← ERateSignal
Methodology

How the data works.

ERateSignal's E-Rate filing and funding data is built on opendata.usac.org— the same public dataset anyone can query. We don't scrape the E-Rate Productivity Center. We don't use authenticated data. We don't have an inside channel. Our job is to make the public record fast, structured, and scannable for sellers working the program.

Sources

Where every record comes from.

We ingest the USAC Socrata Open Data API directly. The datasets we rely on:

  • jt8s-3q52 — Form 470 filings (opportunities)
  • avi8-svp9 — Form 471 commitments
  • 59r2-zbdq — Entity / BEN registry
  • jpiu-tj8h — Invoices and disbursements

Every raw API response is archived to object storage before any transformation. If USAC publishes it, we can show you the original payload we ingested and the timestamp we ingested it.

Some non-E-Rate enrichment fields may come from other public government datasets, such as SAM.gov, GSA public exclusion data, or NCES records. Those fields are treated as enrichment, not as USAC filing data, and should be labeled separately in the product.

Refresh cadence

How fresh the data is.

Form 470 filings are polled continuously during the filing window. New filings typically appear in ERateSignal within minutes of USAC publishing them. Form 471 commitments and FRN status changes are polled against the USAC :updated_at field so we only pull records that actually changed.

We rate-limit ourselves to 50 requests per hour and back off exponentially on errors. This means we're a polite tenant on the USAC API — and it means our data is as fresh as USAC publishes, never fresher.

Scoring

How the 0–100 deal score is built.

Deal scoring is a transparent rubric, not a black box. Each filing is scored against three input groups:

  1. Territory fit — your states, your district size range, your category focus.
  2. Product-line fit — whether the filing's service request matches the OEMs and product lines you sell.
  3. Filing characteristics — deadline, value tier, competitive bid count, prior vendor history.

Hover any score in the app and you'll see the three factors that drove it. No hidden weights, no model drift surprises.

Classification

How service requests get grouped.

USAC service descriptions are free-text. We classify each FRN line item into a product line taxonomy (firewalls, switches, wireless, ISP, cabling, etc.) and detect the OEM when one is named. Classification is rule-based today, with an ML pass planned as the training corpus grows. The classifier's output is stored alongside the raw text so you can always see both.

Compliance

What we will not do.

  • We do not scrape or redistribute data from the E-Rate Productivity Center (EPC). EPC data is authenticated and non-public.
  • We do not sell harvested contact lists or send bulk unsolicited mail to applicants.
  • We do not show bid prices or any information USAC has not yet published. No hidden-data layer — just public record.
  • We do not build features that help a service provider prepare an applicant's Form 470 — this is prohibited under 47 CFR § 54.503.

Our design rule: if a feature could be construed as providing advance or non-public information in the competitive bidding process, it does not ship.

Open questions

What we're still building.

ERateSignal is pre-launch. Deal scoring weights, product-line taxonomy, and territory analytics will evolve as founding members use the product against live filing windows. If you want a direct line into those decisions, join the founding cohort — that's literally what it buys you.

Join the waitlist →

ERateSignal is not affiliated with, endorsed by, or sponsored by USAC, the FCC, or the U.S. Government. All E-Rate program data displayed on this platform is sourced exclusively from publicly available datasets published by USAC on opendata.usac.org. Data is provided "as-is" — verify all information directly with USAC before making business decisions.