Ground it
Connect customer signals, business context and engineering reality. Surface contradictions and explicitly mark assumptions.
Private working brief · August 2026
A product hypothesis for Picnic
Build the decision system,
not another AI workspace.
Murdo — this is an outside-in hypothesis, not a claim to know the business better than the people building it. I may be wrong. The question is whether Picnic can become the place product teams trust before they build.
Signal Evidence Decision Shared PRD Slices Observe
Prepared by Jeetesh Khodiyara · Senior Product Manager candidate
01 / The bet
AI has made output cheaper. It has not made judgement easier.
Picnic says that deciding what is worth building is now the bottleneck. Its product signals a compelling ambition: bring customer signals, business context and engineering complexity together, then help teams create and align.
My hypothesis: Picnic’s early advantage will come from making a small number of consequential decisions evidence-linked, challengeable and execution-ready—before it expands into a broad AI workspace.
Why this mattersA fast PRD, wireframe or pull request is valuable only when the decision behind it is clear, credible and owned.
02 / The initial wedge
Make the Opportunity Brief the trusted record of a decision.
Rather than treating it as one more generated artefact, make it the structured object that connects a product opportunity to the evidence, trade-offs and delivery work behind it. Everything that follows should preserve that decision’s intent.
Decide it
Record the problem, alternatives, confidence, accountable owner and the outcome the team expects to create.
Carry intent forward
Use the brief to create a living PRD, plan and system design—not disconnected AI documents.
Learn from it
Bring post-launch outcomes back to the original decision so teams improve judgement, not merely delivery speed.
03 / What follows from the wedge
If the Opportunity Brief is trusted, it must carry intent all the way through delivery.
The first downstream object is a living PRD—not a hand-over document. It is the shared contract between the person accountable for product intent and the engineer, whether that engineer is human or agentic AI. The plan, system design and operational feedback must remain connected to it.
Make it buildable
Turn the decision into a PRD holding the problem, scope, acceptance criteria, constraints and the specificity an engineer needs to build the right thing.
Keep it current
When evidence, scope or a trade-off changes, update the PRD. It remains the source of truth for what is being built and why.
Deliver in slices
Let the living plan follow the PRD. Each slice leaves the product usable, protects against regression and makes learning possible.
Explain the system
Use a design document to make architecture, interfaces, data flows, failure modes and operational choices explicit before complexity becomes hidden.
Design for real use
Keep usability and performance in the PRD, plan and acceptance criteria—not as end-stage checks.
Learn from reality
Log the right product and technical events so teams can find issues, understand performance and improve the next Opportunity Brief.
04 / A deliberate trade-off
Trust before breadth.
Picnic’s visible surface area is exciting. It also creates a familiar early-stage risk: becoming a place where everything can be produced, but no critical decision is reliably improved.
Prioritise
- One repeatable, expensive product decision
- Evidence provenance and confidence
- Clear decision rights and hand-offs
- Decision-to-delivery traceability
- Outcome learning and reuse
Defer until trust is earned
- Generic all-purpose co-pilot behaviour
- Feature breadth for its own sake
- Complex multiplayer workflows without a core loop
- Outputs that look finished but lack a decision record
05 / What I would test
Decision confidence
Share of opportunity briefs supported by attributable evidence; time from signal to an agreed decision; cross-functional reuse.
Decision quality
Repeat use by a team, lower rework after hand-off, and a demonstrable connection between the chosen opportunity and an outcome.
False confidence
Evidence must be inspectable. Assumptions and uncertainty must not be hidden by polished AI output.
06 / The first 90 days
Find the decision worth owning.
Diagnose
Work with design partners to map the current decision journey: signal sources, recurring trade-offs, hand-offs, workarounds and where confidence is lost.
Output: one sharply defined, repeatable decision to own.Prove
Prototype the evidence-linked Opportunity Brief with a small number of real decisions. Instrument time-to-alignment, evidence use and rework.
Output: a trusted loop, not a feature catalogue.Productise
Turn what repeats into a clear workflow, then connect its decision record to the downstream artefacts, usable slices and observable outcome review.
Output: an evidence-backed product and commercial learning plan.07 / Why I care about this problem
I work from first principles when the apparent problem is not the real one.
I found the structural transaction-data issue beneath four unsuccessful programmes, then helped reduce unrecognised-transaction calls by more than 50%.
I led integrated product strategy across product, commercial and technical constraints; active users grew from 6m to 12m and revenue-share ARR from £3.6m to £7.2m, while costs stayed flat.
I explored a governed specialist-agent workflow for Product Managers—from concept and research through strategy, roadmaps and technical delivery—with explicit hand-offs, traceability, entitlements and tenant isolation.
08 / The questions that would change my view
Where am I wrong?
- Which decision currently creates the most avoidable product rework for Picnic’s design partners?
- Who must trust the decision record first: the Product Manager, product leader, designer or engineering lead?
- What evidence do teams have today but cannot reliably connect to a decision?
- Is the initial commercial promise improved decision quality, faster alignment, less rework—or another outcome?
If this is directionally useful, I would welcome the opportunity to test the hypothesis with you and turn it into a sharper product decision.