The Levers That Actually Drive Growth

When teams say they're "doing growth", what they usually mean is that they're running acquisition activity. Paid social. Paid search. Creative testing. Funnel tweaks. Maybe a referral programme. In practice, growth becomes marketing without the brand bit, plus a layer of speed and buzzwords.

That misunderstanding is surprisingly persistent. It shows up in startups, scale-ups, and increasingly in corporates that are seduced by the romance of "growth hacking". The result is the same almost everywhere: lots of activity, some movement in metrics, and very little real leverage.

The problem isn't effort. It's focus. And more specifically, it's a failure to understand what growth levers actually are, and how to choose them.

Growth levers are not channels

A growth lever is not a channel.

Channels are where something happens. Paid search. Paid social. Email. App notifications. Platforms like Meta, TikTok, LinkedIn.

Growth levers are different. They are rooted in behaviour.

A lever is something that changes how people behave in a way that meaningfully alters the trajectory of the business. It's not confined to a single touchpoint. In fact, the best levers tend to cut across product, marketing, engineering and customer experience at the same time.

Referral is a good example. Referral is often treated like a channel: "we'll launch a referral programme". But referral is really a behaviour: someone valuing a product enough to share it with their network. That behaviour can show up in word-of-mouth, waitlists, incentives, social sharing, or simply conversation. The lever isn't the mechanic. The lever is understanding why people want to share, and removing friction so they can.

The same applies to retention and reactivation. These aren't just lifecycle campaigns. They're about understanding why someone left, what changed in their world, and whether the product now solves a problem they care about again.

Seen this way, growth levers are almost always about reducing friction, increasing control, or amplifying something people already value. This is why design thinking matters upstream — if the proposition doesn't address real unmet needs, no lever will save it.

Small changes can unlock disproportionate impact

Some of the highest-impact growth work looks almost trivial in isolation.

In one subscription business, acquisition was working well but lifetime value was poor. Customers churned early, often with reasons like "I just wanted to try it" or "I don't want a subscription right now".

Instead of pushing harder on acquisition or discounts, the team focused on a simple behavioural lever: control. They redesigned the account area to include a clear delivery calendar. Upcoming orders were visible, highlighted, and easily paused with a single click.

The engineering effort was minimal. The impact was not.

Churn dropped. Lifetime value increased. Word of mouth improved because customers felt safe trying the product. Giving people an obvious exit paradoxically made them stay longer.

What's interesting is that this change faced internal resistance. There was fear it would encourage pausing or cancellation. That fear was psychological, not data-driven, and it was wrong.

In another example, a fintech product used referral to drive growth. The initial referral programme worked well, but a small tweak unlocked much more value: timed nudges. If a referred friend hadn't completed onboarding after a few days, the original user could send a gentle reminder through the product. The nudge came from a friend, not the company.

Again, minimal effort. Massive effect. Acquisition costs dropped dramatically because trust and distribution were outsourced to real human relationships.

These changes didn't just improve one metric. They altered behaviour across the system.

Growth is constrained by psychology, not just data

Psychology is the thread that runs through effective growth decisions.

Consumer psychology is obvious, but often treated superficially. It's easy to say "put the user first". It's much harder to truly understand how people think, what they fear, what they expect, and how much cognitive load they can handle at any given moment.

Less discussed, and just as important, is company psychology.

Many powerful growth levers feel uncomfortable internally. Giving users control feels risky. Making cancellation easy feels like inviting churn. Sequencing work instead of stacking it can feel slow, unambitious, or politically difficult.

Teams also fall into the trap of equating busyness with progress. Shipping lots of things feels productive. Roadmaps look impressive. Updates sound substantial. But activity is not leverage.

The best growth work often requires resisting internal incentives that reward visibility over impact.

Choosing the right lever is a situational decision

There is no morally "correct" growth lever.

Acquisition is not bad. Discounts are not evil. Incentives are not inherently wrong. What matters is context.

Growth decisions are constrained by:

• Runway
• Burn
• Fundraising timelines
• Business maturity
• What must be true in the next three to six months

Sometimes you need to push acquisition hard to survive. Sometimes you need to preserve cash and unlock efficiency. Sometimes you accept unsustainable growth in the short term to buy time.

Scalable growth is only useful if the company still exists.

This is where understanding unit economics becomes critical. Once you understand what actually drives value, it becomes easier to judge which levers are worth pulling now.

A useful way to think about prioritisation is in terms of total unlockable value:

• If this works, what does it unlock?
• Revenue?
• Lower CAC?
• Higher LTV?
• Faster learning?
• Better future decisions?

Frameworks like ICE can help, not because they're precise, but because they force discipline. Impact becomes unlockable value. Confidence is tempered by humility. Ease becomes a question of resources and speed to learning.

Why sequencing beats stacking

Growth is not about stacking levers. It's about sequencing them.

Stacking feels productive. Teams ship more. Updates sound impressive. But stacking collapses feedback loops.

If you run five initiatives at once, you lose the chance for learning from one to inform the next. An idea that could have worked brilliantly might be written off because it was launched alongside four others.

Sequencing preserves clarity. It allows learning to compound. It keeps doors open.

There's also a human cost to stacking. Teams lose focus. Execution quality drops. Consumers get overwhelmed. Releasing too many changes at once dilutes perceived value and slows behavioural adoption.

Good growth work moves quickly, but in small, deliberate steps. Act. Observe. Adjust. Act again. Speed comes from tight iteration, not from doing everything at once.

The real job of a growth leader

At its core, growth is a decision system.

The job isn't to find more channels. It's to understand behaviour, of customers and of the organisation, and to build a system that learns fast enough to adapt.

The most effective growth leaders aren't the busiest. They're the ones who:

• Choose levers with intent
• Sequence work to preserve learning
• Resist internal psychological traps
• Let data justify decisions ("we did this because we saw that")
• Stay comfortable under pressure and uncertainty

Growth doesn't reward noise. It rewards judgment.