Start with a comparable baseline.
Choose a period that covers your usual buying cycle. Write down the source, audience, offer and owner for each step. Compare like with like: a self-serve trial and an enterprise sales conversation need different measures.
Map the path you actually operate: discovery → qualified interest → trial or meeting → first useful outcome → paying customer → repeat value. Mark missing information explicitly. A step you cannot measure is an investigation task, not proof that it is the bottleneck.
01 / CHECK THE EVIDENCE
Acquisition: the right buyers are not arriving
Are you reaching people with the problem, budget and authority your offer needs?
What to examine
Compare qualified conversations by source. Read the reasons people reply, decline or disappear. Separate relevant buyers from visits, impressions and unqualified contacts.
A useful first experiment
Test one audience and offer combination in one channel. Record qualified responses and follow-up outcomes, with an owner for every interested buyer.
Explore distribution systems →02 / CHECK THE EVIDENCE
Conversion: interest does not become commitment
Can a relevant buyer understand the value, believe the evidence and take the next step?
What to examine
Review the path from first interest to trial, meeting or payment. Check the experience on a phone, the clarity of the offer, the objections raised and the delay before follow-up.
A useful first experiment
Remove one observed obstacle: an unclear promise, missing decision information, a broken action or an unowned handoff. Measure the same step before and after the change.
Discuss the conversion path →03 / CHECK THE EVIDENCE
Activation: new users do not reach value
What must a new user accomplish before the product becomes useful to them?
What to examine
Define a meaningful first outcome for your product. Follow new-user cohorts from signup through that outcome, including elapsed time and the step where they stop. Login counts alone do not establish value.
A useful first experiment
Improve the step preventing that outcome. It might be setup, confusing instructions, missing data, an integration failure or a need for assisted onboarding.
Explore product and workflow systems →04 / CHECK THE EVIDENCE
Retention: useful early activity does not last
Do customers return to the useful outcome often enough to keep paying?
What to examine
Compare customers who joined in similar periods, plans and use cases. Review cancellations, repeat usage, support conversations and renewal reasons. A blended average can hide one segment's problem.
A useful first experiment
Choose a specific segment and investigate its reason for leaving. Test a change tied to that reason before increasing acquisition into the same experience.
Review the retention constraint →05 / CHECK THE EVIDENCE
Operations: demand creates work the team cannot carry
Where does useful customer activity wait for a person, a tool or missing information?
What to examine
Trace ownership and elapsed time through qualification, onboarding, delivery and support. Record repeated manual tasks, failed transfers and exceptions that need recovery.
A useful first experiment
Fix the most costly handoff. Agree the source of truth, owner, exception path and recovery procedure before adding an automation.
Explore revenue operations →Use the numbers to locate a question.
Illustrative example: 100 relevant trials produce 20 users who reach first value, and 10 of those become customers. That means 20% reach first value and 50% of the activated users become customers. Those percentages are invented for this example; they are not targets or client results.
This suggests investigating why 80 trial users fail to activate. It does not prove that onboarding is the only problem. Check traffic quality, event tracking and the product's definition of value before choosing a change. Adding trials would amplify the same unresolved path. Use our SaaS onboarding audit checklist to inspect setup friction and define first value.
Write one decision before starting work.
Use this brief with your team. Keep the experiment narrow enough to interpret, and decide what evidence would change your mind.
- Constraint: which step is limiting a useful customer outcome?
- Evidence: what have you observed, and what is still unknown?
- Hypothesis: why would the proposed change affect that step?
- Scope: what changes, who owns it, and how can you recover if it fails?
- Measure: what baseline and outcome will you compare?
- Decision: when will you continue, change or stop?
Aim toward $100k MRR by improving a path that carries customer value and revenue. The order of work depends on your stage, market, pricing and retention; this framework does not promise a revenue result.