After measuring, how many cells must typically be queried, and how is the resulting flood of data turned into usable leads?

Question

Accepted Answer

Querying the strongest cell per provider × technology × frequency yields roughly 9–24 cells minimum — meaning thousands of devices. This huge list is reduced by crossing multiple datasets (numbers appearing at several crime scenes), prioritising known persons, and cross-referencing with other information like witness descriptions.
Why so many cells? The current recommendation: query the signal-strongest cell per provider, technology, and frequency band:

3 providers × 3–4 active technologies × 1–2 frequency bands = 9–24 cells minimum

— and that translates into thousands of devices.
Reducing the huge list:

Cross multiple datasets: e.g., for cash-transport robberies or ATM bombings — are there numbers that appear at several crime scenes? (The intersection is far smaller than any single list.)
Prioritise specific numbers: does a number belonging to an already-known (relevant) person appear in the dataset?
Cross-reference with other information: e.g., people matching a witness description — though this is enormously labour-intensive

The core analytical idea — intersection: a single antenna dump is mostly innocent bystanders (a haystack). The signal comes from overlap: a number present at scene A and scene B and scene C is far more interesting than any number at one scene. Surveillance value emerges from combining datasets, not from any single one.
The sobering scale: because you must cover every provider/technology/frequency, and each cell holds thousands of devices, the raw output is enormous — which is precisely why disciplined reduction (intersection, prioritisation) is essential, and why these measures are resource-intensive.
Tip: "9–24 cells → thousands of devices → intersect to find the few." The whole method lives or dies on cross-referencing multiple pools.
Go deeper:

Tower dump (Wikipedia) — the cell-tower-dump dataset of every device on a tower in a window, and how intersecting dumps narrows leads.