Executable

When will this actually be sent to me?

ALS-COFFEE-1.3-DELIV-001

This question is asked as a test and produces a result for a real page.

Assertion

delivery_window_stated is_stated true

Applicability

applies when
the product is physically shipped to the buyer
signal
any purchasable physical coffee product

Accepted evidence

  • product_description — a stated dispatch or delivery window with a number and a unit e.g. Orders ship within 2 business days.
  • product_faq — the same statement in FAQ structured data
  • shipping_policy — the same statement in the publicly readable shipping policy

Explicitly insufficient evidence

What does not count, and why. Published because the omission is where a check quietly becomes a claim.

  • "free shipping" or a shipping price
    Price is not speed, and crediting it was the live false positive that caused the engine's entire evidence-validation layer to be written.
  • "fast shipping", "ships fast", "quick dispatch"
    An adjective with no number and no unit states no window a buyer or a machine can act on.
  • a returns window such as "returns accepted within 30 business days"
    A returns period answers a different question, and the engine currently credits it as a delivery window — a pinned corpus defect this entry inherits.
  • "roasted to order" or "always freshly roasted"
    These describe the workflow rather than the latency and are compatible with a single weekly roast day that adds days before dispatch.
  • a transit estimate such as "delivered in 3-5 days" with no handling time
    Transit alone omits the handling component, which is the part that varies most between roasters and the part a roast-to-order operation controls.
  • a delivery estimate shown only in the cart or after a postcode is entered
    It is not on the product page, so a pre-purchase reader and an automated check cannot see it.

Conflict rules

  • When the product page and the shipping policy state different windows — then the engine reports the first supporting sentence it finds and names its surface; the standard records that the page disagrees with the policy precedence: product_description > product_faq > shipping_policy
  • When the shipping policy could not be read — then report that the surface was not inspectable rather than that no window is published — the difference between a finding and an accusation

What a pass licenses

A pass establishes: That readable text on a product surface or the public shipping policy states a dispatch or delivery interval with a number.

A pass does NOT establish: That the interval is honoured, that it covers handling as well as transit, that it applies to this product rather than to orders generally, or that the interval is about delivery rather than returns.

Discrimination

Measured fail rate: 60.8% (45 of 74 adjudicated; 26 of the 100 rows it was asked returned "requires store access" and are excluded from the denominator, because counting an undecided row as a pass is a different measurement)

95% interval: 49.4% – 71.1%, against a 15-85% target band (wilson95 vs target band)

Verdict: Discriminating — the measured rate lies inside the target band, so the answer separates stores from one another.

26 of the 100 rows this entry was asked returned `requires_store_access` — the engine could not decide them. They are excluded from the denominator, which is what `n_adjudicated` means. v1.1 published 45% for this entry, which is the same failures over the ASKED rows: that reading counts an undecided row as a pass. Both numbers are in this record so the reading is visible. Recomputed by `standards/coffee/issue_v1_2.ts` from experiments/v3-2/standard_run3.jsonl through the shipped `verdictFor` in standards/discrimination.ts; no re-run, no network, no spend. Cross-checked against experiments/v3-2/discrimination.json and standards/coffee/v1.0/fitness.json.

Authored expectation: no prediction, confidence low. The numeric band this standard used to carry is not published at this grammar version.

The engine has measured this row failing for about seventy-one per cent of stores in a general sample, which sits inside this band; coffee may differ in either direction because roast-to-order sellers discuss timing more than average while also being the least likely to commit to a number. The band is anchored on a real measurement of a different population, which is the strongest basis any prediction in this standard has.

The numeric fail-rate band this reasoning was authored with is NOT carried at grammar 1.2 and is not reproduced here. Authored bands held 1 of 10 against this category's own sample and every miss was HIGH, so a direction derived from the band would inherit the error that condemned it (SCHEMA.md §9.3 step 3). The band survives in ALS-COFFEE v1.1 (standard_hash f8ec2780f60c38931913e5b6cd37506500c8462709209de7180ba6691d6137e7), which is byte-frozen and still served.

Known gaps

  • "Returns are accepted within 30 business days of receipt." — false_pass
    A returns window credited as a delivery window. Pinned open, and coffee shipping policies routinely place both in the same document.
  • "We do not offer weekend pickup, or overnight shipping." — false_pass
    Negation does not reach past the comma, so the second conjunct of a denial reads as a stated delivery speed.
  • a denial, then a colon, then a delivery window — the ordinary FAQ and spec-label shape — false_pass
    A colon is an unconditional clause boundary, so any colon between the negator and the term resets the clause and the denial reads as a stated delivery speed. The corpus records this case with the assurance verb the claim linter forbids, which is why it is described here rather than quoted.

For a shopper

For coffee the handling time matters more than the courier, because a bag can sit for days waiting for the next roast. Look for a stated dispatch window with a number in it, and treat a transit estimate or a free-shipping badge as answering a different question.

For a merchant

State a dispatch window with a number and a unit on the product page or in the shipping policy, for example that orders ship within two business days, and state it separately from courier transit time.

Grounding