹40 lakh saved, a wife who was quietly worried, and no day of his life spent inside a factory. His first choice was PET preforms, the test-tube-shaped blanks that get blown into water bottles. He had read that bottled water was growing and had built a spreadsheet with forty tabs.

His own sheet scored it 15. Mine scored it 11.

Capital came down to a 3 once I asked him to price the moulds as well as the injection machine, and to hold three months of resin at prices that had moved by close to 20 per cent within a single year. His buyer row said "bottling plants, Karnataka". No name. When he did call two bottlers near Hoskote, both already had preform suppliers delivering daily and asked whether he could beat them by a rupee per thousand.

"So I just drop it?" he said.

"You score it again in three years, when you know what a mould costs to repair," I said.

His second sheet was humbler and better. For eight years he had sat through meetings with the facilities team at a Whitefield tech park, and he knew the canteen operator who served four of its buildings, about 9,000 paper cups a day. He had the man's number and had once helped his son with a college application.

Paper cups. Crowded, thin margins, and I told him so. Capital 5, raw material 4, buyer 4, licence 4.

Seventeen. He sent it with a single line underneath: "Less exciting. Is that the point?"

Raw Material or Buyer: Which Should Be Closer

Arun's next question was where to rent. He had found a shed near Nelamangala, on the far side of Bengaluru from Whitefield, for ₹9 a square foot less than anything east of the city. Cheaper rent is the first thing every founder sees and the last thing that matters.

The answer depends on one physical fact: does your product get heavier and bulkier on the way through the factory, or lighter? Whichever end of the process is harder to move is the end you sit next to.

My own trade makes the point better than any diagram. Kraft paper arrives in reels, dense and tightly wound, and a truck carries close to its legal weight. The boxes that leave are mostly air. Even flat-packed and strapped in bundles, they fill a truck long before they get heavy. Here is roughly what one 32-foot truck carried for us last year:

Load	Approx. value per truck	Freight, 40 km
Kraft reels in	₹6.5 lakh	₹5,200
Finished 5-ply boxes out	₹1.6 lakh	₹5,200

Same truck, same diesel, same driver. Freight on the outbound side costs about four times as much per rupee of goods. That is why box makers cluster around their buyers in Chakan, Ranjangaon and Bhosari, and why I have never minded that the paper mills are 150 km away in Kolhapur district. Paper travels well. Boxes do not, and they hate rain.

Meena's phenyl works the same way, more extremely. Black phenyl is mostly water, and the water comes out of her own tap. The concentrate and pine oil she buys from Ambad would fit in a tempo; the finished cans for three schools and a hostel need several trips a month. She should be within a short drive of the schools on the Nashik Road side, which is where she already lives. That rarely happens so neatly.

Now reverse it. Take dehydrated onion, which a Nashik founder might reasonably consider, since Lasalgaon is the largest onion market in the country. Fresh onion loses something like 85 to 90 per cent of its weight in the drier. Ten tonnes in, a little over one tonne out. You would be mad to truck wet onions 300 km to a shed near your buyers in Mumbai. You put the drier near the mandi and send the light, dry flakes wherever they need to go. Cold-pressed groundnut oil is similar, though less dramatic: the oil is roughly a third of the seed's weight, and the cake left behind sells locally as cattle feed.

Harpreet's case shows the third possibility. Steel bar is heavy and his hydraulic pins are heavy, so freight hardly decides anything. What decides it is the buyer's habits. A tractor assembler wants parts delivered in small lots, often twice a week, and when a batch is rejected he wants someone at his gate within the hour with the gauge and an explanation. Gill Road to Focal Point is under half an hour. For engineering components that closeness to the customer is worth more than any saving on bar. Karthik in Hosur told me that his vendor code was partly decided by the fact that he could be on the assembly line in twenty minutes.

So the rule of thumb runs like this:

• Product gets bulkier or heavier (boxes, cups, phenyl, bricks, foam): sit near the buyer.
• Product gets much lighter (dried vegetables, oil, rice milling, stone crushing): sit near the raw material.
• Both dense, buyer demanding (turned parts, fasteners, gaskets): sit near the buyer for service, not freight.

I am less certain about products sold through distributors across a whole state, where the "buyer" is a godown that can move. There I would sit wherever the labour and the power are reliable and accept the freight.

Arun's cups fall plainly in the first group. Paperboard reels come in tight; nested cups go out in cartons that are nine-tenths air, and the canteen wanted deliveries every three days. Nelamangala to Whitefield is 50 km through the worst traffic in the country. I asked him to price the extra freight over a year against the ₹9 he would save on rent.

He rang back the next morning. "The rent saving is ₹1.6 lakh," he said. "The freight is ₹2.3 lakh. And that's if the driver turns up."

By Thursday he was looking at sheds in Hoskote.

Sell Someone Else's Product for Thirty Days

Before any of the three signed a lease, I asked them to do something that sounds backwards: sell the product for a month without making it. Find a working unit, buy its finished goods at whatever price it charges, put them in front of the buyer named on your sheet, and see what happens. Buyer before machine. If the buyer will not take a product you bought in, he will not take one you made yourself at three times the investment.

I did not do this. I bought Vilas's machines first and found my customers afterwards, which is the order in which most people lose their savings. I was lucky, and I have watched luckier men than me run out of it.

Arun went to a cup maker in Peenya, a two-machine unit run by a man called Ravi, and bought 2,70,000 single-wall cups, a month's supply for the canteen, at ₹0.48 each. He sold them to the canteen operator at ₹0.55. On paper that was ₹18,900 for thirty days of driving, phoning and loading cartons into a hired tempo. In practice he learned four things that no spreadsheet had told him:

• The canteen paid on day forty-one, not day thirty, and only after Arun went in person.
• One batch of about 6,000 cups leaked at the side seam when filled with hot sambar. The operator deducted the whole carton and a little more for his trouble.
• The operator wanted 150 ml cups for coffee and 210 ml for tea, which meant two sizes and two sets of dies.
• A second canteen in the same tech park asked for a quote after seeing the cartons arrive.

The leaking seam was the most useful ₹2,880 he has ever lost. He now knew which defect his customers would notice first, and he asked Ravi exactly how it happened. Ravi said, "Paperboard was damp. Monsoon." Arun wrote that down.

Meena's month was simpler and more surprising. She bought 400 litres of black phenyl from a small unit in Ambad at ₹19 a litre, in 20-litre cans, and supplied her three schools and the hostel at ₹26. Two schools paid within a week. The hostel warden paid in cash on delivery. The third school's bursar, a careful woman named Mrs Joshi, refused the 20-litre cans because the ayahs, the cleaning staff, could not lift them onto the shelf, and asked for 5-litre cans instead. That changed Meena's packaging cost by nearly ₹2 a litre and told her exactly what her filling line should look like.

Harpreet's version was shorter, since he could already turn the pins. What he could not do was harden them, so for the first trial order of 800 pins he paid a heat-treatment shop on Industrial Area A ₹14 a kilo and stood beside the furnace for half a day. The tractor assembler rejected 31 pins for a hardness reading at the low edge of the drawing. Harpreet learned which number the quality inspector actually checks, and which furnace operator in Ludhiana he would trust with his name.

Keep a notebook for the thirty days. One line a day: what you delivered, what you were paid, what was rejected and why, and anything the buyer said he wanted that you had not offered. At the end, count three things. Did the buyer reorder without being chased? Did he pay within the terms he promised? Did your margin survive the rejections? Two yeses and you have a business worth scoring again with real numbers. One, and you have saved yourself ₹12 lakh, or ₹40.

There is a practical catch, and it arrived for all three in the same week. Schools want a proper invoice. Tech-park canteens want a GST number on it. Tractor assemblers want both, plus a vendor registration form with a PAN, a bank letter and a cancelled cheque. You cannot sell someone else's product for long without becoming, in the eyes of the law and the bank, a business.

Meena rang me on day twenty-two. Mrs Joshi had placed a standing order for 60 litres a month in 5-litre cans, and the school's accountant had asked for her GSTIN before he would release the payment.

"I don't have one," she said. "Where do I even start?"

I told her to bring a photocopy of her PAN card, her Aadhaar, her electricity bill and a blue file. The file was important. It was about to get very thick.