Each decision gave something up, and the record says what

Pricing new films by the hours they would earn, not the star on the poster

When I became chief strategy officer in April 2021, Ilavenil was spending ₹310 crore a year on satellite and streaming rights to new Tamil and Telugu films, close to two-fifths of its content budget. Prices for films led by the biggest stars had risen by about 60 per cent in three years as broadcasters competed for festival premieres. Nobody in the company could say what a film earned after its premiere week, because premiere ratings were reported to the board and repeat telecasts were not.

My team built a model from five years of our own ratings and streaming data. For each film it estimated the viewing hours it would earn over ten years: premiere, repeats on our film channels and streams on Ilavenil Play. Twelve star-led films bought in 2019–20 had taken 41 per cent of the spend and earned 19 per cent of the hours, because their repeat viewing faded quickly. Mid-budget family films and older library titles earned about three times as many hours per rupee.

I proposed a price ceiling for every new film, set by its expected hours, and a separate budget for library purchases. Arulmozhi Senthilnathan, our head of Tamil programming, disagreed. Festival premieres, she argued, are when families decide which channel is theirs, and a channel that loses them loses more than the model can see. The managing director accepted the ceiling and asked for a review after a year.

We lost the Pongal 2022 premiere of a major star's film to a rival, and the Sankranti premiere in Telugu the same week. Our Tamil prime-time share fell from 21 to 16 per cent that week before recovering within a month. We spent ₹96 crore on 140 library titles and moved two bids to revenue-sharing terms.

By 2023–24 film-rights spending had fallen to ₹246 crore while hours watched from films rose 9 per cent, and the cost of a thousand viewing hours from films fell by 27 per cent. Every film above ₹15 crore now reaches our content committee with an hours case beside the price.

I would do one thing differently. Arulmozhi was right that premieres carry value the model cannot measure. I would have kept one festival premiere a year in each language as a deliberate brand cost, priced openly, rather than letting the ceiling give away both in the same month.

Buying a Marathi broadcaster instead of a stake in a Hindi network

In early 2022 two acquisitions were put to us within a month. The first was 51 per cent of a cluster of three Hindi entertainment channels, fifth in its genre and losing about ₹90 crore a year, for ₹1,380 crore. The second was Ghatmatha Entertainment, a family-owned Pune broadcaster with two Marathi channels, 18 per cent of Marathi entertainment viewing, revenue of ₹128 crore and a library it had never exploited, for ₹465 crore outright.

Our managing director preferred the Hindi stake. It would make Ilavenil a national broadcaster, and our head of advertising sales argued that national advertisers would pay a premium for one buy across Hindi and the South. I recommended Ghatmatha. A fifth-placed Hindi cluster would need years of losses to become a channel distributors had to carry; Ghatmatha already was one, in a state where most entertainment viewing is in the local language. Because I grew up in a Marathi-speaking home, the managing director asked, fairly, whether I was the right person to make the case. I asked for an outside firm to prepare the valuation and for my team to write the full case for the Hindi stake too.

The board approved Ghatmatha in August 2022; the purchase completed in October. We gave up the story of becoming a national network, and with it the premium the sales team believed it could earn; two national advertisers told us that they would not change how they bought.

We kept the Ghatmatha brand, brought its channels into our negotiations with distribution platforms, put its library on Ilavenil Play and raised its fiction budget by ₹70 crore a year. In 2023 we moved its commissioning approvals to Chennai. Two of its head writers left, and I reversed the change after nine months.

Ghatmatha's revenue rose from ₹128 crore in 2021–22 to ₹236 crore in 2025–26, its EBITDA from ₹19 crore to ₹58 crore and its share of Marathi entertainment viewing from 18 to 24 per cent. The Hindi cluster was later sold to another buyer at a lower valuation.

I would do two things differently. I would have left commissioning in Pune from the first day, because the writers were most of what we bought. And I would have offered the founding family an earn-out tied to viewing share, to keep their judgement in the business for two years.

Keeping the tenth year in a deal with a global streaming service

By early 2023 Ilavenil Play had 1.4 million paying subscriptions, and its original series cost us ₹160 crore a year. A global subscription video service offered to co-commission six Tamil and Telugu series a year for three years. It would pay ₹210 crore in return for exclusive worldwide rights to the series in perpetuity.

The offer would have covered almost all the production cost. Our CFO wanted it as offered, because it took about ₹70 crore a year of cost off our books. Our head of Ilavenil Play opposed any version of it: he argued that originals were why people subscribed to our app, and that giving our best series to a larger competitor would slow our growth. I proposed a third course: a two-year exclusive worldwide streaming window for the platform, after which all rights would return to us, with television rights in India for us after twelve months.

The platform accepted the structure but lowered its payment to ₹168 crore over three years. We gave up ₹42 crore of fees for the rights after year two, and we accepted the slower growth our head of Ilavenil Play had predicted: the series launched elsewhere, and the app added subscribers more slowly in 2023–24 than our plan assumed.

The agreement was signed in August 2023. Fourteen of the eighteen series had been delivered by June 2026. Platform fees covered 78 per cent of their production cost. The first five series premiered in our afternoon television slots in 2024–25, drawing audiences about 30 per cent larger than the films usually placed there, and have since returned to us in full and now stream on Ilavenil Play. The app reached 2.3 million paying subscriptions by March 2026, helped largely by bundles with broadband and mobile operators.

What changed was how we price partnerships. Every licence we sign now states which rights return, when, and what we expect them to earn in the years afterwards, and the board sees that estimate beside the fee.

I would renegotiate two things I conceded. I gave the platform the soundtrack rights during its window, and the music income from two successful series went with them. And I did not ask for viewing data from the platform's window, so when the series came home our partner knew more about their audiences than we did.

Trading exit charges for new tenancies as mobile operators merged

Between late 2016 and 2017, a new mobile operator's free voice calls pushed the industry into consolidation. At Varanya Towers, two of our six tenant operators announced a merger and a third began withdrawing from several circles. Together they gave notice on 6,900 of our 31,200 tenancies. Our master service agreements entitled us to exit charges, which we calculated at ₹540 crore, and our lenders were watching a tenancy ratio that could fall from 1.9 towards 1.5.

Our CFO and the director nominated by our larger investor wanted the agreements enforced in full, through arbitration if necessary, because the cash would protect our debt covenants. I thought the claim was strong on paper and weak in practice. The merging operators would become our largest customer, they were planning their 4G networks, and an arbitration would take years while they decided whose towers to load.

I proposed a settlement: ₹310 crore of exit charges, paid over eighteen months, and a commitment from the merged operator to place 4,100 new 4G tenancies on our existing towers, each with a ten-year lock-in. Prasanna Deodhar, our managing director, supported it. The investor director agreed only after we modelled the settlement against an arbitration lasting four years. We gave up ₹230 crore of contractual claims, and our lenders reset two covenants at a cost of 35 basis points on our loans.

The settlement was signed in November 2017. Our tenancy ratio bottomed at 1.58 rather than the 1.5 we had feared, and it was rising as the new tenancies arrived in early 2018, around the time I left. Varanya then rewrote its agreements with every operator so that exit charges and upgrade commitments were negotiated together rather than in separate clauses.

I would do two things differently. I would have modelled consolidation in 2015, when it was already visible in operators' accounts, and renegotiated our agreements before any customer needed to leave. And I would have briefed our lenders when the merger was announced, not when the notices arrived.