Product led growth tactics: What to ship first

Open any product-led growth tactics guide and the first thing it does is list every tactic that has ever worked for anyone: freemium, viral loops, PQL scoring, in-app nudges, expansion pricing. Useful enough in theory, but not much help when you're trying to decide what to ship this week. This guide does the opposite. It ranks product-led growth tactics by what to implement first, what to instrument when you do, and how to tell whether it actually moved revenue.

The ordering problem is the whole game. A seed-stage team that ships viral loops before fixing activation is collecting referrals for a product nobody sticks with. A growth-stage team that skips PQL routing and hands every signup to sales is burning rep capacity on users who were never going to buy. The wrong tactic at the wrong stage just creates noise.

Which product-led growth tactics to ship first by stage

The stage filter that keeps you from doing too much

Early teams, usually pre-product-market fit and with fewer than a few hundred active users, need exactly two things: one activation event they can instrument and one onboarding path that gets users there reliably. Everything else is premature. Viral loops require a product worth sharing. Expansion revenue requires users who stayed. PQL routing requires enough volume to set a threshold that means something. None of that helps until activation is real.

Later teams can layer more on top. Once activation is instrumented and cohort retention is visible, network effects, PQL handoffs, and expansion loops become useful instead of distracting. PostHog's PLG overview says it plainly: the self-serve motion has to work before the sales-assist layer on top of it earns its keep.

A simple prioritization matrix

Score every candidate tactic on three axes: effort to implement, expected impact on activation or revenue, and measurement complexity. High-effort, hard-to-measure tactics, like viral loops and expansion pricing, belong later. Low-effort, directly measurable tactics, like onboarding simplification and activation event instrumentation, belong first.

One team building a B2B collaboration tool skipped referral mechanics until month six, after they had instrumented activation and confirmed retention cohorts were healthy. The referral loop they eventually shipped landed on a product people already used daily. The same loop at month two would have pushed people into a broken onboarding experience.

Choose freemium, free trial, or a hybrid PLG motion

The motion choice is a bet on where your product creates value and how fast users feel it.

Freemium only works when usage naturally expands

Freemium is a retention-and-expansion bet, not a default acquisition strategy. It works when free usage creates a real habit: daily active use, data accumulation, or collaboration, so users eventually hit a limit they want to pay through. The failure case is a product where free usage is complete. The user gets everything they need on the free tier, never collaborates, never accumulates enough data to feel the ceiling, and never converts. Freemium for a product with a single-player, bounded use case is a permanent discount, not a growth motion.

Free trial works when the aha moment is fast

A time-boxed trial beats freemium when the product has one clear activation milestone and the buyer can feel value within days. If a user can reach the aha moment in 48 hours, a 14-day trial creates urgency without giving away the product indefinitely. The First Round Review's PLG deep-dive notes that trial length should match time to value. A trial that ends before the user reaches activation is just churn with extra steps.

Hybrid motion when the product has both paths

A hybrid motion lets users self-serve into activation, then routes higher-intent accounts to sales once PQL signals appear. The product handles the top of the funnel; sales handles accounts that show buying intent. This only works when the self-serve path is genuinely complete. If users need sales to get started, the "hybrid" is just sales-led with a free trial bolted on.

Define activation without guessing

Pick one aha moment the product can actually prove

Activation is a specific user action the product can observe, not "the user understood the value" or "the user had a good first session." Guessing here poisons everything downstream. If you instrument the wrong event, every cohort analysis after that is measuring the wrong thing, and every product change you make in response is optimizing for a proxy.

The test is simple: can your analytics stack return a list of user IDs who completed this action in the last 30 days? If not, the event isn't instrumented. If yes, can you correlate that list with 30-day retention and revenue? If the correlation is weak, it's the wrong event.

Use an event schema the team can query later

A useful activation event schema looks like this:

• Event name: `demo_activated` / `project_created` / `report_shared` — whatever the action is, named once and used consistently
• Properties: `user_id`, `account_id`, `timestamp`, `plan_tier`, `days_since_signup`
• Account identifiers: both user-level and account-level, so you can cohort by company, not just individual

The account-level identifier is the part teams skip and regret. If you can only query by user, you can't answer "did this account activate?" That's the question that matters for B2B PQL scoring. PostHog's event instrumentation guidance covers the cohort-query requirement directly: an event that can't support a cohort query isn't an activation metric, it's a counter.

Instrument the events that show onboarding, retention, and revenue

Onboarding events that show where users stop

Map the onboarding path to a sequence of observable events: `signup_completed`, `invite_sent`, `key_action_completed`, `onboarding_finished`. The gap between `signup_completed` and `key_action_completed` is where most users fall out. That's the first place to intervene.

Retention events that prove the product got habitual

Retention is repeated behavior, not repeated logins. Pageviews are noise. The event that proves retention is the one that reflects the core product action: a report run, a message sent, a file exported, a workflow triggered. Track it weekly per account. If the same accounts are running that action week after week, you have retention. If different accounts run it each week, you have exploration, not habit.

Revenue events that tie usage to expansion

The same event stack can surface upgrade signals: `feature_limit_hit`, `seat_invite_blocked`, `export_quota_reached`. These are the moments a user wants to pay. Instrument them separately from retention events. They're not the same signal. A user who hits a feature limit and doesn't upgrade has a conversion problem. A user who hits a limit and upgrades is an expansion trigger. Mixing them with pageview traffic is how vanity metrics crowd out the real ones.

Set PQL thresholds and sales handoff rules that do not flood the team

What makes a PQL worth a sales follow-up

A PQL threshold that uses only behavioral signals, like number of logins or actions taken, without firmographic fit, like company size, industry, or role, produces junk leads. A solo user at a five-person company who has run 40 actions is not the same as a VP at a 200-person company who has run 20. The threshold needs both sides: behavior that shows the user found value, and fit that shows the account can buy.

A working example: PQL = activated, meaning they completed the core action within 7 days of signup, and the account has 3+ users and 50+ employees. Each of those criteria filters out a different kind of noise.

How to route the right accounts without breaking self-serve

The handoff rule has to be specific enough that sales knows exactly what to do and the product doesn't feel like it got hijacked. One clean version: when an account hits PQL status, sales gets a notification with the account's activation date, actions taken, and seat count. Sales sends one personalized outreach. If there is no response in five days, the account stays in self-serve. The product never changes for the user. No forced demo calls. No paywalls triggered by PQL status. The First Round Review's self-serve strategies discussion makes the same point: the self-serve motion has to stay intact even when sales is in the loop.

Use onboarding, paywalls, and in-product changes to improve time to value

Small onboarding fixes that usually matter first

The highest-leverage onboarding changes are the ones that shorten the path to the activation event. That usually means removing steps, not adding them. A five-step onboarding wizard where step three is optional is a three-step onboarding wizard with a bug. Find the step where users drop and ask whether it is required to reach activation. If it is not, cut it.

Paywalls that create focus instead of friction

A paywall placed before activation blocks learning. A paywall placed after activation, at the moment a user wants to do more, creates focus. The distinction matters. Gating a feature the user has not tried yet teaches them nothing about the product. Gating a feature they just discovered they need tells them exactly what they are paying for. The paywall is a PLG tool when it sits at the expansion moment, not at the entry point.

In-product changes that move the metric fastest

The product changes that move time to value fastest are usually the simplest: a clearer default state, a shorter path to the first meaningful action, or a prompt that tells the user what to do next. One common example is the empty-state problem. A product that shows a blank screen on first login forces the user to figure out where to start. Replacing that blank screen with a guided first action or a sample dataset can cut time to activation by days without touching the core product. Measure it with a before-and-after release cohort: compare activation rates for users who signed up the week before the change versus the week after.

Test whether a product-led growth tactic actually moved revenue

What to compare before you trust the lift

Signups went up after you shipped the tactic. That tells you almost nothing. Compare activation rates, not signup counts. Then compare retention cohorts: users who went through the new onboarding versus users who went through the old one. If the activation rate improved and the 30-day retention cohort is healthier, the tactic worked. If only signups went up, you changed acquisition, not product-led growth.

The cleanest way to avoid vanity PLG metrics

Vanity metrics are the ones that go up whether the product is working or not: total signups, total pageviews, total sessions. The metrics that reflect actual product movement are activation rate, week-4 retention by cohort, and expansion revenue per account. Run a treatment-versus-control cohort, users exposed to the change versus a holdout, and measure those three numbers. If the treatment cohort shows higher activation and better retention, the tactic moved the metric. If only signups moved, the tactic moved the top of the funnel, which is a marketing win, not a PLG win. YC's growth team guidance makes the same distinction: growth metrics tied to revenue are the ones worth optimizing; everything else is a proxy that can mislead.

FAQ

Q: Which product-led growth tactics should a B2B SaaS team prioritize first if it wants the fastest lift in activation and revenue?

Instrument one activation event and fix the onboarding path to it before anything else. Viral loops, PQL routing, and expansion pricing all depend on activation being real. A team that skips this step and ships viral mechanics is collecting referrals for a product nobody sticks with. The stage filter is simple: if you can't answer "what percentage of last month's signups activated?" you're not ready for the next tactic.

Q: How do you define the activation event for your product without guessing?

Pick the one user action that, when completed, correlates most strongly with 30-day retention and revenue. It has to be an observable event, something your analytics stack can return a list of user IDs for. If the correlation between the candidate event and retention is weak, it's the wrong event. Instrument it with both user-level and account-level identifiers so you can cohort by company, not just individual users.

Q: What events and user actions should you instrument to measure PLG properly?

Three separate buckets: onboarding events (`signup`, `key_action_completed`, `onboarding_finished`), retention events, meaning the core product action repeated weekly, not logins or pageviews, and revenue events like `feature_limit_hit`, `seat_invite_blocked`, and `upgrade_triggered`. Each bucket answers a different question. Mixing them produces dashboards that look busy and tell you nothing.

Q: When should a company use freemium, free trial, or a hybrid product-led sales motion?

Freemium when usage naturally expands and the free tier creates a real habit. Free trial when the aha moment is fast and a time-boxed window creates urgency without giving away the product indefinitely. Hybrid when the self-serve path is complete and sales only needs to engage accounts that show buying intent, not accounts that need help getting started. The wrong choice is freemium for a product where free usage is complete, or a trial shorter than time to value.

Q: How do you identify product-qualified leads and route them to sales without adding noise?

Use both behavioral and firmographic signals. Behavior alone, like logins and actions, produces junk leads. Fit alone, like company size and role, produces cold outreach. A PQL threshold that requires activation plus fit, for example, completed the core action within 7 days and company has 50+ employees and 3+ users on the account, filters out most of the noise. The handoff rule has to keep the self-serve experience intact: sales gets a notification, sends one outreach, and the product never changes for the user regardless of whether they respond.

Conclusion

Pick one tactic from this list, instrument the one event path it depends on, and run a cohort comparison this week. The ordering problem from the intro is still the whole game. Doing the right things first beats doing more things at once. If activation is not instrumented, start there. If it is, look at the onboarding drop-off before you touch anything else. The fastest PLG wins come from fixing what is already broken, not from adding what is theoretically possible.