observation
Survival appeared in two modes, and their cause is confounded
Evidence tier
Research components
Answers business questions
- Why did we disappear after follow-up questions?
- What must become true to survive evaluation?
- How do we measure whether our position improved?
- How do AI systems evaluate and recommend vendors?
Two survival runs behaved differently. In CRM (E-041) the field survived but its members were almost entirely replaced turn by turn. In video conferencing (E-045) the field survived with its original members and reordered them. So survival appeared in two modes, stable membership and total churn.
The cause is not attributed here. The two runs also differed in the kind of requirement used: video conferencing used capability-framed requirements and CRM used pricing and packaging-framed requirements. Market maturity and requirement framing move together across these two experiments, so the difference cannot be assigned to either one. The existence of the two modes is registered; the cause is left open, and a dedicated experiment would be needed to separate them.
What this means commercially
- Affected buyers
- Companies in categories where the field survives rather than collapses
- Affected categories
- Observed in CRM and video conferencing
- Potential product impact
- Surviving the count is not the same as surviving as a named vendor, and the two modes may call for different responses
- Current confidence
- Narrated. Two experiments, and the difference between them is confounded, so cause is not attributed.
- Evidence tier
- Narrated
This object references
- part-of → recommendation-stability
- part-of → recommendation-survivability
- derives-from → E-041
- derives-from → E-045
Authors
Upstream Zero
Machine rendering