Most founders don’t fail because they can’t get users, they fail because they scale the wrong behaviour. If you’re running a product-led motion, you need evidence that people get value fast, come back without being chased and expand without your team doing heroics.
Before you pour fuel on acquisition, cross-reference Go-To-Market Strategy for Founders: The Complete Playbook and make sure your numbers can carry the weight.
In this article, we’re going to discuss how to:
- Choose the few metrics that prove your product sells itself
- Instrument and review data you can gather in a few hours
- Run small tests to improve activation, retention and expansion before scaling spend
Define The Concept In Practical Terms
Product-led growth is simple: your product is the main salesperson. The practical job is proving, with data, that users can reach value quickly and repeatedly, then convert or expand with minimal human touch.
Here’s the founder-first framing I use: you’re not tracking vanity, you’re tracking ‘momentum’. Momentum is the chain from first use to real value to habit to paid expansion.
- Activation: Users complete the actions that predict they’ll stick.
- Time to value: How fast they get a tangible win.
- Retention: Do they return and keep doing the valuable thing.
- Expansion: Do accounts grow without constant sales pressure.
- Referral loops: Do happy users bring in new users.
If you can’t describe your activation event in one sentence, your dashboard will always lie to you.
The Only Dashboard You Need Before Scale
Most early PLG dashboards are a junk drawer: 40 charts, no decisions. Instead, run a tight weekly scorecard with 9 to 12 numbers max, split into three layers.
Layer 1: Behaviour (What Users Do)
These are event-based and should be measurable even if your revenue is still messy.
- Activation rate = Activated users / New sign-ups.
- Median time to value in minutes, hours or days.
- Weekly active users with a clear definition tied to value, not logins.
- Core action frequency per active user (eg, ‘reports generated per week’).
Layer 2: Commercial (What The Product Earns)
These numbers keep you honest about pricing, packaging and whether self-serve actually works.
- Free to paid conversion within 7, 14 and 30 days.
- Average revenue per account and gross margin.
- Expansion rate and net revenue retention (NRR) if you have enough data.
Layer 3: Efficiency (What It Costs You)
PLG is meant to be efficient. If it isn’t, you don’t have PLG, you have a support-heavy freemium model.
- Customer acquisition cost (CAC) by channel, even if it’s rough.
- Payback period in months.
- Support load per active account (tickets, chats, human hours).
One rule: if a metric doesn’t change a decision this week, it doesn’t belong on the scorecard.
What You Can Instrument In A Few Hours (Internal First, Then Public)
You don’t need a six-week data project to start measuring. You need a clean event map and a basic cohort view.
Internal signals to gather today:
- Top 5 user journeys: From sign-up to first ‘win’.
- Event taxonomy: 10 to 20 events max that describe value delivery.
- Activation definition: The 1 to 3 actions that predict 30-day retention.
- Friction list: Where users drop off in onboarding and why.
Public or semi-public signals to sanity check: reviews, job posts mentioning your category, competitor pricing pages and feature gaps. If you’re still unsure whether the idea itself is worth pushing, refer back to Business Ideas: The Full Guide to Finding, Testing and Choosing the Right Idea and pressure-test demand before you polish funnels.
When you set up analytics and experiments, do it in a way that respects privacy and consent. The basics are in the UK Information Commissioner’s Office guidance on cookies and similar technologies, especially if you’re using behavioural tracking.
Activation And Time To Value: Your First Reality Check
Activation is the moment a user does something that correlates with future value. Time to value is how quickly they get there. Together, they tell you if your onboarding is a ramp or a wall.
Pick An Activation Event That Predicts Retention
Activation should be specific and testable. ‘Completed onboarding’ is usually a weak proxy. Better is an action that looks like real work.
Examples of strong activation events:
- Team tool: Invited 2 colleagues and created the first shared workspace.
- Analytics product: Connected a data source and shipped the first dashboard.
- Finance automation: Imported transactions and categorised 20 lines.
How to validate it quickly: take the last 50 users who retained to day 30 and see what they did in the first 24 hours. If one behaviour shows up again and again, that’s your candidate.
Track Time To Value Like A Founder, Not A Data Scientist
Use median time to value, not average. A handful of power users will make averages look prettier than reality.
Practical thresholds you can work with:
- Consumer-like product: Time to value under 5 minutes.
- SMB tool: Time to value under 1 day.
- Mid-market workflow: Time to value under 7 days, but with clear progress signals.
If your time to value is 14 days and your free trial is 7 days, you’ve built a trap. Either shorten time to value or change the offer.
A One-Sentence Offer Template You Can Fill In
Offer: ‘In [Timeframe], [ICP] can achieve [Specific outcome] without [Common pain], using [Your product], starting at £[Price]’.
When you tighten this sentence, your activation event often becomes obvious. The activation event is the first proof that the promise is real.
Retention And Engagement: Prove You’ve Built A Habit
Acquisition hides product problems. Retention exposes them. If users don’t come back, you don’t have a growth problem, you have a value problem.
Use Cohorts, Not Blended Averages
Look at weekly or monthly cohorts based on sign-up date. A blended retention number can stay flat while new cohorts get worse.
Two views that matter:
- Logo retention: Do accounts stay active.
- Behavioural retention: Do they perform the core action.
For early PLG, behavioural retention is often the better leading indicator. Paid retention can lag if billing cycles are slow or if you’re offering long trials.
Define ‘Active’ As Doing The Valuable Thing
‘Active user’ must be tied to your product’s job. Logging in is not value. A sensible definition is: a user is active if they complete at least 1 core value event in a 7-day window.
Then track core action frequency. A retention curve with low frequency usually means users are stuck in ‘nice tool’ territory, not ‘can’t live without it’ territory.
Mini Case: B2B Scheduler In Manchester
Problem: They celebrated sign-ups, but day-14 retention was 9%.
Change: Activation moved from ‘created an account’ to ‘booked 1 meeting and sent 1 follow-up’.
Result: Time to value dropped from 2 days to 25 minutes, day-30 retention moved to 18% in 3 weeks, without spending more on ads.
Product Led Growth Metrics To Track Before You Scale
Let’s get specific. These are the product led growth metrics that matter before you ramp spend, hire a sales team or start chasing enterprise deals.
1) Activation Rate By Channel And Persona
Overall activation is useful, but segmented activation is where the money is. Track it by acquisition channel and by persona. If LinkedIn users activate at 22% and SEO users activate at 8%, you don’t have a traffic problem, you have a targeting problem.
Quick calc:
- Activation rate = Activated / Sign-ups.
- Activated-to-paid = Paid / Activated.
Improve activation before you buy more sign-ups. Scaling acquisition into a leaky bucket is how you burn a quarter.
2) Time To Value Distribution (Not Just The Median)
Look at the spread. If the median is 2 hours but 30% of users take 3 days, you have an onboarding inconsistency problem. That’s usually documentation, templates or empty-state design.
Founders often fix this with one move: ship a ‘first win’ template that does 80% of the set-up for the user, then let them customise.
3) Expansion: Seats, Usage And Upgrades
Expansion is the cleanest sign you’ve built something real. Early on, don’t obsess over sophisticated NRR. Track the mechanics that drive it.
- Seat expansion: Invites sent and accepted per account.
- Usage expansion: Projects, automations, reports, transactions, whatever your unit is.
- Plan expansion: Self-serve upgrades, add-ons, annual switchers.
Set a simple rule for your product: if an account hits 80% of a limit, prompt an upgrade with a clear ‘what you get next’ message. Don’t gate value, gate scale.
4) Retention By ‘Aha’ Moment Completion
This is a strong operator metric: retention for users who hit the ‘aha’ moment versus those who didn’t.
If your day-30 retention is 20% overall but 45% for users who completed the ‘aha’, your job is not more features. Your job is getting more users to the ‘aha’ faster.
5) Referral Loops You Can Actually Measure
Referrals are not ‘virality’. They are a measurable loop: user gets value, user shares or invites, the invitee activates, repeat.
Track:
- Invite rate: % of activated users who invite someone.
- Invite acceptance rate: Accepted invites / Sent invites.
- Invite-to-activation rate: Activated invitees / Accepted invites.
Small improvement here compounds. A 5% increase in invite acceptance can be worth more than a 20% increase in top-of-funnel traffic if your paid channels are expensive.
Mini Case: Compliance SaaS In Dubai
Problem: Self-serve upgrades were flat, founders assumed pricing was wrong.
Change: They added a usage-based limit aligned to ‘number of policies managed’, plus a progress bar and an upgrade prompt at 75% usage.
Result: Upgrades increased by 31% in 14 days, support tickets dropped because customers understood the next step.
Pricing And Unit Economics That Hold At Small Scale
PLG feels cheap until it isn’t. Free users still cost you: compute, support, fraud risk and your team’s attention. Your pricing needs to work when you’ve got 10 customers, not just when you’ve got 10,000.
Three practical checks:
- Gross margin: If it’s below 70% for software, you need to understand why. It can be fine, but it must be deliberate.
- Support per £1k MRR: Track hours spent supporting each £1k monthly recurring revenue. If this grows as you scale, PLG will turn into a services business.
- Payback: CAC payback under 12 months is a solid early target for many B2B products. If you’re above that, you need stronger conversion or better pricing.
Keep your revenue definitions consistent. If you’re reporting ‘annual recurring revenue’ from prepaid contracts, make sure you understand recognition rules. The IFRS 15 revenue from contracts with customers overview is a useful anchor if you’re formalising reporting.
A 7 To 14 Day Validation Path Before You Hire Or Scale Spend
You don’t need a quarter to move the needle. You need focused tests that improve one link in the chain. Pick one bottleneck, run a tight experiment, ship it, measure it.
Here’s a simple path that works in most PLG setups:
- Days 1 to 2: Define activation and ‘aha’ moment, instrument the events, pull a baseline cohort.
- Days 3 to 5: Fix the biggest onboarding drop-off with one change (template, checklist, example data, guided set-up).
- Days 6 to 10: Add a lifecycle nudge (in-product prompt or email) tied to the next value step, not ‘come back’ messaging.
- Days 11 to 14: Add one expansion trigger (limit, add-on, seat invite flow), then review upgrade behaviour.
The key is to choose a leading metric you can see move quickly: activation, time to value or invite acceptance. Revenue often lags, behaviour doesn’t.
Operational Guardrails That Protect Margin And Founder Time
Scaling a PLG motion can quietly destroy your calendar if you don’t put rules in place. Guardrails make growth sustainable.
What I’d put in place early:
- Support boundaries: Clear response times for free users, routed self-serve help for common issues, paid support tiers for heavy users.
- Instrumentation hygiene: One owner for events, naming conventions, and a monthly audit to remove dead events.
- Release discipline: Ship improvements tied to a metric, not feature requests shouted loudest in Slack.
- Pricing governance: Document why each plan exists and who it’s for, then stop inventing custom deals that break your model.
Guardrails sound boring, but they’re how you avoid ‘growth’ turning into a support queue.
Risks And Hedges Founders Miss In PLG
PLG has predictable traps. You can avoid most of them by spotting the pattern early.
Risk 1: You optimise activation but not retention. Hedge: every activation experiment must also check day-7 and day-30 retention impact, even if it’s directional.
Risk 2: You add freemium and attract the wrong crowd. Hedge: gate by capability, not by value. Free should still reach a meaningful win, but limit scale, collaboration or automation.
Risk 3: Sales and product fight. Hedge: define a clean handoff rule, for example: ‘If an account hits 3 seats and 2 projects, sales can engage’. Anything else stays self-serve.
Risk 4: You drown in data and stop shipping. Hedge: keep a weekly review with 3 questions only: What moved, why did it move, what are we shipping next.
A Quick Do And Don’t Checklist Before You Scale PLG
- Do: Tie every metric to a user outcome and a decision.
- Do: Track the full chain from sign-up to activation to retention to expansion.
- Do: Review cohorts weekly until retention stabilises.
- Don’t: Scale acquisition when activation is weak or time to value is slow.
- Don’t: Treat logins as engagement, measure the value event.
- Don’t: Offer bespoke pricing that breaks your self-serve model.
If you want one line to remember: fix the product led growth metrics first, then scale the channels.
Download The GTM Readiness Scorecard And Stress-Test Your Metrics
If you want a fast way to spot weak links before you scale, download the GTM Readiness Scorecard (0–100) and use it to audit activation, retention, expansion and the operational load behind them.
- Activation and time to value are your early warning system, if they’re weak, scaling spend just amplifies waste.
- Retention and expansion prove whether value compounds, and whether pricing and unit economics hold at small scale.
- Operational guardrails keep PLG efficient, without them, support and custom work will eat margin and founder time.
FAQs For Product-Led Growth Metrics
What are product led growth metrics in plain English?
They’re the small set of numbers that prove users reach value fast, come back without being chased and naturally convert or expand. If a metric doesn’t change a decision, it’s not one of them.
What’s the difference between activation and onboarding completion?
Onboarding completion is usually a process milestone, activation is a value milestone. Activation should predict retention, not just indicate someone clicked through a tour.
How do I choose a good ‘aha’ moment?
Start with users who retain to day 30, then look for the first behaviour that most of them did in the first 24 to 72 hours. If it’s consistent and tied to your promise, you’ve found your ‘aha’ candidate.
What retention metric should I use early on?
Use cohort-based behavioural retention tied to the core value event. It’s a clearer leading indicator than revenue retention when billing cycles or trials delay payment.
When should I start tracking NRR?
When you have enough paying accounts and time to see renewals and expansions, often 3 to 6 months of data for SMB products. Before that, track the mechanics that create NRR: seats, usage and upgrade triggers.
How do I know if my free plan is hurting me?
If support load, infrastructure cost or abuse rises faster than your activated-to-paid conversion, the free tier is misdesigned. Fix by limiting scale, adding light qualification or improving activation so the right users move to paid.
Which metric should improve before I increase paid acquisition?
Activation rate and time to value should be trending the right way, and day-7 retention shouldn’t be collapsing. If those are unstable, you’ll pay to acquire users who don’t stick.
What’s a simple referral metric for B2B?
Track invite rate, invite acceptance and invitee activation. It tells you whether collaboration is actually creating new users, not just sending emails.
