ELDHURSTLLC

Editorial standards

These commitments bind every Eldhurst brand. They are published so readers — and anyone doing business with us — can hold us to them.

Sourcing

Every statistic originates in a named public dataset and carries its as-of date. We do not publish estimates dressed as measurements, and we label the difference.

Restraint

Publication is earned. Where the public record is too thin to say something useful about a place or product, our brands route readers to broader guidance instead of publishing thin pages.

Methodology

Our brand sites link each published figure to a methodology record — the dataset, the definition, the time window, and how to check the number against the public source yourself.

Disclosure

Where a brand earns a referral or connection fee, that relationship is disclosed next to the action it relates to — plain language, readable size, every page where it applies. Providers reached through our brands are independent businesses; we say so, and we do not claim vetting we have not performed.

Corrections

Errors get fixed, and the fix gets dated. Reach us at support@eldhurst.com for brand content or contact@eldhurst.com for anything else.

The corrections log

How one number proves itself

Commitment 03, worked through on a figure printed today on a Countyguard county guide.

  1. Step one · what the reader sees

    “14 federal disaster declarations naming Harris County since 2015 (all incident types)”

    Harris County guide — the figure carries a “method” link beside it.

  2. Step two · the record behind it

    Source dataset
    FEMA — OpenFEMA Disaster Declarations Summaries v2
    Window
    2015-01 (registered start) through 2026-07
    Date semantics
    declaration issuance month, not incident month
    Counting rule
    county-designated rows only, counted once per distinct disaster number
    Rounding
    none — an integer count

  3. Step three · check it against the public source

    The record prints the filter to run against FEMA’s public interface and the count to take from the result, so the figure can be recomputed without us. It also prints the number that would have flattered us: an all-time count would show 41 — the window is why they differ. 41 is the bigger number; 14 is the one the window supports, so 14 is what the page says.

What this rules out

Commitments are easy to write and hard to check. These are the specific things they forbid us from doing, stated so that a reader or a partner can tell if we have broken one.

  • We do not publish lead forms. Our brands print a number; the reader decides whether to dial it.
  • We do not sell or share reader contact information with providers — we do not collect it.
  • We do not publish rankings, ratings or “best of” lists that a business can pay to enter.
  • We do not claim to screen, vet, license-check or endorse the providers a reader may reach.
  • We do not publish a page about a place merely because the place exists.
  • We do not present an estimate as a measurement, or a model output as a record.