aswritten.ai · public perspective

Perspective › NorthStarMetrics

NorthStarMetrics

6 nodes in this area of the perspective

This hub gathers the candidate \"north star\" signals — and they do not agree on one number. The strongest product claim insists [extraction quality is paramount] (Claim_ExtractionEvalsAboveLine), proposing eval data capture (Postgres table: memory text, TX output, claim count, timestamp; manual review at launch, judge workflow post-launch) and declaring it \"the most important part of the product, bar none\" (dated 2026-03-26, [echoed as an observation] (Obs_ExtractionEvalsQuote)). Against that, the [single business metric] (Narrative_KeyMetric) is named as Pilot-to-Full Conversion — the number separating a [amount] practice from a [amount]+ business. A separate [financial frame] (Obs_NorthStarMetrics) tracks the [amount] F&F raise, [amount] engagement price, and monthly burn. A measured [coverage reading] (Obs_CoverageMetric) (0.87, 13/15 claims, 2026-05-23) and a [weighting scheme] (Claim_9) for compile inputs round out the operational signals.

Witness phrases

verbatim
the most important part of the product, bar none.
verbatim
Single number that determines whether this is a [amount] consulting practice or a [amount]+ business.
verbatim
Pilot-to-Full Conversion
verbatim
Cite tool returned 13 supported citations against 15 claims (coverage 0.87)
verbatim
[amount] F&F raise target. [amount] per engagement price point.
verbatim
continuous-valued, not binary
verbatim
The hub holds two rival north stars: product-quality (extraction evals, "bar none") vs. business outcome (Pilot-to-Full Conversion as THE single number) — neither is reconciled as primary.
verbatim
A third, financial framing (raise / price / burn / invisible debt) sits alongside as its own scorecard, distinct from both product and conversion metrics.
verbatim
Coverage (0.87) is a concrete measured output, but it's a citation-quality reading, not declared a north star — an operational signal that may be mistaken for the headline metric.
verbatim
The weighting claim insists on continuous-valued, not binary scoring — a methodological stance that pushes against any single yes/no metric.

In this area

Claim Claim_9
Compile weights by: review count, recency, reviewer diversity, lens coverage — continuous-valued, not binary
Observation Obs_ExtractionEvalsQuote
Extraction quality declared the most important part of the product

4 further nodes are not published in the public projection. Each node links from the entity index.