Three decisions that each cost something visible

Building one multipurpose block at Sarigam when the plan was a new site

By mid-2022 our enquiry book for custom synthesis and halogenated intermediates had nearly tripled in eighteen months. The sales head and our managing director proposed a ₹610 crore greenfield plant near Jhagadia, with 380 kilolitres of reactor volume and its own effluent treatment, to be commissioned in 2025. Its case rested on enquiries from buyers seeking a source outside China. With revenue heading towards ₹1,240 crore that year, the project would have committed roughly half a year's sales.

I recommended something smaller. When my office sorted the enquiries by stage, about a quarter had reached validated batches, and signed agreements covered less than a fifth of the proposed volume. I proposed a ₹236 crore multipurpose block at Sarigam with 140 kilolitres of reactor volume, plus a rule that further dedicated capacity would need signed agreements covering 60 per cent of its first three years. We kept the Jhagadia plot allotment for ₹14 crore.

The sales head disagreed, arguing that customers would not sign until they saw capacity being built, and he had a point. A US crop-protection company wanted two intermediates from dedicated equipment by mid-2024. Our smaller block could not provide them, and the contract, worth about ₹90 crore a year, went to another Indian producer. The board approved the Sarigam block in November 2022.

It was commissioned in March 2025, four months late after a scrubber delivery slipped. Meanwhile, in 2023–24, crop-protection customers cut inventories, Chinese prices for several of our intermediates fell by about a third, and our revenue dropped to ₹1,050 crore. A greenfield plant would have opened into that market. In 2025–26 the block ran at 58 per cent of available reactor hours and earned ₹96 crore of revenue. The company now applies one evidence standard to every capacity proposal.

What I would do differently is give the sales head a way to test his argument: we could have offered the US customer shared capacity with an expansion trigger instead of treating the contract as a yes or no.

Selling the dye-intermediates business our founder started

Saranvik began in 1984 making reactive-dye intermediates in Vapi, and for twenty-five years that was the company. By 2023 it was a small part of a larger group: ₹118 crore of revenue in 2023–24, an operating margin below 8 per cent against about 18 per cent elsewhere, and a heavy claim on our Vapi effluent-treatment capacity. Chinese material set its prices. Our chairman, Jayantilal Sodhavaria, had founded it, and his younger son Pratik, an executive director, had run it for twelve years.

In early 2023 I proposed that we look for a buyer. I had learnt that such a proposal could not arrive as a surprise in a board paper. I spoke first to Pratik, then to the chairman and the managing director, and only then took the analysis to the board. The chairman opposed the sale for most of a year. His concern was not sentiment alone: the business had supplied Surat customers through every cycle, and he doubted a buyer would keep its people. Pratik accepted the economics earlier than his father did, which cannot have been easy for him.

We agreed three conditions before approaching buyers: a sale as a going concern with its Vapi unit, transfer of every employee on existing terms, and six months' notice to customers. Ravelkund Dyes, a Surat reactive-dye maker that was already the unit's largest customer, bought it for ₹68 crore in September 2024, taking 164 employees. Our independent directors reviewed the valuation with an external adviser, and the board approved the sale with the chairman's reservations recorded in the minutes.

We gave up a stable, low-return revenue line and part of the company's identity. The proceeds funded part of the Sarigam block and repaid ₹40 crore of working-capital borrowing, and Pratik now leads performance chemicals. What I would do differently is set exit criteria for every business in 2021, before the dye unit's weakness made the discussion feel like a verdict on one member of the family.

Reducing our reliance on Brakenholm Agro while its orders were falling

In 2022–23 Brakenholm Agro, a European crop-protection innovator, accounted for 31 per cent of our revenue, about ₹384 crore, across four intermediates for its fungicide and herbicide range. The relationship was the foundation of our custom-synthesis business, and I had built much of it. It also meant that one customer's inventory decisions could move our quarterly results more than anything that happened in our plants.

In May 2023 I proposed to the board that no customer should exceed 20 per cent of revenue by 2026–27. The difficult part was the resources behind that limit. Our process-development team had twenty-two chemists and two kilo-laboratory suites. I recommended moving six chemists and one suite, for eighteen months, from Brakenholm's next scale-up to qualification work for innovators in Japan, the United States and India. Bhavik Sanjelia, who heads custom synthesis and manages the Brakenholm account, argued that we were weakening our best relationship when it most needed us. He was right that Brakenholm would notice.

We told Brakenholm's procurement head in person what we were doing and why. That did not prevent a cost. Brakenholm placed the validation of a new fungicide intermediate with a second supplier, a molecule we had expected to bring about ₹60 crore a year from 2025–26. One of our new qualifications also failed on its impurity profile after ₹3.8 crore of development spending.

By 2025–26 four new innovator customers were buying ₹142 crore a year from us, and Brakenholm's share had fallen to 21 per cent, about ₹248 crore. I report that figure with its causes separated: roughly half the reduction came from our new customers and half from Brakenholm's own destocking, which we did not control. What I would do differently is measure concentration by share of gross margin from the start. On that basis Brakenholm still represents nearly 27 per cent, and the board deserved to see it earlier.