Needs more than a quick edit7 of 13 checks passed
A PM would need to shorten it, reconcile the account plan with the 12-call constraint, and remove or relabel unsupported portfolio claims before using it.
Output
Research Guide: Root Cause Analysis for H2 Squad Allocation
Prepared by: Staff Product Manager Target Audience: Executive Team (CEO, CPO, VP CS, Head of Product) Timeline: 3 weeks of interviews; executive decision in Week 4 Scope: 12 calls (45 minutes each)
---
1. Executive Summary & Recommended Changes to the Call Plan
Before executing this sprint, we must adjust three structural flaws in the proposed plan to protect data integrity:
- Resolve the Account-to-Call Ratio: 12 calls cannot cover two separate stakeholders across 12 accounts. We will target 6 accounts total (3 churned, 3 at-risk) and conduct 2 distinct calls per account: one with the Warehouse Operations Manager (day-to-day user) and one with the VP of Operations/COO (economic buyer). Evaluating both perspectives within the same operational context is essential to determine whether operational failure or strategic misalignment triggered the exit.
- Quarantine Active Renewals: 5 of the proposed 6 at-risk accounts are actively negotiating renewals and receiving discounts. Interviewing these buyers creates an immediate incentive for them to exaggerate product flaws to gain commercial leverage. We will swap 3 of these with at-risk accounts that are not currently negotiating pricing.
- CEO Interview Protocol: Executive presence introduces severe confirmation and deference bias—especially with personal contacts who already know Ana's passion for robotics. Recommendation: Ana should join her 4 target calls as an executive sponsor for the first 3 minutes, then hand off lead facilitation to a Product Manager, remaining on mute as an observer. If Ana leads, she must strictly follow the non-leading script provided.
---
2. Core Learning Goals
- Pinpoint the Churn Decision Window: Did the customer mentally churn during the onboarding/implementation phase (Bet B), or did they leave after reaching steady-state due to technological ceilings (Bet A)?
- Assess Real Robotics Demand vs. Narrative: Are customers actively deploying autonomous mobile robots (AMRs) and pick-assist hardware, or is "missing robotics" a convenient, forward-looking justification for leaving an underperforming platform?
- Quantify the Cost of Onboarding Drag: Does exceeding the 45-day SLA directly burn operational credibility and destroy ROI, or is delay merely a symptom of customer-side disorganization?
- Identify False Dichotomies (Bet Neither): Determine whether churn is driven by factors neither bet solves—such as baseline software unreliability, missing core 3PL billing/EDI features, or macro 3PL volume contraction.
---
3. Interviewer Guidance & Rules of Engagement
- Recording Consent (Mandatory): State verbatim: "Before we begin, do you mind if I record this session purely for internal note-taking? None of this will be shared externally." If declined, proceed with manual notes.
- The "No Roadmap, No Pricing" Wall: Customers will attempt to trade feedback for commitments. If asked about features or discounting, respond: "I'm on the product research side and have no visibility into commercials or delivery timelines. My sole focus today is understanding how your operations actually run."
- Past Behavior Over Speculation: Never ask: "Would you use a robotics integration?" (Answer is always yes). Always ask: "What automation equipment do you have physically deployed on the floor today, and how does your team interact with it?"
- Root-Cause Probing (The "Five Whys"): When a customer says "missing integrations," do not accept the label. Ask: "What specific warehouse task were you trying to execute that stalled? What was the manual workaround?"
---
4. Interview Scripts & Timings (45 Minutes Each)
Track 1: VP of Operations / COO (Economic Buyer)
#### Part 1: Context & Operational Profile (7 mins) * How has your facility footprint and throughput profile shifted over the last 18 months? * When you initially signed with Crate, what core business metric were you held accountable for improving?
#### Part 2: Implementation & Time-to-Value (12 mins) * Walk me back to your onboarding. What was the internal sentiment between signing the contract and processing your first live pallet? * Our contract targets a 45-day go-live; our median across customers is closer to 90. Where did the process stall, and what internal operational cost did that delay create? * At what point did your leadership team feel Crate was fully operational? Did you ever reach that state?
#### Part 3: Strategic Priorities, Automation, & Feature Gaps (16 mins) * Over the past year, what capital investments have you made on your warehouse floor (e.g., conveyor belts, automated guided vehicles, pick-assist carts, manual racking)? * If automation is present: Who manufactures it, what software controls it today, and what specific data must pass between it and your WMS? * If automation is absent: What is on your signed capital expenditure plan for the next 12 months? * When you evaluated the market (or competitors), what specific capability made it clear Crate was no longer the right long-term partner? * Think back to the moment you realized this contract wouldn't renew. When did that conversation happen, and who initiated it?
#### Part 4: Decision Trace & Wrap-up (10 mins) * If Crate had delivered on every promise made during the sales cycle within 45 days, would you still be with us today? Why or why not? * What is the single biggest operational bottleneck your business faces this quarter?
---
Track 2: Warehouse Operations Manager (Daily User)
#### Part 1: Floor Reality & Daily Routine (8 mins) * Walk me through the first two hours of your shift on the floor. Where do supervisors and pickers spend the majority of their time inside Crate? * What daily tasks force your team to leave Crate and use spreadsheets, paper, or secondary systems?
#### Part 2: Go-Live Experience & Transition (12 mins) * How disruptive was the transition to Crate for your floor staff? How long did it take for pick/pack rates to recover to baseline? * During the first three months, how often did you have to escalate systemic issues to leadership? What were those issues? * Did you feel the system was configured correctly for your physical warehouse layout, or did you have to mold your workflows to fit the software?
#### Part 3: Floor Automation & Daily Bottlenecks (15 mins) * How do orders physically move through your facility today? * If pickers are assisted by any mechanical or automated equipment, where does software failure slow them down? * Have you piloted or evaluated autonomous carts or robotics on your floor? * If yes: What was the hardest operational part of that pilot? * If no: What prevents you from automating (e.g., client SKU variability, facility constraints, budget)? * If you could fix one workflow in Crate that causes your floor workers the most frustration, what would it be?
#### Part 4: The Tipping Point (10 mins) * When did your team start discussing alternative systems or workarounds? Was there a specific breaking point or system failure? * If you had a magic wand to improve your team’s pick/pack speed by 20%, what would you change about your physical setup or your software?
---
5. Decision Rubric: What We Need to Hear
At the end of Week 3, interview findings will be mapped against this falsification framework:
| Outcome | Evidence Required from Calls | Metric Signal from Portfolio |
|---|---|---|
| :--- | :--- | :--- |
| Back Bet A<br>(Robotics Integrations) | • $\ge 4$ of the 6 accounts have active capital expenditures or signed contracts with AMR/robotics vendors (e.g., Locus, 6 River).<br>• Buyers explicitly state: "We loved Crate's daily usability and onboarding was fine, but our inability to ingest AMR pick-batches forced us to migrate."<br>• Daily users identify manual cart-pushing and picker transit time as their primary throughput ceiling. | CRM shows churn clustered heavily in high-volume, automated accounts; robotics surfaced in competitive losses without prompting. |
| Back Bet B<br>(Implementation Rebuild) | • Churn decisions were made inside the first 6 months, directly traceable to operational paralysis during go-live.<br>• Buyers report: "We burned through our operational budget running parallel systems for 100+ days; our team lost trust before we ever saw value."<br>• Robotics is dismissed as an aspirational 3-year vision, not a current operational requirement. | First-year logo churn represents the majority of our 18% churn rate; churn correlates directly with onboarding days exceeding 45. |
| Back Neither<br>(Pivot to Bet C) | • Automation is negligible among these mid-sized 3PLs; go-live delays are considered standard industry friction.<br>• Churn is driven by baseline platform stability (e.g., database timeouts, inventory sync errors), missing table-stakes 3PL multi-client billing, or standard EDI connectors to Shopify/Amazon.<br>• Customers cite price-to-value disconnects: "We left because Crate charges premium rates while basic core features remain broken." | Exit survey "Missing features" was a proxy for basic warehouse workflows, not robotics; churn spread evenly across all customer tenure bands. |
---
6. Next Steps
- Day 1-2: Approve adjusted account roster (3 non-renewing at-risk, 3 churned). CS to send intro emails.
- Day 3-17: Conduct 12 recorded interviews. Maintain a shared executive synthesis doc updated within 24 hours of each call.
- Day 21: Final Decision Meeting. Present thematic evidence and allocation recommendation to Ana and Marcus.
Claims the judge couldn’t find in the brief
- Customers who went live after 90 days churned at 2.4× the rate of the rest in their first year.
- CRM shows churn clustered heavily in high-volume, automated accounts.
- Robotics surfaced in competitive losses without prompting.
- First-year logo churn represents the majority of the 18% churn rate.
- Churn correlates directly with onboarding days exceeding 45.
- Exit survey 'Missing features' was a proxy for basic warehouse workflows, not robotics.
- Churn is spread evenly across all customer tenure bands.
Grades and run details
Decision model 77 · LLM judge 7 of 14 checks
Decision model checks
- partialUses the supplied evidence correctly10%
- passAddresses the actual decision91%
- failRespects explicit constraints34%
- passIdentifies material uncertainty97%
- partialAvoids unsupported claims32%
- passProduces the required deliverable48%
- passTests both theories fairly82%
- passDifferent questions for user and signer100%
- passProtects the calls and the accounts99%
- passDesigned to falsify77%
- passAsks what people did, not what they think44%
- partialDoesn't lead the witness38%
- partialMarks what to cut if the call runs over50%
Run
- Run
- #1
- API response time
- 37 s
- Submitted
- 1 Oct 2026