EN Every sale still needs four things done: find out where he's stuck, get him to trust you, prove the thing actually works for him, and ask for the money. Product-led growth doesn't delete those four — it moves them into the product, so onboarding does the asking, a free tier does the trusting, running his own data through it does the proving, and hitting a usage wall does the asking-for-money. So the only real test is this: with nobody holding his hand, can a stranger get one real success on his own? If he can't, PLG hasn't saved you anything — it has simply moved the cost from a salesperson's salary to your churn rate, and churn doesn't send you an invoice. What actually decides the model isn't price, it's how complicated the decision is: when the user is the person who pays and the first success takes minutes, self-serve carries the whole thing; when the user is not the one who pays and it takes integration, procurement or a security review, there's a human in the middle your product can never touch.
EN Freemium dies two ways: cut too deep and he never gets one real success before the wall; cut too loosely and he never hits the wall at all. The fix isn't tuning the ratio, it's changing the axis. Don't cut along "how advanced a feature feels" — he won't miss what he doesn't use. Cut along a unit that grows as he succeeds: people collaborating, things stored, messages sent, money handled. On that axis the better it works for him the more certainly he hits the wall, and by then he doesn't want to move house. Test it with one sentence: "you'll start paying when you ____." If the blank takes a good thing — more people, more volume, more revenue — the axis is right. If it takes a bad or irrelevant one (wanting to export, wanting dark mode), it's wrong. And the free period's job isn't showing off features, it's getting him to put things in: data, settings, habits, colleagues he invited himself. One red line goes with all of it — if his data can't be exported in one click, that isn't stickiness, it's a locked door.
EN Most people who don't convert didn't compare you against a competitor and pick them — they never started. So a trial has exactly one job: get him to finish one piece of real work in the first few days, however small. Three levers, strongest first. Kill the empty state: open onto something already running with sample data, not a "new project" button. Set the defaults to the path you want him on, so he isn't re-deciding at every step. And push one action, not five — a five-step tour is a zero-step tour. Two honest warnings come with this. Defaults work because attention, not laziness, is the scarce thing, which also means they can be used to keep people who never noticed they stayed; the line is whether he'd feel tricked if he found out later. And the famous activation thresholds you've read about are correlations dressed as recipes — someone who already intended to stay naturally does more of everything. Find your own candidate behaviour, then A/B test whether pushing it actually moves retention.
EN In subscription businesses the money is in year two and year three, so your price has to be able to grow with the customer. Net revenue retention — what the same cohort of existing customers pays this year divided by last year, counting upgrades, downgrades and churn but never new logos — is the one number that compounds by itself: at 100% you tread water without new sales, at 120% revenue grows even if you never sign anyone again, and at 90% a chunk of every new deal is just patching the bucket. The way above 100% isn't nagging people to upgrade, it's pricing against a unit that grows as they succeed — seats, usage, volume handled — the same axis your free tier should be cut on. Charge a flat price per company and their growth is none of your business. Two cautions: NRR built on price rises isn't the same animal as NRR built on usage, and high NRR with a shrinking customer count means concentration risk, not success.