Method
See how the research gets from question to recommendation.
Every Parse service runs through the same seven stages. What varies is the evidence and the decision in question — not the discipline.
Three phases
Seven stages · One report
Phase 1
Frame the decision
01
Agree the decision
02
Set the scope
Phase 2
Build and test the evidence
03
Collect the evidence
04
Check what the evidence supports
05
Challenge the findings
Phase 3
Prioritise and deliver
06
Prioritise what to change
07
Deliver the report
Decision Report
Findings · Evidence · Priority · Next actions
Phase 1
Frame the decision
01
Agree the decision
Define the question the research needs to answer.
02
Set the scope
Agree what will be investigated, what evidence is available and what sits outside the work.
Output — agreed decision brief and scope
Illustrative example
Not client research
Decision brief
Trial activation
01
Decision
Understand what prevents trial users reaching first value before investing further in acquisition.
Question
Why are trial users signing up but failing to reach first value?
Evidence available
Onboarding analytics
Support tickets
Previous customer interviews
Boundaries
Trial sign-up through the first meaningful product outcome.
Example content demonstrates the research structure only. Actual evidence, findings and priorities depend on the client question.
Evidence and challenge
Illustrative example
Source 01
Product analytics
Observation
Users complete sign-up, but many do not complete initial workspace setup.
What it may support
Early setup friction may be blocking progression.
Source 02
Support tickets
Observation
Repeated questions concern workspace configuration and what to do after account creation.
What it may support
The first-run experience may not make the next action clear.
Three sources point toward one explanation
Provisional finding
Setup friction is preventing some trial users from reaching first value.
Supporting evidence — Product analytics · Support tickets
Conflicting evidence
Some users complete setup successfully but still fail to reach meaningful value.
Status
Needs refinement
Not yet strong enough as a final finding
Refined interpretation
Setup friction appears to be one barrier, but it does not fully explain the activation problem.
Example content demonstrates the reasoning structure only.
Phase 2
Build and test the evidence
03
Collect the evidence
Gather the relevant client and public evidence.
04
Check what the evidence supports
Separate strong support, weak support, contradictions and missing evidence.
05
Challenge the findings
Test whether each proposed finding survives alternative explanations and conflicting evidence.
Output — evidence map, contradiction register and reviewed findings
Phase 3
Prioritise and deliver
06
Prioritise what to change
Turn supported findings into a clear order of attention.
07
Deliver the report
Show the conclusion, supporting evidence, priorities and next actions in a usable final report.
Output — priority view and the Decision Report
Priority view
Illustrative example
01
Clarify first-run setup
Evidence support — Strong
Relevance to decision — High
Next action — Reduce ambiguity around workspace setup and the immediate next step.
02
Test expectation-setting before sign-up
Evidence support — Moderate
Relevance to decision — Potentially high
Next action — Test whether pre-sign-up messaging is setting an expectation the early product experience does not meet.
03
Investigate post-setup loss of momentum
Evidence support — Incomplete
Relevance to decision — Still uncertain
Next action — Gather evidence on what prevents users who complete setup from reaching meaningful value.
Priorities belong to the illustrative example only. Actual priorities depend on the client question and the evidence that survives challenge.
Decision report
Illustrative structure
01
Finding
Early setup friction contributes to activation loss, but does not fully explain it.
02
Evidence
Product analytics and support evidence, with conflicting evidence from users who complete setup.
03
Confidence
Moderate.
04
Priority
High.
05
Next action
Clarify first-run setup while investigating post-setup loss of momentum.
Not client research
Example only
Why the order matters
Evidence is not interpretation. Interpretation is not recommendation.
01
Evidence
What was observed, recorded or supplied. Kept identifiable by source, before anyone decides what it means.
In the report
Source named · Record kept
02
Interpretation
What the evidence appears to explain, tested against alternative explanations, with confidence stated honestly.
In the report
Strength stated · Alternatives tested
03
Recommendation
What to change, test or investigate next, stated separately so the reasoning behind it stays checkable.
In the report
Linked to a finding · Priority given
If a report skips straight to recommendation 3 without showing 1 and 2, there is nothing to check it against.
See which service fits your question.
Book a 30-minute call, or look through the three services first.