PLG basics. What is it?
Product-led growth, reduced to three things
Product-led growth gets said in board meetings with roughly the same reverence, and the same vagueness, as “AI strategy”. Everyone wants it. Ask for a definition and you’ll usually get a slide with a flywheel on it.
At CaptionHub we’re exploring PLG as a way to reduce friction for our enterprise clients rather than an entire GTM Motion in and of itself.
On our weekly management standup my CTO rightly asked me to clarify what I mean by PLG. My answer was: simply think about whether these three things are satisfied:
- Can your user find you and sign-up to your product by themselves and without a sales person? This includes finding you as well as the process. Of course, there is a binary yes/no answer to this, but the real answer is, how well they can find us and sign-up: how easy, how quickly, how much delight in the process and so on.
- Can your user get what they came for quickly. Can they walk out of the showroom with the thing they came for, solved, without any other human interaction?
- Once they are in and getting something back in terms of value, can they pay for it there and then and how easily? Once they’re paying for the thing, can they pay for more of it, and can they pay for other related things in your product set.
There’s more to PLG, such as bringing in sales teams when product activity indicators are triggered, but if you can prove that initial set of satisfied arguments, then you have a PLG motion as a foundation in place.
Now, credit to my CTO, he knows what PLG is. He knows the above. The question forced alignment on a set of definitions that open the conversation to why we want PLG. Notice, we excluded ‘spending millions of budget / reallocating capital to a massive marketing campaign’. This doesn’t have to include the awareness and acquisition piece to have PLG in place, though that is critically important to its success.
From someone who has sold software for a living since 2010 and has seen the GTM playbook rewritten more than once in the last few years, PLG is a go-to-market model where the product does work a sales team used to do: getting found, convincing the buyer, closing and expanding. It comes down to three things, and they're so foundational to PLG that I'll repeat them again with a little more detail:
1. Self-service entry
Customers can find you and sign up on their own. There’s no “book a demo” wall and no form asking for company size and budget before they’ve seen a screen.
Most B2B companies still don’t do this, because their front door was built to qualify leads for sales. PLG reverses the order. People get in first, and you qualify them later on what they actually do.
2. Self-service time to first value
Once they’re in, they can get what they came for quickly and without speaking to anyone. That first useful result should take minutes, and nobody should need an onboarding call to reach it.
Most attempts fall over here. If the sign-up is self-service but the product needs a 45-minute implementation call before it does anything, you’ve built a lead form with extra steps.
3. Self-service billing and adoption
Customers can pay, and pay more, without negotiating. Expansion goes two ways. Usage grows from the bottom up as they add users, minutes or credits, and it spreads sideways as they pick up adjacent features like pro tiers and bolt-ons.
The product has to drive this itself, with a prompt that arrives when the customer needs the next thing. An upsell email at renewal is too late.
What PLG isn’t
Plenty of PLG companies have sales teams. They point them at accounts the product has already warmed up. And a freemium tier bolted onto an enterprise product, then left alone, doesn’t count.
Why it matters more now
AI has made software far cheaper to build. Earlier this year I built and shipped an internal tool in a Saturday morning with no coding knowledge, which is great, and a bit unnerving if selling software is your job. We have people in our company building products in a day or a week. The bottleneck is GTM. Buyers now have more alternatives and can try three in an afternoon. They’ll pick whichever one gets them to value before anyone asks for a call.
AI products also tend to be priced on usage (tokens, credits, minutes), which fits the third pillar well.
This piece is the introduction. Next I’ll take each pillar in turn: what good looks like, how to measure it and where it breaks.