How much time does a listing leave you?

Measure the listing, not the buyer.

This notebook follows a configuration at a store from its first recorded return to the next unavailable observation. The result describes visibility in the source feed. It cannot time a customer’s checkout or show when the physical phones arrived.

Read a range such as 45–60 seconds as two observations: still available at the lower bound and unavailable by the upper bound, measured from the first recorded return. The summary uses upper bounds consistently. A single displayed value is still limited by source observation precision.

Rows marked approximately use the historical 15-second fallback bound because a last-available timestamp was not recorded. Initial sightings do not establish a return, and open intervals have not yet supplied a completed duration. Both are excluded from the completed sample.

Inspect a few intervals before trusting a summary.

The ledger keeps store, configuration and timing together. A row that remains listed has an open clock; it does not prove the listing will remain available when you open Apple. The clock is based on the recorded feed, which can arrive after collection and synchronization delay.

Use your acceptable configurations as the comparison set.

Read the completed-row count before comparing medians. 4 of 8 configurations in the detailed sample reach the 10-row display threshold. A lower count is marked thin; it is still shown so that a missing comparison is not mistaken for a zero-second listing.

A longer median may be interesting if you already want that configuration. It is not evidence that the next listing will last longer, and it should not override your storage needs or budget. Return frequency is a separate measure in the configuration notebook.

Check whose observations a location comparison includes.

A state with more rows has more representation in this sample, not necessarily better access to stock. Its stores, configurations and observation gaps may differ from another state’s. Use the links to inspect a location rather than converting a national ranking into travel advice.

The imported US record begins on October 29, 2026; the selected international records begin on November 2, 2026. Those are dataset boundaries, not ShelfSpring operating dates. Different coverage periods and sparse groups limit direct country comparisons.

A timing experiment, with its assumptions visible.

Imagine checking at a fixed interval while a listing’s start is uniformly positioned between checks. A duration of d seconds and an interval of T seconds give a modeled chance of observation of min(1, d ÷ T). The values below average that model over the separate seven-day sample across all tracked countries; they are not measured delivery rates.

Real collection does not follow every assumption of that experiment. A listing can fall entirely between source checks and leave no recorded interval. This selection effect means the observed distribution cannot establish the distribution of every listing, including missed ones.

Take preparation from the chart, not a countdown.

Decide on acceptable configurations and reachable stores before you need to evaluate a match. When one appears, inspect the current Apple offer. Neither a short median nor a long interval tells you whether that particular order will succeed.

A notification adds another timing stage after the original observation. All ShelfSpring plans share the inventory feed; a higher rule limit does not buy a faster source. Use your account to confirm enabled matches and delivery settings.