The Four-Step Growth Levers Process
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- —
Matt Lerner's core process for early-stage companies that have a product and some customers but no scalable acquisition. It starts from the observation that roughly 90% of a company's growth comes from about 10% of what it does, so the job is to find that 10% fast rather than doing a little of everything. Step one maps the growth model, a mathematical picture of how the business acquires, engages, retains and monetises customers, so points of highest leverage can be derived rather than guessed. Step two studies the customer's journey, which has nothing to do with your product. Step three filters ideas against the model and tests them in a week or two. Step four is the mindset shift: fewer things, faster, with mistakes treated as learning.
Origin
Extracted from Deep Dive with Ali Abdaal
How to run it
- 1
Accept the 90/10 reality
Assume a small handful of actions will produce almost all your growth, and that a startup with twelve months of runway cannot afford the big-company luxury of trying hundreds of things. The goal of everything that follows is to locate that handful as quickly as possible.
Pro tip Read the well-documented case studies (Dropbox, Canva, HubSpot) and look specifically for the two or three mechanisms that carried the growth, not the tactics list.
Watch out Large, cash-rich companies can afford to waste money on dozens of failed products. Copying their breadth on a startup budget is how runway disappears.
- 2
Map your growth model
Build a mathematical representation of how the business acquires, delights, engages, retains and monetises customers. Put real data into it, then derive mathematically where the points of highest leverage sit.
Pro tip Keep the top-level map to one simple page; the detailed spreadsheets sit underneath each box.
Watch out A growth model with no numbers in it is a diagram, not a model. It cannot tell you where the leverage is.
- 3
Map the customer journey
Study what the customer was actually trying to do, what they thought they were looking for, where they were looking and what questions they had. Nobody woke up looking for your product, and often not even for your category, so the task is to turn up where they already are and look like the thing they think they need. Use jobs-to-be-done interviews rather than surveys.
Pro tip Interview people who became customers recently so they can still remember the sequence of events.
Watch out Do not use these interviews to ask what people think of you or your product. Empty your mind of preconceptions and ask about their life.
- 4
Filter the ideas, then experiment fast
You will have lots of ideas and most will be bad. Filter first by the growth model, keeping the ones that hit the rate-limiting step, then design experiments that test the riskiest assumption within a week or two rather than building the whole thing.
Pro tip If an idea is big, on-model and carries no real risk, skip the experiment overhead and just do it.
Watch out Hard, slow ideas such as writing a book or building a product are exactly the ones people skip testing, and they are the most expensive to get wrong.
- 5
Shift the mindset
Succeeding at school or in a corporate job means learning everything, doing every task well and making no mistakes. Growing a startup requires the opposite: doing fewer things, moving fast and generating mistakes you learn from. Document experiments and learnings so the mistakes compound into knowledge instead of being repeated.
Pro tip Make predictions before each experiment and have the whole team place bets, which kills hindsight bias when the results come in.
Watch out People who fear being blamed hide their mistakes, so nobody learns and the next person repeats them.
In the wild
Looking back at almost eleven years at PayPal, Lerner realised about 90% of the company's growth came from roughly five things: getting onto eBay, a bot that bid on eBay items and asked sellers whether they accepted PayPal, outreach to the web developers building e-commerce sites, pre-integration with the shopping carts and hosting platforms that most stores were built on, and international expansion. By the time he left, the company had 24,000 employees and was spending $5 billion a year, having tried hundreds of things and launched dozens of products. The waste did not matter at that scale, but it proves the point for a startup with twelve months of runway.
→ About five levers accounted for roughly 90% of the growth of a company that scaled to 24,000 employees.
Lerner cites Canva as a clean example of the same pattern. The company runs roughly 750,000 landing pages targeting search terms such as award certificate template, birthday card template and social media post design, so people googling those terms land in the product. The second half of the engine is an extremely well-optimised self-serve onboarding experience, so a person with no Photoshop experience can get into the tool and be successful immediately. Two mechanisms, not twenty, carry the growth.
→ Search-driven acquisition paired with frictionless onboarding became the dominant growth lever.
Common mistakes
Trying a little bit of everything
The single mistake Lerner says he saw over and over as a VC: startups spread themselves across dozens of half-executed initiatives instead of finding the few levers that actually scale. Every channel has a power-law distribution, so half-hearted effort in seven channels produces nothing in any of them.
Skipping the model and going straight to tactics
Without a growth model, idea selection becomes a debate about which suggestion sounds best. With one, you can point to the rate-limiting step and rule most ideas out on arithmetic.
Running experiments and never writing down the learning
Ali admitted his team often remembered a past attempt only two years later, after repeating the same mistake. Documenting the hypothesis, result and open questions is part of the process, not admin.
From the transcript
“these are the things the actions that will have the biggest impact on the growth of your business”
“okay step two is you need to really understand your customer's Journey which has nothing at all to do with your product nobody if you're…”
“and you need to move quickly which means you can't do all the things and you've got to make mistakes and the value is those…”
From the episode
How to Scale your Startup with Growth Levers: Matt Lerner