Should we buy the company that threatens us?

The question behind the question

When a young company threatens an established one, buying it is the obvious response, and often the right one. But an acquisition is two decisions at once: whether the target is worth owning, and whether you can own it without destroying what made it valuable. Most acquisitions that fail do so on the second question.

The base rate is sobering. Sara Moeller, Frederik Schlingemann and René Stulz found that shareholders of acquiring companies lost about $240 billion in the merger wave of 1998 to 2001, mostly in a small number of very large deals. Studies of thousands of acquisitions find that, on average, the seller’s shareholders capture most of the value a deal creates, and that the buyer’s often lose. Richard Thaler’s account of the winner’s curse explains part of the reason: in a competitive auction for an uncertain asset, the winner is often the bidder who overestimated its value most.

When buying worked

Facebook bought Instagram in 2012 for about $1 billion, when the photo app had thirteen employees and no revenue, and WhatsApp in 2014. Emails later released in litigation show that Mark Zuckerberg wanted to buy Instagram before it could grow into a threat. Both were bought young, before their prices reflected their eventual size, and both were allowed to keep their products and much of their character for years. Microsoft bought GitHub in 2018 for $7.5 billion and, to the surprise of many developers who feared it would close or exploit the platform, kept it open to all and invested in it. Investors who bought Iridium’s satellite network out of bankruptcy in 2000, at a tiny fraction of its cost, and served niche markets, made a good acquisition by paying a price set by someone else’s failure.

When it did not

Nokia agreed in 2007 to buy the digital map maker Navteq for about $8.1 billion, betting that maps would be a scarce asset on phones. Within two years Google made navigation free, and the asset stopped being scarce; Nokia sold the mapping business in 2015 for €2.8 billion. Dropbox bought Mailbox, a hit email app built on top of Gmail, in 2013, and closed it in 2015, losing what it had bought. Autodesk invested in Ted Nelson’s Xanadu hypertext project in 1988 to turn it into a commercial product, and withdrew in 1992 without one; the simpler web soon made the vision moot. Oracle’s acquisitions of PeopleSoft and Siebel consolidated the market for installed enterprise software at the moment that market began to move to the cloud, which the record judges mixed.

A diagnosis to start from

The diagnosis starts with why the target is succeeding. If its growth depends on something the buyer has, such as distribution or capital, the acquisition can accelerate it. If its growth depends on something the buyer lacks and might damage, such as a culture, a community of users who distrust the buyer, or independence from the buyer’s other products, the acquisition can destroy it. The second part of the diagnosis is whether the asset’s value depends on conditions likely to change. Navteq’s maps were valuable because maps were scarce; the deal was a bet that they would stay scarce.

The guiding policy that the successes share is to buy early, pay for growth that the buyer can protect, and commit publicly to what will not change. The coherent actions include leaving the acquired product’s team and brand in place, giving it resources it lacked, and setting a review after two years with a willingness to admit failure rather than escalating commitment, a danger described in Sunk Costs and the Escalation of Commitment.

The strongest counter-case

The strongest objection comes from competition law and from society rather than from shareholders. Buying young rivals before they can compete may be good for the buyer and bad for everyone else. The US Federal Trade Commission’s case against Meta argued that the Instagram and WhatsApp acquisitions eliminated competitors; a court ruled for Meta in 2025, but the argument has changed how regulators review acquisitions of young companies. A strategy that depends on buying every threat also invites the scrutiny that eventually prevents it.

There is a quieter counter-case too. Facebook’s acquisitions are famous because they worked. Many acquisitions of promising young companies by large ones failed quietly, and their absence from the record flatters the strategy of buying young, a problem the site’s idea Silent Evidence describes.

What the wider research says

Daniel Kahneman and Dan Lovallo argued in 1993 that managers make bold forecasts and evaluate each deal in isolation, ignoring the base rate of similar deals; acquisitions are a textbook case. Studies commissioned by investment banks and consultancies, including Michael Mauboussin’s work at Credit Suisse, find that deals tend to create more value when the buyer pays in cash, when the premium is modest and when the acquisition is related to the buyer’s existing business. The record adds a point those studies cannot easily measure: what matters most may be whether the buyer protects the reason the target was growing.