Not All Permit Data Is Created Equal

By Raf Howery | Kukun

The building permit data market has a dirty secret: most providers are selling the same upstream data.

A handful of original sources aggregate raw permit records from county and municipal portals across the country. Everyone else (the resellers, the data bundles, the platforms marketing “comprehensive permit coverage”) is licensing from those same upstream feeds, marking them up, and shipping them to you with minimal additional processing.

That matters for three reasons: coverage, freshness, and quality. When you’re making underwriting decisions, pricing risk, or building a product on top of permit data, the difference between a purpose-built data layer and a resold feed shows up in your outputs.

Here’s what to actually look at when evaluating permit data providers.

1. Source count determines your coverage ceiling

Raw permit data comes from thousands of individual county and city portals. No single aggregator covers all of them. The question is how many your provider has direct integrations with and how they handle the rest.

Resellers are constrained by their upstream source. If that source covers 2,700 jurisdictions, that’s your ceiling regardless of what the marketing says. You inherit their gaps.

Kukun aggregates directly from 5,600+ sources, more than double what most resellers can access through their upstream providers. That coverage gap is largest in secondary markets and rural counties, which is precisely where underwriting decisions are hardest to make from public records alone.

2. Refresh rate is not a marketing claim; it’s a data quality metric

Permit data goes stale fast. A permit issued last week that shows up in your system next month is worse than no data at all; it creates false confidence that you’ve checked when you haven’t.

Most permit data in the market refreshes on a biweekly or monthly cadence. You’re not getting current permit activity. You’re getting a snapshot that’s already 2–4 weeks old by the time it reaches you.

Kukun’s median data lag is 8 days from permit issuance to availability via API. For insurance carriers monitoring portfolio risk, 8 days versus 30 days is the difference between catching an undisclosed renovation before renewal and missing it entirely.

3. Raw labels vs. enriched labels

Raw permit data uses whatever language the issuing jurisdiction used. “Rem exist lease space for restaurant.” “Sheet metal work exhaust ventilation fiberglass ductwork.” These descriptions mean something to a building inspector. They mean nothing to a machine learning model or an underwriting rule.

Most providers apply a basic keyword match to assign a category. That’s not enrichment: that’s a find-and-replace.

Kukun applies a multi-stage processing pipeline: normalization across all 5,600+ sources, standardized label generation, and, for Permits Intelligence, real-time enhanced labeling triggered per API call, applying updated models rather than batch-processed labels from days ago.

The difference shows up downstream. A roof permit mislabeled as general construction doesn’t trigger your HVAC age flag. A kitchen remodel miscategorized as cosmetic doesn’t surface the renovation trigger signal your HELOC team is looking for.

4. What you’re actually getting from a reseller

When you buy permit data from a reseller, you get: a subset of the upstream source’s coverage, a data lag that compounds, labels applied by the upstream source’s methodology, and no recourse on quality issues that originate upstream.

The reseller cannot fix what their source didn’t capture. They cannot refresh faster than their source refreshes. And they cannot re-label data that was never correctly labeled to begin with.

5. The join: permit data in context

Permit data in isolation answers one question: what work was permitted at this address?

The more powerful question is: what does this permit mean for this property, given its age, its systems, its value, and its history?

Answering that requires joining the permit record to property-level context. That join (permit data alongside property records, valuation, system lifecycle, and condition signals) is what turns raw permit activity into actionable intelligence. It’s the difference between a data feed and an analytics layer.

Read more: Permit data for lenders the HELOC blind spot

Comparison: what to look for

→ Explore the Kukun Permits API at mykukun.com/developers 5,600+ sources · 800M+ permits · 8-day median lag · Free tier available

Raf Howery

Raf Howery is the founder and CEO of Kukun. Howery, a self-professed "serial renovator" and entrepreneur, started Kukun to transform the home remodeling industry from an offline, fragmented, and frustrating undertaking to an online, modern, and user-friendly experience.

Published September 28, 2026