Target SelectionNarrow moat

The first and most underappreciated skill in Broadcom's acquisition machine is knowing what to buy. Under Hock Tan, Broadcom has pursued a remarkably consistent template: acquire established franchises with sticky, entrenched customers and durable market positions — not speculative growth stories or turnaround hopes, but mature businesses whose customers are, in one way or another, stuck. From LSI and Brocade to CA, Symantec's enterprise unit, and VMware¹, the targets share a common feature: high switching costs and a captive base that can be relied upon to keep paying.

This discipline is what separates Broadcom's M&A from the value-destroying acquisition sprees that litter corporate history. Most serial acquirers overpay for growth or buy businesses they do not understand; Broadcom buys a specific kind of asset it understands intimately — a franchise with pricing power waiting to be exercised and costs waiting to be cut — and applies the same playbook each time. Buying only businesses whose customers cannot easily leave is the foundation on which everything else rests.

The limitation is that targets fitting the template are finite, and each deal must be larger than the last to move a company that has grown enormous. The universe of mature, sticky, acquirable franchises of sufficient size is not infinite, and as Broadcom grows it must reach into new domains — from chips into infrastructure software — to find deals big enough to matter, stretching the template into less familiar territory. Target selection has been superb, but the supply of ideal targets is a real constraint on a model that must keep feeding.