What product-led growth actually means (and when it's the wrong model)

What is PLG?

Product-led growth (PLG) is a go-to-market model in which the product itself drives acquisition, conversion and expansion — rather than a sales team doing it. The term was popularised by the venture firm OpenView, and the products it describes are the familiar ones: sign up without talking to anyone, get value before paying, invite a colleague, upgrade when you hit a limit.

The phrase gets used loosely to mean “we care about product quality,” which is not what it means. Every company should care about product quality. PLG is a specific claim about who does the selling, and it has hard requirements.

The mechanics, not the vocabulary

Four things have to be true, and if any one is missing you have a free trial rather than a growth model:

1. Time-to-value measured in minutes. If someone can’t get a real result before deciding, there’s nothing for the product to sell with. This is the most common failure and the least often named — teams add self-serve signup to a product that takes two weeks to configure, and conclude PLG doesn’t work for them.

2. A natural expansion mechanic. Seats, usage, or a limit people hit while succeeding. Expansion has to correlate with the customer getting more value, not with them being squeezed. If your upgrade prompt fires when someone is frustrated rather than when they’re winning, you’ve built a toll booth.

3. A reason to invite someone else. Not a referral bonus — a reason inherent to the work. Products where the job is collaborative spread; products where it’s solitary don’t, and no amount of incentive design changes that.

4. Instrumentation good enough to see all of it. You need to know which actions in week one predict retention in week eight. Without that, you can’t improve any of the above, and you’re running a self-serve funnel on instinct.

Where PLG is the wrong model

This is the part that gets left out, and it costs companies a lot:

High-consideration, high-risk purchases. Anything where a wrong choice gets someone fired — core financial systems, regulated data, security infrastructure. Buyers want a salesperson, because they want someone accountable.

Products requiring integration before value. If you need three systems connected and a data migration before anything works, there is no self-serve path to value. You can bolt a free trial on; you cannot make the product do the selling.

Small numbers of very large customers. If ten accounts are most of your revenue, the economics of PLG don’t apply. You’re not optimising a funnel, you’re managing ten relationships.

Products the user doesn’t choose. Compliance tooling, HR systems, anything bought for people rather than by them. The user’s enthusiasm doesn’t drive the purchase, so product-led motion has nothing to push on.

The hybrid reality

Most successful PLG companies aren’t purely product-led, and the marketing around the term obscures that. The common shape is self-serve at the bottom, and a sales team that engages when an account crosses a usage threshold — the product generates the qualified lead, and a human closes the expansion.

That’s a good model. It’s just not the pure version anyone describes at conferences, and pretending otherwise leads teams to fire a sales function they needed.

What it means for product work

If you’re genuinely running PLG, three things change about the job:

• Onboarding is a product surface, not a project. It’s where the selling happens, and it needs an owner permanently rather than a redesign every two years.
• Your activation metric matters more than your acquisition metric. More signups against an unchanged activation rate is just a bigger leaky bucket — which is why activation, not signups, tends to be the right north star input.
• Pricing is product. Where the limits sit determines both revenue and the upgrade experience, and it can’t be owned by a separate team that ships quarterly.

Before adopting the label

The useful question isn’t “should we be product-led.” It’s: can a stranger get a real result from our product, alone, in under fifteen minutes?

If no, PLG is a project to make that true — usually a large one — not a go-to-market strategy you can adopt this quarter. And it’s worth being honest about whether the answer is no because of a fixable onboarding problem or because of what the product fundamentally is. Those need very different plans, and confusing them is how a year gets spent.

That’s a diagnosis worth getting right before committing a roadmap to it — the kind of bounded question a short consulting engagement answers better than a quarter of internal debate.