How to achieve product-led growth?

Summarize this article with LLMs

I was a VC for 6 years and founders I worked with had the same problem:

The pitch deck said "product-led growth". The actual growth model was "founder-led sales plus vibes".

Dashboards looked nice. Activation funnels were defined. But when we looked at who drove revenue, it was still:

Founders and early AEs pushing deals over the line.
Customer success teams manually onboarding key accounts.
"PLG" flows that looked good in screenshots but didn't move expansion or retention.

This post is a practical playbook for getting closer to real product-led growth, based on patterns I saw across dozens of early-stage teams I've worked with, invested in, or advised.

Who this is for

This is mainly for:

Early-stage SaaS founders who have a PLG slide in the deck but close deals through founder-led sales.
Product leaders who are being asked to "make it product-led" without clear metrics.
Growth / success people who want the product to do more heavy lifting before humans step in.

If you are evaluating Skene, you can also keep this post open next to the main site and compare the patterns here to how your own onboarding works today.

Key takeaways

You need clear "aha" and "habit" moments, not just signups.
Onboarding should lead to one real outcome, not a feature tour.
The product should own more of the repetitive work so sales and CS can focus on the parts that actually need humans.

TL;DR - what actually works

Pick one "aha" and one "habit" metric, and align the team around moving those, not vanity signups.
Design onboarding around a critical first use case, not a tour of every feature.
Make your product do the first 80% of success work automatically, and reserve humans for the last 20% that's truly bespoke.
Give sales and CS strong product signals (who's activated, who's stuck, who expanded) so they amplify, not replace, PLG.
Ship small experiments weekly in onboarding, pricing, and packaging, and read them through revenue, not just engagement.

Step 1: Define "product-led" in numbers, not adjectives

In most board meetings, "PLG" meant:

"Users can sign up without talking to sales."
"We have a free tier."
"We built an onboarding checklist."

None of that guarantees product-led growth.

For early-stage teams, a more useful working definition is:

"New revenue and expansion are mostly predicted by what users do in the product, not by how many calls we had with them."

To operationalize that, sit down with your team and answer three questions:

What is our 'aha' moment?
- The first action that strongly correlates with higher conversion or retention.
- Examples:
  - A payroll tool: “Run first payroll with real (or test) employees.”
  - A dev tool: “Deploy first change to production.”
  - A sales tool: “Send the first sequence to a real list.”
What is our 'habit' moment?
- The recurring behavior that separates casual users from power users.
- Examples:
  - “Log in and complete at least one key workflow per week for four weeks.”
  - “Invite at least two teammates who also complete one key workflow.”
Which of these can we measure reliably today?
- If you can't measure it, PLG is a story, not a system.

Write these down. For the next quarter, your product-led work should mostly be about:

Getting more signups to the "aha" moment.
Turning more "aha" users into "habit" users.

Everything else is a nice-to-have.

Step 2: Fix onboarding so it leads to one real outcome

Across my portfolio, the common trap was feature tour onboarding:

Users got a tour of 10 features.
They touched none of them deeply.
They left with no clear sense of “this solved X for me”.

The teams that made progress on PLG did the opposite: they picked one critical first job-to-be-done and designed onboarding only around that.

A practical pattern you can copy

Pick a single first outcome.
- "Get your first dashboard live."
- "Import your first dataset."
- "Send your first campaign."
Design a 3–5 step path to that outcome.
- Each step is concrete, user-facing, and verifiable in data.
- Example for a sales engagement tool:
  - Connect your email account.
  - Import or create your first contact list.
  - Choose a starter sequence template.
  - Send the first sequence to at least 10 contacts.
Measure completion of each step.
- Not just "onboarding completed", but which step users drop off at.
Instrument "time to first outcome".
- How long does it take a new account to hit the first "aha"?
- Early-stage teams that won here often cut this from weeks → days → hours.

The inflection point that really matters is when founders can tell you, without opening a deck:

"50% of new workspaces hit first value within 24 hours."
"Users who hit first value are 5x more likely to convert to paid."

If you can't say that yet, that's your PLG roadmap.

Step 3: Decide what the product does vs. what humans do

In almost every early-stage portfolio company, there was a moment where:

The product could in theory guide users to success.
But the actual guidance lived in:
- Notion docs.
- CSM’s heads.
- Miro or Figma flows that never made it into the product.

Product-led growth doesn't remove humans; it moves their effort later in the journey, and makes it more leveraged.

A simple exercise

Take your current onboarding / success process and list out the steps a new customer goes through, from "heard about you" to "renewed or expanded".

For each step, ask:

Can the product detect if this step is done?
Can the product suggest or automate the next best action?
Where do we still truly need a human, and why?

Examples:

Product should own:
- Checking whether key integrations are connected.
- Detecting that a workflow has never been run.
- Nudging users when they’re close to finishing a setup task.
Humans should own:
- Navigating complex stakeholder politics in the account.
- Helping customers re-think a process (not just clicking through a wizard).
- Strategic conversations about pricing and roll-out.

Your product-led roadmap becomes a list of “move this step from human-only to product-first” items.

Step 4: Give sales and CS better product signals

One of the biggest misconceptions I saw in PLG pitches was:

"We'll turn off sales and let the product do the work."

In practice, the strongest PLG companies used sales and CS as multipliers:

The product handled signup, onboarding, and early activation.
Sales came in where there was real intent and real usage.

In the healthiest setups I’ve seen, the internal dashboards had:

A clear view of which accounts were:
- Signed up but not activated.
- Activated but under-using key features.
- Power users who could become champions.
A way for sales/CS to:
- See that view without asking an analyst.
- Trigger targeted outreach based on product events (not just renewal dates).

If your PLG motion feels "not working", check:

Do sales and CS have a shared, trusted view of product usage?
Can they see which users actually hit the "aha" moment?
Do they know who invited teammates, who created key objects, who is blocked?

If not, your next PLG investment might not be a new growth hack, but a simple internal product usage dashboard and alerts.

Step 5: Run small, honest experiments

A final pattern from the portfolio: the teams who made PLG work didn't win with one big bet. They:

Shipped small onboarding and pricing changes weekly.
Measured them with simple, clearly owned metrics.
Were brutally honest when an experiment didn't move the numbers.

Here’s a template you can reuse:

Pick a single metric.
- Example: "% of new workspaces that hit 'aha' within 7 days."
Define one small change.
- Example experiments I saw work:
  - Swap the order of steps in onboarding to front-load value.
  - Add in-product templates tailored to 1–2 concrete use cases.
  - Gate advanced features until the basics are set up (to avoid overwhelm).
Decide in advance what success looks like.
- "We'll roll this out if we see a +15% relative lift in activation over four weeks."
Review with the team and write one paragraph of learning.
- Even if it fails, document: what did we learn about our users and our product?

Over a year, this compounds. The signals are obvious when it's working:

Onboarding metrics steadily improved.
Expansion happened earlier and with less hand-holding.
The team talked in terms of user behavior and experiments, not just roadmap features.

A sanity checklist for your next board or investor update

If you want to show that your PLG story is real, not just a buzzword, here's a short checklist you can use:

We've written down our 'aha' and 'habit' moments, and we can measure them.
Our onboarding is designed around one clear first outcome, not a generic feature tour.
We know which parts of onboarding are product-led vs. human-led, and we have a plan to shift more to product over time.
Sales and CS work from shared product usage signals, not just gut feel.
We run small, measurable experiments, and can show at least a few that improved activation, expansion, or retention.

If you can confidently say "yes" to most of these, you're much closer to true product-led growth than many companies whose decks say "PLG" in big letters.

And if you're still in the messy middle, that's normal. The teams I saw succeed didn't start as "PLG-native". They just got very honest, very early, about what the product was actually doing for growth and then iterated relentlessly from there.

Common PLG mistakes to avoid

Treating PLG as a marketing slogan instead of a quantitative model. If you cannot describe your aha and habit moments in data, you are not yet running a real PLG motion.
Building feature tour onboarding instead of outcome-led onboarding. Walking users through every feature once does less than helping them reach one meaningful outcome quickly.
Assuming PLG means “no humans”. The best PLG companies still use sales and CS, but they plug them in after product signals show real intent.
Ignoring internal product data quality. Without trustworthy product events and funnels, you end up debating anecdotes instead of improving activation and expansion.
Over-investing in tools before behavior. Buying a CDP or analytics stack without a clear PLG model often creates more dashboards, not more revenue.