# Method: how these Texas license figures are built

> Source: PermitPulled, https://permitpulled.com/method
> Compiled from the public TDLR license file. Published under CC BY 4.0; attribution to PermitPulled required.
> For an individual license record, use the official TDLR search: https://www.tdlr.texas.gov/LicenseSearch/

[Texas license lookup](</>) / Method

# How these figures are built

Everything on this site comes from one public file. This page explains exactly what we do to it, what we leave out and where the numbers can be wrong, so you can decide how much weight to give them.

## The source

TDLR publishes a file of every license it issues, refreshed daily, as part of its public license data. We use the version dated **20 September 2026**. Nothing on this site comes from scraping the TDLR website, and nothing comes from a third-party broker.

## What we keep

From each row we keep the license type, license number, holder name, business name, city, county, license class and expiration date. We use those to produce counts. The source file also contains mailing addresses and phone numbers, which we discard at import and never store.

## What we count

  * **Active** means the expiration date has not yet passed on the date of the snapshot. It is a date comparison, not a status field, so a license suspended by an enforcement action can still be counted as active here.
  * **Renewals in the next 12 months** are licenses whose expiration date falls inside the twelve months after the snapshot.
  * **Per month** figures are the twelve-month total divided by twelve, not the count for any particular month.

## What we exclude, and why

The source file contains about **51,000 secondary rows with no license type**. Roughly 31,000 of them duplicate a row that does have a type, and the remaining 20,000 cannot be classified. All of them are excluded, which is why our record count is 946,223 rather than the raw line count of the file.

We also suppress small cells: a county is not shown for a license type when fewer than five licenses of that type are active there. With one or two holders, a county count can point at a specific person even though no individual record is published.

## Where these numbers can differ from TDLR's

  * **Definition.** TDLR may count as active only licenses in good standing; we count by expiration date.
  * **Timing.** The file is a daily snapshot. A renewal processed after the snapshot is not reflected until the next refresh.
  * **Classification.** Where the file gives a license class or subtype, we use it as published; we do not reinterpret it.

For any individual license, TDLR is the official record and should be checked directly.

## How often it refreshes

The statistics are rebuilt from a new copy of the file periodically, and every page prints the date of the file it is built from. If a page shows an old date, the figure is old: that is the point of printing it.

## Reuse

The compiled dataset is published under [CC BY 4.0](<https://creativecommons.org/licenses/by/4.0/>). You may republish it, including commercially, with attribution to PermitPulled. The [CSV](</downloads/texas-license-renewals.csv>) carries the same figures as the [renewals report](</reports/texas-license-renewals>).

If you need a cut that is not published, such as a single county, a single trade or a longer series, write to [info@permitpulled.com](<mailto:info@permitpulled.com>) and we will prepare it.

## Corrections

Errors get fixed and the fix is dated. If you can show that a figure or a rule is wrong, send it over: it is the most useful message this site can receive.