The Founder’s Growth Playbook: Loops, Retention, and What AI Just Changed

Acquisition gets the attention and the budget, but it’s rarely the real constraint on growth. The frameworks that best explain durable growth put activation and retention first — because spending to pull users into a product they’ll leak back out of is money set on fire.

This is a founder’s playbook built on ideas from two people who shaped modern growth thinking. Sean Ellis coined the term “growth hacking,” built early growth at Dropbox and LogMeIn, and gave us the 40% product-market-fit test. Elena Verna — who has led growth at Amplitude, Miro, Dropbox, and SurveyMonkey, and is now head of growth at Lovable — has the clearest thinking anywhere on growth loops and growth motions. What follows organizes their ideas into a practical sequence for founders.

The whole playbook follows one arc, and the order is not optional: earn product-market fit, instrument it so you can see value, then drive growth. Skip a step and the next one won’t hold. Read it top to bottom, or jump to the part you need.

What’s inside

Start here: is it even a growth problem?
The three levers — retention first
The three motions — how you reach and convert
Nail activation: the “aha moment”
Instrument it: one metric, tested relentlessly
Then drive growth: loops, not funnels
What AI changed
The founder’s checklist

1. Start here: is it even a growth problem?

Ellis and Verna agree on the order of operations, and getting it right is the most valuable thing you can do early: product-market fit, then data, then growth. Skipping straight to growth is a common and expensive mistake. A company that looks like it “has no growth” often has a product-market-fit or activation problem wearing a growth costume — you can’t compound something people don’t miss.

So how do you know if you have product-market fit? Ellis gave us a clean test. Survey your active users with one question:

“How would you feel if you could no longer use this product?”

40% or more say “very disappointed” — you likely have fit. Green light to pour on fuel.
Below 40% — stop. Segment the users who love you, learn why, and rebuild toward them before you scale spend.

Superhuman famously used this exact test to manufacture product-market fit — narrowing to their most-disappointed users and rebuilding for them until the score cleared the line.

2. The three levers — retention first

Verna’s growth model has two axes: three motions (how you reach and convert users) and three levers (what you’re actually growing). Take the levers first — that’s where the foundation is.

Acquisition — bring the right users in.
Monetization — convert and expand revenue.
Retention — keep users engaged and paying. The foundation.

The hard truth is that retention is the lever that makes the other two worth pulling. A great acquisition loop on a leaky product just fills a bucket with holes faster. Diagnose retention first: if your retention curve decays toward zero, you have a product problem, not a marketing problem. If it flattens — a core of users who keep coming back — you have something worth scaling.

3. The three motions — how you reach and convert

The other axis is motion — the way a company actually acquires and converts users. Pick the one that matches how your customers buy; most durable companies eventually run more than one.

Product-led — the product itself acquires, converts, and expands users through a free tier or self-serve. Works when time-to-value is fast and friction is low. Dropbox, Figma, Calendly, Canva, Lovable.
Marketing-led — content, brand, SEO, and demand gen create and capture intent. Works when the purchase needs education. HubSpot, Ahrefs, Semrush.
Sales-led — people close and expand, with the product supporting the motion. Works for high-price, complex, or regulated buyers. Salesforce, Snowflake.

The strongest companies layer motions as they scale. Slack and Figma start product-led, then add a sales motion for enterprise — often called product-led sales.

4. Nail activation: the “aha moment”

Between acquisition and retention sits the most underrated work in all of growth: activation. Ellis calls it finding the “aha moment” — the specific behavior that predicts a user will stick. The pattern is always the same: a concrete action, hit fast, that correlates with long-term retention.

Facebook — reaching 7 friends in 10 days.
Slack — a team sending 2,000 messages.
Dropbox — putting one file in one folder on one device.

Find yours, then redesign onboarding to drive every new user to it as fast as possible. It’s often the highest-return work available early on.

5. Instrument it: one metric, tested relentlessly

This is the step that’s easiest to skip, and the one that makes everything after it possible. Before you chase growth, you have to be able to see it. Once you have product-market fit, instrument the product so value is visible: wire up event analytics (Amplitude, Mixpanel, or similar) so activation, engagement, and retention show up in the data, not just in your gut. You can’t improve — or even diagnose — what you can’t measure.

Anchor that instrumentation to a single North Star Metric — one number that captures the core value you deliver. It should reflect customer value (not company vanity), be a count rather than a ratio, and correlate with revenue without being revenue itself.

Airbnb — nights booked.
Spotify — time spent listening.
Slack — messages sent within teams.

With fit earned and the business instrumented, growth becomes an experimentation engine. This is Ellis’s high-tempo testing: a weekly loop of analyze, ideate, prioritize, and test — where the output of every test is a learning that feeds the next cycle.

Prioritize with ICE — score each idea 1–10 on Impact, Confidence, and Ease, and run the winners first. The point isn’t any single test. It’s your rate of learning. Mature teams run dozens of experiments a week; start small and build the muscle.

6. Then drive growth: loops, not funnels

Now — and only now, with retention holding and the business instrumented — you drive growth. And the goal is a loop, not a funnel. A funnel is linear and lossy: you spend to put people in the top, most fall out, and you spend again tomorrow. A growth loop is a closed system where the output of one cycle becomes the input of the next — every new user helps create the next one.

Dropbox’s referral loop is the classic example: users invited other users in exchange for storage, and each new cohort fed the next. But it only worked because people already loved the product and the team could measure what was working. That order matters, and it’s easy to get backwards.

There are only a handful of loops worth knowing. The goal isn’t to run all four — it’s to find the one that fits your product and invest disproportionately there.

Content loop — users generate content that gets indexed and pulls in new users. Pinterest, TripAdvisor.
Viral loop — using the product invites others in. Calendly, Slack, Loom.
Paid loop — a user’s revenue funds acquiring the next one faster than payback erodes. HubSpot, Booking.com.
Sales-assisted loop — usage inside a company surfaces the next account to expand into. Slack, Notion, Figma landing and expanding team by team.

Most growth hacks are a distraction from the growth model.

7. What AI changed

If you’re building an AI-native product, the ground has shifted under all of this. Verna now argues that 60–70% of the traditional playbook no longer applies, based on what she’s seen taking Lovable from zero to $200M in ARR in about a year. Three changes matter most for founders.

Product-market fit is a treadmill, not a gate. Models, competitors, and expectations move so fast you have to re-earn fit every few months. Keep running the 40% test; don’t treat it as one-and-done.
Optimization collapsed in value. At Lovable, tuning the funnel drives roughly 5% of growth; shipping new features and products drives the other 95%. The compounding wins come from innovation, not from squeezing another point of conversion.
Pricing became a primary lever. Lovable changed pricing about ten times in a year and runs usage-based billing, so you pay as you get value. Giving the product away — free credits, hackathons — is treated as marketing spend, not a cost to minimize. Reckless a few years ago; one of the strongest acquisition strategies going today.

8. The founder’s checklist

Run this against your own company, top to bottom. A “no” high on the list usually invalidates the work below it — fix it first.

☐ Product-market fit confirmed? Run the 40% test. Below the line, fix the product before you scale spend.
☐ Retention curve flattens? If it decays to zero, no amount of acquisition will save you.
☐ Right motion for how customers buy? Product-led, marketing-led, or sales-led — and is a second motion warranted yet?
☐ Activation moment quantified? Do you know the behavior that predicts a user sticks — and drive people to it fast?
☐ Instrumented and measurable? Can you see activation, engagement, and retention in the data — anchored to one North Star Metric?
☐ One primary loop identified? Content, viral, paid, or sales-assisted — and are you investing disproportionately in it?
☐ A weekly testing cadence? Analyze → ideate → prioritize by ICE → test — not one-off hacks.
☐ Building in AI? Re-earn fit continuously, favor innovation over optimization, and treat pricing as a core experiment.

Growth is a system, not a hack. The companies that win aren’t the ones with the cleverest tactic — they’re the ones that build the engine and let it compound. If any of this resonates and you want to talk it through, I’m always happy to.