PracticeCritical Thinking & First Principles
What do we do with this?
The claim with four citations
The situation
Oksana asked an AI research assistant whether offline mode is table stakes in your category before the company commits a quarter to building it. She shared the brief raw, then spent twenty minutes spot-checking its sources. The decision meeting is Thursday.
Oksana · Commissioned the brief: “Started spot-checking. Two of the four don’t hold up and I haven’t got to the paywalled one. Decision meeting is Thursday. What do we do with this?”
Two sources failed, one checked out, one sits behind a $1,200 paywall. First say which claim the recommendation stands on; then decide how to spend the time before Thursday.
What you can see
E01 · The brief's summary and recommendation
Prepared by an AI research assistant at Oksana's request, dated Oct 30. Summary: offline capability has moved from a differentiator to an expectation in field-work software; buyers treat it as a baseline requirement; vendors without it face measurable competitive disadvantage. Recommendation: prioritise offline mode for Q1. 'The retention evidence in particular suggests the payback is faster than the engineering estimate implies.'
E02 · The four findings
1. A 2024 study found 68% of field technicians report losing connectivity at least once per shift, rural deployments substantially higher. [1] 2. Field-service SaaS products with full offline functionality retained users at approximately 1.4× the rate of those requiring persistent connectivity. [2] 3. A majority of established vendors in adjacent categories now ship some form of offline mode, most commonly read-only caching with deferred write sync. [3] 4. Sync-conflict resolution is the dominant engineering cost; last-write-wins, operational transform and CRDTs have well-documented tradeoffs and mature library support. [4]
E03 · The sources as cited
[1] Whitmore, J. & Adeyemi, K. (2024). Connectivity Patterns in Distributed Field Operations. Journal of Mobile Workforce Systems, 12(3). [2] Fieldstone Research Group (2025). Retention Drivers in Field-Service Software. Industry report. [3] Vertical SaaS Quarterly (2026). Offline capability across field-service platforms. Market overview. [4] Kleppmann, M. et al. Local-first software and CRDT literature.
E04 · Oksana's spot-checks, twenty minutes later
[1]: the link resolves to a real paper by those authors, but it is about scheduling latency and the 68% figure is not in it. [3]: the link is a vendor blog post citing 'recent industry analysis' without naming it. [4]: checks out; the CRDT material is real and the summary is fair. [2]: not yet checked; Fieldstone's report is paywalled at $1,200. Decision meeting is Thursday.
E05 · What finding 2 does not say
The brief gives no methodology or sample definition for the 1.4× figure. It does not say retention for whom, measured how, or whether the comparison is between products that added offline mode and products that did not, or between users who use offline mode and users who do not. Offline mode is used by field technicians in rural deployments, the population finding 1 describes as losing connectivity most.
Step 1 of 2
Which claim does the recommendation stand or fall on?
Your first choice is kept. Changing your mind later counts as a retry.