Choose a growth engineer when the main constraint is implementing and measuring changes in the product or technical acquisition flow. Choose a growth marketer when the main constraint is understanding audiences, shaping campaigns, and operating distribution channels.
Both can run experiments. The distinction is which work they must execute directly and which support they can rely on.
Identify the blocked experiment
Write down an experiment the team already wants to run. What prevents it from happening?
| Constraint | Hiring implication |
|---|---|
| Unclear audience or message | Investigate marketing judgment and customer understanding |
| Limited channel execution | Investigate hands-on growth marketing capability |
| Product changes wait in an engineering queue | Investigate growth engineering ownership |
| Events do not describe user behavior reliably | Clarify analytics and engineering responsibility |
| No coherent product value proposition | Resolve product and positioning questions before assigning a growth target |
A new title will not fix an experiment whose purpose is undefined.
Specify ownership across the cycle
Name who proposes the hypothesis, builds the change, checks measurement, interprets the result, and maintains successful work.
A growth engineer still needs product judgment. A growth marketer still needs to understand measurement limits. Do not make either person accountable for an outcome while withholding the systems or decisions needed to influence it.
Define quality boundaries
Growth work affects real users. State who reviews privacy, accessibility, reliability, and product consistency. An experiment should not bypass ordinary quality expectations because it is temporary.
Also define what happens when an experiment ends. Temporary code and abandoned campaigns need owners for removal or maintenance.
Assess the bottleneck you need solved
For an engineer, use a fictional product flow and ask for a bounded implementation and measurement plan.
For a marketer, use a fictional audience and channel context and ask for a testable hypothesis, message, and distribution approach.
In both cases, ask what result would be inconclusive and what would make them stop the experiment.
Write the mandate around work
Specify the product surface or acquisition stage, the available support, and the decisions the person can make. Avoid “own growth” as the entire job description.
Use the growth engineering exercise and first marketing mandate. Discuss a growth hire with Refery.