Growth marketing and product growth: experiments beyond acquisition
Growth marketing applies research and experiments across the customer journey. For a product team, that means connecting acquisition with the first useful result, repeat use and revenue instead of evaluating each stage in isolation.
Follow the journey beyond signup
Map how a relevant user discovers the product, reaches a useful result, returns and pays where the model requires it. Choose a meaningful event for each stage and check how the same cohort moves between them.
A cheaper signup is not necessarily a better acquisition result. If those users rarely complete the core task or return, the lower signup cost may hide a worse outcome for the product.
What a Growth Product Manager owns
A Growth Product Manager works with design, engineering, analytics and marketing to choose product changes that could improve growth. The work includes framing the problem, prioritizing opportunities, defining the behavior to build and evaluating what happened.
The exact boundary varies by team. Agree who owns onboarding, invitations, paywalls, lifecycle messages and acquisition experiments, and who has capacity to implement changes. A shared metric does not by itself resolve those responsibilities.
Use funnels and loops for different questions
A funnel helps locate a drop-off between steps. A loop describes how one cycle can produce input for another: for example, a user shares a useful plan, a colleague opens it and later shares their own.
That mechanism still needs measurement. Count who can share, who actually does, which recipients find value and how long a cycle takes. Do not infer self-sustaining growth from an invitation count or assume every product needs a referral feature.
Choose a testable product hypothesis
Start with evidence of a user problem and explain why the proposed change could help. In a hypothetical planning app, people may stop at an empty project screen because they do not know how to begin. A relevant test could offer a starter plan while preserving the option to create one from scratch.
The intended outcome is completion of a useful plan, not simply clicks on the template. Decide which users qualify, what result would justify further work and what must not worsen, such as completion time or the quality of the saved result.
Match the experiment to the evidence you can collect
For a confusing workflow, observing people use a prototype may answer the immediate question. To estimate the effect of a product change at scale, a randomized comparison may be appropriate if enough eligible users and a suitable observation window are available.
Before reading an A/B result, check event quality, exposure and unexpected differences in how many users entered each group. Plan the sample and analysis rather than stopping at the first encouraging number. If the result is inconclusive, keep that uncertainty in the decision.
Read the effect across the product
Evaluate the intended behavior alongside later outcomes and unwanted effects. More invitations may be useful only if recipients can understand the shared object and the sender is not pushed into unwanted contact.
Decide whether to keep, revise or stop the change based on the result, its uncertainty and the cost of maintaining it. Record what the test changed in your understanding of the user problem so the next experiment does not simply repeat the same guess.
| Work | Typical focus | Needed connection |
|---|---|---|
| Product management | User problem, priorities and product behavior | Design, development and user evidence |
| Growth product work | Activation, retention, monetization and sharing | Acquisition context and sustainable product value |
| Performance marketing | Campaigns, audiences, creative and acquisition cost | Quality and later behavior of acquired users |
FAQ
Does growth marketing replace performance marketing?
No. Paid acquisition can be part of growth work, while product experiments address what happens before and after the conversion being bought.
Does a small product need an A/B test for every change?
No. Choose a method that can answer the question with the users and evidence available. A small sample can leave a quantitative experiment inconclusive, while an observed usability problem may be clear enough to address.
How long should a growth experiment run?
Plan around eligible traffic, the effect you need to detect and the behavior's natural cycle. A fixed weekly schedule does not make a retention result mature sooner.
Is a growth loop proof of product-market fit?
No. A loop describes a mechanism. You still need evidence that the intended audience gets value, returns when the need recurs and supports the business model.
Sources
Discuss your product's next step
I help define product priorities, plan launches and investigate where users drop out.
