How to choose: the decision spine

The choice between a product-led and a sales-led model is not a vibe, and it is certainly not a matter of which one is currently fashionable. Decide it on four properties of your actual business. Product complexity: can a buyer understand the value alone, or do they need a human to translate it. Average contract value: are you selling something at a price point that can sustain self-serve economics, or one that demands and can pay for white-glove attention. Buyer expectation: does your buyer expect to try before they talk, or expect to talk before they try. And time-to-value: how quickly a new user reaches the point where the product has obviously paid for itself.

Product-led growth runs on self-serve adoption and low cost to acquire, and it suits products where a buyer can reach value alone. Sales-led growth carries higher contract values and longer cycles, and it suits buyers who expect a human in the loop before they commit. The fit between these four properties and the motion is the decision. Get the fit right and the motion feels effortless; force the wrong motion onto the product and you fight your own buyers every day.

Let CAC payback set your spending ceiling. The model also sets your cost-of-acquisition discipline, and this is where the wrong lens turns into a cash trap. The median B2B SaaS CAC payback ran around eighteen months in 2024, up from fourteen the year before, but the median hides the point. Payback varies enormously by deal size: a low-ACV self-serve tool (under 15,000 USD) commonly recovers its acquisition cost in eight to twelve months, mid-market sits around fourteen to eighteen, and a six-figure enterprise deal commonly takes eighteen to twenty-four. The consequence is that the same paid-search spend is efficient under one model and reckless under another. A long payback is perfectly sensible against a large, durable enterprise contract and a cash trap against a low-ACV self-serve product where the maths has no time to work. You do not judge a tactic in isolation; you judge it against the payback your model permits.

Expect hybrid, but never run all three at once. Most B2B SaaS does not stay purely product-led or purely sales-led as it scales; the majority land in a hybrid that combines self-serve adoption with a sales motion for larger deals, and hybrid models routinely out-perform pure approaches on retention. So hybrid is a destination, and that is fine. The discipline that matters is sequencing, not purity. Hybrid is not permission to run product-led, sales-led, and expansion motions all at full tilt simultaneously while you still have one pair of hands. It is the knowledge of which model you are in now, this quarter, with the others sequenced behind it. Drift into running all three at once and you are back to half-optimising everything, which is exactly the unfocused state the whole exercise exists to escape.