Step 3: Experiment Growth experimentation should follow some permutation of the scientific method: Stick to a Regular Cycle 1. Pose a Research Question Challenge your team’s of Tactical A/B Testing, shared understanding of how different combinations of marketing tactics are likely to impact Analysis, and Iteration your pipeline targets. This can be as general or specific as necessary, but keep in mind that it must be Growth marketing thrives on actionable within a finite time period. experimentation. 2. Form a Hypothesis Estimate the impact your efforts will The core tenet of growth marketing is a commitment have based on group experience, to evaluating efforts by measurable outcomes in a previous tests, or trusted industry recurring, iterative cycle. The first step is to develop benchmarks. Record it for later a backlog of experiments to run, each of which reference and think critically about ways should target one or more of the key success metrics it could be reliably validated/invalidated. identified earlier. 3. Conduct an Experiment Coming up with the initial list should be a Design and launch your experiment. collaborative effort and can be facilitated through Maintain proper scientific hygiene processes like pipeline assessments, persona by setting aside control groups, evaluations, or other creative exercises. Ideas ensuring statistical significance, and should then be prioritized based on the metrics partitioning measurement systems they target, potential benefit to the business, and to ensure clean, trustworthy data. resource constraints. 4. Analyze Results & Draw Conclusions It’s also important at this stage to determine the Examine the objective results of your length of your experimentation cycle. Many growth efforts. Did they improve or damage teams find running experiments on a quarterly your target metric? Can you say with basis to be an effective balance between freedom of confidence that any change was experimentation and urgency of execution. Choose directly attributable to your efforts? a cycle that’s right for your business — and once Come together with your team and it’s been determined, hold yourself accountable for develop a shared understanding of the executing within that time frame. Don’t allow your experiment’s impact. Recognize the team to fall victim to scope creep. difference between assumptions and With a prioritized list of experiments in place, the next additional hypotheses in need of testing. step is to start running tests. 5. Repeat Reprioritize your backlog of experiments and move on to the next one. 10