Back to the Playbook
Define & DesignDD-15
Run an adoption experiment
Predefine the experiment and protect its interpretation from contamination, insufficient evidence and outcome-driven metric changes.
Product statusAssisted workflowTeklens designs and assesses the experiment; running it in production is a separately authorised action by your team.
Input3 required inputsAdoption change and behavioural hypothesis · Baseline funnel, eligible traffic and instrumentation quality · Initiative focus and approved Product Brain guidance
SkillsDesign or assess an adoption experiment · Prepare task context · Review an artifact against agreed criteriaResearcher + Product Researcher Agent
OutcomeAdoption experiment: hypothesis and design, measures and checks, feasibility and analysis plan, then results and the decision.Artifact: Adoption experiment
Owner
Researcher
Human-led
Participants · Agents
Product ManagerData AnalystEngineerProduct Researcher Agent
Trigger
A proposed adoption change comes with a behavioural hypothesis.
Required context
Product BrainAdoption change and behavioural hypothesisAnalyticsBaseline funnel, eligible traffic and instrumentation qualityProduct BrainInitiative focus and approved Product Brain guidance
Optional context
Product BrainMetric definitionsFilesActual results
Skills
Design or assess an adoption experimentPrepare task contextReview an artifact against agreed criteria
Suggested conversation arc
- 1Agree the adoption hypothesisState the expected behavioural change and mechanism, primary outcome, guardrails and what worthwhile effect would change the decision.Skills: Prepare task context
- 2Design a trustworthy comparisonSpecify eligibility, randomisation unit, collaboration spillovers, sample feasibility, measurement checks and analysis plan before starting. Flag insufficient traffic.Skills: Design or assess an adoption experiment
- 3Collect experiment evidenceAfter separately authorised execution, collect actual exposure, outcomes, guardrails and quality checks under the agreed plan; record deviations and do not claim unrun experiments produced results.
- 4Decide from outcomes and uncertaintyInterpret outcomes alongside guardrails and uncertainty without changing the hypothesis after results. Record adopt, revise, reject or inconclusive and its rationale.Skills: Design or assess an adoption experiment · Review an artifact against agreed criteria
Output
Adoption experiment: hypothesis and design, measures and checks, feasibility and analysis plan, then results and the decision.
Artifacts
Creates: Adoption experiment
What good looks like
- Comparison design and analysis assumptions are explicit.
- Claims of effects or uncertainty are backed by actual data.
- Insufficient traffic or invalid measurement prevents a winner claim.
Quality gate
The Product Manager records adopt, revise, reject or inconclusive against the pre-agreed plan.
Destination
Product BrainTeklens only
Usually next
People and agents work from the same Product Brain. The owner stays accountable. Assigned agents prepare and check. A named person approves at the gate.
Choose the path that matches your role.
Set the Product AI direction with us, or test the shared Product Brain on real work with Jira and code.