HubSpot Contact Enrichment Not Working? Here's Why It's Filling In the Wrong Data

You turned on HubSpot contact enrichment (Breeze Intelligence) expecting it to quietly top up thin contact and company records. Instead, a 12-person consultancy is showing 4,000 employees. A UK manufacturer's industry field says "Financial Services." A contact you manually corrected last month has drifted back to the wrong job title. None of this throws an error. There's no failed sync to notice, no red status badge in Settings. The record simply looks a little more complete than it should - and everyone downstream, from lead scoring to segmentation, trusts it.

That's the part that makes enrichment errors harder to catch than most HubSpot data problems: a blank field gets noticed and gets fixed. A wrong field that looks plausible gets used.

↓

↓

✅ CLEAN MATCH Domain maps to exactly one company. Fields fill in correctly.

⚠️ AMBIGUOUS MATCH Shared domain, stale record, or thin coverage. Fields fill in anyway - just wrong.

Both outcomes write a value to the record. Only one of them is describing the right company.

What HubSpot's Enrichment Actually Does

Breeze Intelligence is HubSpot's built-in enrichment engine, built on the Clearbit dataset HubSpot acquired in 2023. When a new company or contact record is created - or when an existing one has gaps - Breeze checks the associated domain or email against its database and writes back company size, industry, revenue, seniority, and a range of other firmographic fields automatically, on paid plans, with no manual research required. HubSpot's own data enrichment documentation confirms this runs on HubSpot's commercial dataset rather than a source you control or can cross-check directly.

That's genuinely useful when the match is clean. The trouble starts with how "clean" a match actually needs to be before HubSpot writes the data back - which, in practice, is a lower bar than most teams assume.

HubSpot markets the underlying dataset behind Breeze Intelligence as covering over 200 million continuously updated profiles. Scale isn't the same as accuracy on any individual record, though - third-party analysis has put Breeze Intelligence's overall data quality at roughly 7 out of 10, noticeably stronger on company-level fields than on individual contact details. That gap between "large dataset" and "reliable on this specific record" is exactly where the errors in this post come from.

Why HubSpot Contact Enrichment Fills In the Wrong Data

In most portals we've looked at, the wrong data doesn't come from a single bug. It comes from a small set of matching and re-run behaviours that are working exactly as designed - just not in the way most teams assume.

• ●Domain matching doesn't respect company structure. Enrichment matches primarily on domain. A subsidiary sharing its parent's domain, or a brand operating under a franchise or platform domain, gets matched to that domain's firmographics - not its own. That's how a small regional office ends up showing the parent group's entire headcount and revenue.
• ●Contact-level matching has nothing reliable to work with on generic addresses. A personal Gmail address, a shared team inbox, or a free email domain gives the matching engine no company signal to anchor to. It either returns nothing, or resolves to the closest fuzzy match in its database - which is not the same as the right one.
• ●Breeze pulls from one proprietary dataset, not a cross-checked blend. If that record is stale - the company changed industry classification, downsized, rebranded, or was acquired - the field gets populated with a confident, specific, wrong answer rather than being left blank pending a second source.
• ●Overwrite rules don't always behave the way teams assume. By default, HubSpot's continuous enrichment is designed to stop updating a property once a person or another system has edited it - a safeguard meant to protect manual corrections. But that protection only holds if overwrite rules are actually set to "fill blanks only." If a property mapping is left on a hard-overwrite setting, or enrichment is re-triggered manually rather than through the passive monthly cycle, a corrected field can still be replaced with the old enriched value the next time HubSpot's contact enrichment runs on that record.
• ●Company-level and contact-level enrichment don't always reconcile. Breeze can enrich a Company record correctly while the associated Contact record keeps an older or mismatched value in the same property, because the two objects aren't guaranteed to re-sync together on the same pass.
• ●Coverage is uneven by company size, industry, and region. Enrichment accuracy is strongest for larger, US-based tech companies and noticeably weaker for SMBs, non-tech industries, and international records. Outside that core segment, a real proportion of "enriched" fields are a best-guess fill from thin source data rather than an honest blank.

The overwrite behaviour above isn't hidden - it's a visible setting on every property mapping, sitting one screen away from where enrichment gets configured:

Each property mapping carries its own overwrite rule - "Fill empty values only," "Fill empty values and overwrite existing values," or "Do not fill any values." Nothing stops adjacent properties from being set differently, or from being left on a setting nobody revisited after initial setup.

Most teams only discover these patterns after enrichment has already run across the full database. HubSpot lets you run a data test on a sample of 50 to 100 records first, to check match rates and flag obviously wrong matches before committing credits - and before HubSpot's contact enrichment reaches thousands of records nobody is going to review line by line. It's a five-minute check that catches most of the domain and coverage problems above before they spread.

Where Does the Wrong Data Usually Show Up?

Check Your Own Records in 5 Minutes

You don't need a full audit to get a first read on whether this is happening in your own portal. Pick 10 recently enriched Company records and work through this:

1	Open the Industry and Employee count properties on each record and check the property history for a Breeze Intelligence entry. Note the date - a change from this month is worth double-checking before you trust it.
2	Compare the domain against the company's actual registered entity - is this a subsidiary, franchise, or brand that shares a domain with a larger parent? That's the single most common source of an inflated employee count or revenue figure.
3	Open the Contact records associated with those same 10 companies and compare the same property values. If the Company record looks right but the Contact record doesn't, you've found the reconciliation gap.
4	If you find a field that was manually corrected before, check whether it's reverted. If it has, look at the overwrite rule on that specific property mapping - it's likely set to overwrite rather than fill blanks only.

If more than two or three of those 10 records show a problem, the issue isn't a handful of bad matches - it's a pattern worth investigating at scale before it spreads to the rest of the database.

This is the same underlying pattern we've seen play out with HubSpot's Prospecting Agent recommending the wrong leads and with a Shopify sync that looks "On" while most of the data quietly fails to arrive - a feature that reports success while the data underneath it is incomplete or mismatched. We've written before about how dirty CRM data undermines every AI feature built on top of it; enrichment filling in the wrong data is simply the earliest point in that chain, because it's often the first automated process to touch a new record.

DIY or Bring in a Specialist?

How Loncom Approaches Bad Enrichment Data

We treat enrichment errors as a matching and governance problem first, not a "turn the feature off" problem. That starts with pulling property history on the affected fields to see exactly when Breeze last wrote to them, then checking whether the match came from a domain that actually represents the record in question or from a parent, franchise, or shared address sitting behind it.

Where a first look at property history isn't enough to tell you how deep the problem runs, our HubSpot CRM Audit gives you a full picture of which fields enrichment has touched and how badly, before any cleanup work starts. Where the data itself needs work before enrichment can be trusted again, our HubSpot CRM Data Cleanup process standardises and deduplicates records first, so enrichment isn't reinforcing a mismatch that already existed. Where the fix needs to hold going forward - excluding specific domains, protecting manually-verified fields from being overwritten, reconciling Company and Contact values automatically - that's typically a custom validation layer built on top of the native feature, paired with proper field validations so a bad match can't quietly slip back in later, rather than a one-off cleanup. It's the same matching-and-governance approach behind the 140+ CRM and ERP implementations we've delivered for growing B2B businesses.

The goal isn't zero automation. It's making sure the fields your sales and marketing teams build segments, scores, and reports on are ones you'd actually trust if someone asked you to defend the number.

In Short

• HubSpot's contact enrichment (Breeze Intelligence) fills fields by matching a domain or email against a single third-party dataset - when that match is ambiguous or stale, it writes back a confident wrong answer instead of leaving the field blank.
• The most common causes are domain matching that ignores company structure, single-source data with no cross-check, scheduled re-enrichment overwriting manual corrections, and Company/Contact records enriching independently.
• Coverage is uneven: accuracy is strongest for larger US-based tech companies and weaker for SMBs, non-tech industries, and international records.
• A wrong field is harder to catch than a blank one - which is exactly why HubSpot contact enrichment needs a manual check built into your process, not blind trust in whatever it fills in.

FAQ

What is HubSpot's contact enrichment feature called?

It's called Breeze Intelligence. It's built on the Clearbit dataset, which HubSpot acquired in 2023 and rebranded, and it automatically fills contact and company properties like industry, employee count, revenue, and seniority based on a record's domain or email. When people say HubSpot's contact enrichment "isn't working," they usually mean the fields are filling in, just not correctly.

Why does HubSpot enrichment fill in the wrong industry or employee count?

Usually because the domain it matched against doesn't map cleanly to the specific company - a subsidiary sharing a parent domain, a franchise on a shared platform domain, or a stale entry in the underlying third-party database. The matching engine still returns a confident value even when the match itself is weak.

Can enrichment overwrite data I already corrected manually?

By default, no - HubSpot's continuous enrichment is built to stop touching a property once a person has edited it. But that protection depends on overwrite rules being set to "fill blanks only." If a property mapping is left on hard-overwrite, or if enrichment is re-triggered manually rather than through the passive monthly cycle, a corrected field can still be replaced with the old enriched value.

Why does my Company record look right but the Contact record doesn't?

Company-level and contact-level enrichment can run independently and don't always reconcile the same property across both associated records. HubSpot's own community has reported this same gap - correct data landing on the Company record without appending to the linked Contact.

Does contact enrichment work the same for every company size and region?

No. Coverage is strongest for larger, US-based tech companies and weaker for SMBs, non-tech industries, and international records. Outside that core segment, expect more inconsistent fills and more fields left as "Unknown" rather than a confident value.

Does bad enrichment data affect lead scoring and segmentation?

Yes, directly. Lead scoring models and segment lists typically run on the same Industry, Employee Count, and Revenue fields that enrichment writes to. A wrong value doesn't just sit there quietly - it can push a record into the wrong tier, route it to the wrong rep, or exclude it from a campaign it should have qualified for, without anyone seeing an error to trace it back to.

How do I check which fields Breeze Intelligence has changed on a record?

Open the property on the record and view its history - each change lists a source. An entry attributed to Breeze Intelligence tells you enrichment wrote that value, and when, so you can compare it against what you know to be accurate.

Can Loncom help fix bad enrichment data in our HubSpot portal?

Yes. We audit which fields enrichment has affected, clean and deduplicate the underlying records where the data itself is the problem, and build validation or exclusion rules so corrected fields stay corrected going forward.