Research
Measurements and methods from OptiVis AI Labs. Each one states its source, its denominators and what it does not show.
12.6% vs 1.9%, n = 19 of 151 and 76 of 3,988 sessions
ChatGPT referrals at an auto-parts store, Q3 2026
Share of sessions that completed checkout, by referral source, in Shopify attribution, at one store over one quarter. It does not show that ChatGPT caused the sales; it is attribution at one store.
Recorded · Updated · Source: Shopify attribution
2 of 37, n = 40 pages in the panel, 37 with a price on two surfaces
Shelf Ledger, issue 0: do a product page's surfaces state the same price?
A frozen panel of 40 public Shopify product pages, each read once: how often the structured data, the visible text and the Shopify product JSON state different prices, before and after a read of each flag. It does not show that a disagreement costs a store sales, and the panel is hand-picked, not a sample of Shopify.
Recorded · Updated · Source: Fact Split, one fetch per page
Method, no figure published
How AI Shelf Check measures readiness
Every check, its weight, its confidence level and the published sources it rests on. It does not report a score or a result for any store.
Updated · Source: Published sources named on the page
Method, no figure published
How Fact Split and Span decide a match
How Fact Split compares structured data, visible text and Shopify JSON, and how Span stamps a clause as a verbatim span. Every normalization and limit stated. It does not report a result for any page or store.
Updated · Source: Fact Split and Span
Illustrations, no figure published
Explainers: eight mechanisms, one sentence each
Interactive illustrations of Falsify, Fact Split, Span, the Jev distillation, the gpur router, span-verified retrieval, fitment search and the gated fitment rebuild. Invented example text; the only recorded figures are the Jev reading. They do not show a measurement; the example text is invented.
Updated · Source: Recorded run, 2026-10-05 (Jev reading only)
Applied work, no figure published
Checking logistics decisions against operating records
OptiVis Logistics is part of OptiVis. It checks every load against the truck's location and the driver's remaining hours, compares the pay with past loads on the same lane, and reconciles fuel records. A dispatcher books the load. It does not show a measured result.
How does OptiVis use AI in its own trucking operation?
OptiVis Logistics applies AI to its own trucking operation. The work includes checking loads against truck location and driving hours, reviewing lane pay history and reconciling fuel. OptiVis AI Labs links to this operation as an applied example. The public description establishes the workflows involved; it does not establish a measured improvement in profit, fuel use or delivery performance.