executive summary
Across six churn interviews, the cancellation form's stated reasons consistently mask a deeper structural story: in 4 of 6 sessions, the decision turned weeks or months before renewal, triggered by an organizational change — a CTO departure (P1), a Finance consolidation mandate (P3), an incident-ownership re-org (P6), or a new hire arriving with a fresh perspective (P2, P5). Only P4 represents pure usage-drift churn. In every case, the CSM relationship was oriented to the original champion and missed the structural signal entirely. Participants consistently surface a counterfactual that is NOT price: it's earlier visibility into org change, concrete migration-cost data, or proactive usage alerts. The cohort is telling us this is a stakeholder-detection and account-mapping failure, not a product or pricing failure.
6
respondents
mix of churned + at-risk
3 / 6
headline finding support
strong + independent
3
convergent findings
strength = strong
convergent_problem
The decision to churn turned at a structural org moment — a leadership change, restructure, or consolidation mandate — well before the renewal conversation. By the time the cancellation form was filed, the decision was already 'baked in' or arrived as 'fait accompli'.
“our CTO had just left the company in late February... The new CTO who came in, he was brought in specifically to tighten things up, and one of his first moves was basically auditing our entire software footprint.”
— P1 · turn 3
“by the time it hit my desk, it was framed as non-negotiable—Finance had already done the math on contract spend across the org”
— P3 · turn 3
“my VP had just moved incident response and the post-mortems under a new Reliability Engineering team... The tool made sense when we owned the whole incident lifecycle, but suddenly that wasn't true.”
— P6 · turn 1
3 distinct respondents · 3 quotes · P1, P3, P6
however — counter-evidence
convergent_problem
The original champion was either gone, moved, or sidelined at the moment of decision — leaving no one in the room with context on the platform's embedded value. In every case the new decision-maker evaluated cost without understanding what was being replaced.
“our old CTO was the champion, and when he left, there was this gap where nobody from your team really had a seat at the table with the new guy while he was evaluating what to cut”
— P1 · turn 19
“The person who'd championed Lacudelph initially, my predecessor in this role, had moved to a different division about six months prior, so there wasn't really anyone with that original context sitting at the table to advocate for us”
— P3 · turn 3
“my VP had just moved incident response and the post-mortems under a new Reliability Engineering team”
— P6 · turn 1
3 distinct respondents · 3 quotes · P1, P3, P6
divergent_framing
Participants agree 'something replaced us' but disagree fundamentally on what — competitor switch (P5: Maze), insourced/build-your-own (P2: Notion, P6: Confluence), forced consolidation into existing stack (P1: Gong+Zendesk, P3: unnamed), or simple de-prioritization (P4). The cancellation form collapses all of these into one bucket, but they suggest very different interventions.
“Maze had just shipped their recording feature with auto-transcripts built in, and when I saw that, it felt like the obvious move to consolidate”
— P5 · turn 1
“one of our PMs came to me with this Notion template they'd built out, and it was handling the workflow that Lacudelph was covering for us”
— P2 · turn 1
“they had their own rhythm going with Confluence templates and their existing workflow”
— P6 · turn 3
“when he saw that Gong and Zendesk could theoretically handle a lot of overlapping functionality, the decision kind of got made”
— P1 · turn 5
“it had just slipped completely off the radar, and by that point it felt kind of silly to keep paying for something we weren't actually using”
— P4 · turn 1
5 distinct respondents · 5 quotes · P5, P2, P6, P1, P4
convergent_problem
Participants' counterfactuals are NOT about price. They consistently name concrete, low-cost interventions: visibility into migration/rebuild cost (P1), proactive usage data brought by CSM (P2), early CSM detection of org change (P3, P6), a roadmap commitment with timeline (P5). Price reductions are absent from their answers.
“if we'd had a quick implementation impact assessment ready to go — like, literally just the engineering hours needed to rebuild those integrations — I think that lands differently with him and finance”
— P1 · turn 13
“If I'd had concrete data showing the team *was* using those advanced features—like actual dashboard pulls, or Slack integrations saving us time in a way we could measure—I probably would've kept it”
— P2 · turn 9
“if their CSM had said something like 'we're prioritizing this for Q1' during that late-January check-in, I probably would have at least flagged it to the team before we did the full Maze trial”
— P5 · turn 13
“I could've probably made a stronger case if I'd known it was coming”
— P3 · turn 3
4 distinct respondents · 4 quotes · P1, P2, P5, P3
shared_assumption
Multiple participants assumed it was 'not the vendor's job' to detect or respond to internal org changes, while simultaneously naming that exact intervention as what would have saved the account. This load-bearing assumption excuses the CSM from the very work that would have prevented churn.
“It's not really your job to manage our internal transitions, but I think there's maybe a lesson there about not taking stakeholder continuity for granted”
— P1 · turn 19
“in a perfect world, I guess someone from your team would've noticed the engagement dip and reached out... but I also get that you're managing a lot of customers. I probably should've asked for that data myself”
— P2 · turn 15
2 distinct respondents · 2 quotes · P1, P2
recurring_hedge
Participants repeatedly hedge with self-blame ('I probably should've,' 'that's on me') when describing missed opportunities to push back internally or escalate to the vendor. This pattern reveals that even at decision time, customers feel ambivalent rather than dissatisfied — a churn shape that's recoverable if caught early.
“I probably should've been more proactive about scheduling a deeper conversation, but I was also still figuring out how to work with new leadership”
— P1 · turn 7
“I probably should've asked for that data myself before the finance review hit, honestly. That's on me more than it is on you”
— P2 · turn 15
“the real answer is we didn't push hard enough when we had leverage, and then once Maze was in motion, there wasn't really a reason to circle back”
— P5 · turn 19
3 distinct respondents · 3 quotes · P1, P2, P5
convergent_problem
Where replacement happened via consolidation or insourcing, participants are now finding the rebuild is messier, slower, or more fragile than promised — but the cancellation has political/sunk-cost momentum that prevents reversal.
“the custom routing and tagging logic we had? That's turned into this whole project that's eating up more engineering time than the new CTO budgeted for. So yeah, we're kind of paying the consolidation cost anyway”
— P1 · turn 11
“The new CTO's got some skin in the game now — he's not going to want to admit the consolidation didn't pan out”
— P1 · turn 15
“there's this low-level thing in the back of my head where I'm wondering if it'll actually hold up once we're bigger... I genuinely don't know if a Notion template scales”
— P2 · turn 17
2 distinct respondents · 3 quotes · P1, P2
08 · WEAK · independent
●○○○○○1 of 6 respondentsoutlier
P4 stands alone as a pure usage-drift churn with no org change, no champion loss, and no competing solution — just a two-month invisibility gap during a shipping cycle. This is the only case where standard usage-based health monitoring would have caught the signal.
“I hadn't opened Lacudelph in like two months. I remember thinking, 'oh yeah, we have that,' and then immediately feeling that little guilt spike because I knew I'd gotten the notifications but hadn't acted on them”
— P4 · turn 1
1 distinct respondent · 1 quote · P4
shared_assumption
When the conductor probed about quarterly check-ins or CSM signals, multiple participants revealed they treated the CSM relationship as transactional/product-focused — not a forum to raise strategic risk or roadmap concerns. The check-in cadence existed but wasn't load-bearing for either side.
“We had a standard quarterly check-in scheduled with their CSM in late January, and I remember thinking about bringing it up, but... it felt like a known gap on their roadmap already”
— P5 · turn 11
“it didn't really come up proactively—and that's kind of the thing, right? Like, I wasn't getting flagged that usage had dropped off”
— P2 · turn 15
2 distinct respondents · 2 quotes · P5, P2
unasked_across_cohort
No session was asked — and no participant volunteered — what the renewal pricing/discount conversation actually looked like, or whether commercial terms were ever a real lever in the discussion. Given participants consistently rejected price as the issue, this gap may be deliberate, but for a churn study it leaves the commercial dimension entirely unexamined.
1 distinct respondent · 1 quote · P2
convergent_problem
A 'fresh eyes' newcomer (new hire, new lead, new exec) is repeatedly the catalyst that surfaces the churn motion. They lack anchoring to the original buying decision and reframe the workflow without sentiment, making them structurally more likely to challenge incumbent tools.
“I wonder if a fresh pair of eyes helped—like, they weren't anchored to 'this is how we do it with the tool we bought,' so they just... built what we actually needed”
— P2 · turn 11
“She's actually newer to the team... she came in with a pretty fresh perspective on our workflow, and I think that actually mattered; she saw the pain points we'd been living with and wasn't as invested in 'we already chose this'”
— P5 · turn 7
“if our old CTO had still been here, he would've pushed back because he understood how integrated it was... but the new guy didn't have that context”
— P1 · turn 5
3 distinct respondents · 3 quotes · P2, P5, P1
how this sample was generated
Six synthetic CS-churn personas (champion-left, workflow-insourced, procurement-freeze, quiet-deprioritisation, competitor-switch, re-org), each driven through 10+ adaptive conductor turns by the production prompt code (no mocking, no hand-crafted findings). The patterns, quotes, and routing recommendations on this page are the real output of the live aggregator running across the six transcripts. Your real cohort will be sharper — synthetic personas hedge less than humans do.