#1577 2004 · Bharti Airtel · Telecommunications
Paying Ericsson to build and run the network—flipping telecom capex into a variable spend
the problem
India's mobile boom made the network the bottleneck; Bharti couldn't fund towers and equipment fast enough to keep up
background
India's mobile market was growing explosively in the early 2000s, and the only way to keep up was to lay enormous physical infrastructure — towers, base stations, backbone, and the IT systems that run them — all of which must be paid for years before the customers who justify it arrive. For a challenger like Bharti, funding that build-out with fixed capital was both impossibly expensive and slow: every rupee tied up in equipment was a rupee it could not spend on marketing, brand, pricing or the customer base. Its balance sheet could not finance a build fast enough to match demand, and a network bought outright carried the risk of stranding capital the moment growth softened.
Bharti's answer in 2004 was to refuse to carry the fixed asset at all. In a deal described at the time as the first such arrangement in the world, it signed a roughly $400 million managed-services contract with Sweden's Ericsson to provide, run and maintain its cellular network in thirteen circles, paying for managed capacity rather than owning the equipment. It applied the same logic to its IT estate (IBM) and its towers, keeping for itself only what it could uniquely own: brand, distribution, customers, pricing and marketing.
what everyone would do
Raise enough debt and equity to buy and run the network yourself, then own the infrastructure as a cost advantage once it is built out and amortised.
what they saw
You need not own the asset to capture the customer: pay a vendor to build and run the network, and the fixed capital throttling your rollout becomes a variable fee that scales with subscribers
the move
Bharti outsourced the physical network to its equipment vendors instead of buying and running it themselves. Under the 2004 deal Ericsson provided, managed and maintained the equipment and guaranteed quality across thirteen of Bharti's mobile circles for three years, with around 250 Bharti network-planning staff absorbed into Ericsson; Bharti funded the arrangement partly with a $300 million ABN Amro loan at roughly 4% including hedging costs, and later extended the model with a $1 billion Ericsson expansion covering more than 70% of the network plus companion deals with Nokia for the rest. The operator focused its own resources on product innovation, value-added services, marketing, branding and pricing, effectively paying a demand-linked fee for managed network capacity instead of sinking fixed capital into owned, idle infrastructure.
why it works
A mobile operator's economics are dominated by the network, which must be paid for years before the traffic that justifies it appears; a challenger that buys it outright spends limited capital on hardware and slows every other driver of growth. By contracting Ericsson to provide, manage and maintain the infrastructure and paying for managed capacity instead of ownership, Bharti moved the capital burden and the build risk to a supplier that already ran networks at scale, kept its own resources on brand, pricing, marketing and customer acquisition, and therefore could turn subscribers into cash faster than a cash-strapped, asset-laden rival ever could. Because the fee scaled with usage rather than being sunk up front, growth did not force the balance sheet to inflate in lockstep — the very asymmetry that let a capital-limited challenger ride an exploding market and outpace the incumbents that insisted on owning their networks.
the payoff
Freed capital and a faster rollout let Airtel ride India's boom; it is now the world's second-largest telco by mobile subscriptions.
where it breaks
The model works only while demand keeps growing, because renting managed capacity is a permanent, usage-linked cost rather than an asset you eventually own: if growth stalls or prices fall, you have no owned infrastructure to lean on and your margin is at the mercy of a vendor's pricing. You also trade direct control for a dependency — outages, upgrade timing and cost escalation sit with the supplier, so the operator's reputation rides on a third party. And the economics of outsourcing are only good while hardware is a commodity: the moment equipment supply or a vendor relationship becomes too concentrated, or a competitor owns and amortises the network over decades-old invested capital, the rented-usage operator can find itself paying more than an owner does.
what came after
The deal was written up at the time as the first managed-network outsourcing of its kind, and lightreading noted within two years that outsourcing network management to an equipment supplier had 'become de rigueur for mobile carriers, particularly in emerging markets where operators prefer to focus on managing their rapidly expanding customer bases.' It legitimised the asset-light operator — renting managed capacity from vendors who carry the capital and build risk — and spread across emerging-market telecoms as a template for companies growing into demand too fast to fund their own infrastructure; the same outsourcing logic later extended to towers and to IT.
references
- [1]Bharti, Ericsson in $400m network outsourcing dealBusiness Standard, 2004business-standard.com
- [2]Ericsson in $1B Bharti ExpansionLight Reading, 2006lightreading.com
- [3]Bharti Airtel: Rise of a GiantTeleGeography, 2026resources.telegeography.com