Growth lever

Definition
A growth lever is the one improvement that lifts the whole business when pulled, because it sits on a mechanism that compounds.

Why it matters

Resources are tight, and spreading effort over a dozen initiatives tends to produce a dozen modest results. Lever thinking asks one question: which single improvement would accelerate growth most right now. It also makes experiments more useful. A test aimed at a real constraint teaches something even when it fails, unlike a random tactic tried without a hypothesis.

How to apply it

  • Map the buyer's path to find where it leaks, using numbers plus what customers and staff say.
  • List three to five candidate levers instead of jumping at the first plausible one.
  • Score each on impact, confidence and effort, then commit to one or two.
  • Decide before starting what changes, who owns it and which metric proves it worked.
  • Keep the first test small, so the team learns whether the lever is real before committing further.
  • Once it delivers, build it into the standard process and move on to the next.

What it is

A growth lever is a change that moves a number much further than the effort it takes, because its effect repeats every period afterwards. Cutting monthly churn from fifteen per cent to ten per cent is one: every future month, more customers remain. Adding a minor feature is not one: it changes almost nothing downstream.

The word comes from the physical tool. A small push at the right point moves a heavy load. In a business, the right point is usually the constraint, the stage where the most people drop out or wait. A growth driver is what already produces growth. A lever is the thing you act on to get more of it.

Common mistakes

  • Picking a lever because it is easy. A task that fits the calendar is not necessarily the one that moves the number.
  • Pulling several at once. If three things change together, you cannot tell which one worked.
  • Guessing the constraint. Look at where people leave or wait, with real numbers, before choosing.
  • Declaring victory early. A lift for one week is not a lever. Check that it holds over a full cycle.
  • Confusing a lever with a driver. A driver is what already produces growth. A lever is what you choose to change to get more of it.
  • Staying on a finished lever. Once it delivers, move to the next constraint instead of squeezing the last 2 per cent.
Worked example

Suppose a software company has a cohort of 200 paying customers, losing 5 per cent of them each month. Cutting churn to 4 per cent looks like a small change, but after a year about 108 of the original customers remain at 5 per cent and about 123 at 4 per cent, and the gain keeps compounding. The team finds that few customers reach the second key action in their first week. It tracks retention and conversion across its web and mobile products in Amplitude, then tests a shorter onboarding path on half of new sign-ups and watches whether the second action happens sooner. The lever is clear before any large investment is made.

Tools in the example

Some links are affiliate links: we may earn a commission at no cost to you. It never decides a ranking. How we work with partners

  1. Article

    Churn rate

    A common lever, since improving it compounds every year.

  2. Article

    Onboarding Funnel

    A frequent home for a lever in the first session.

  3. Article

    Prioritisation

    How to choose between candidate levers.