Enrichment fills the fields a CRM record is missing: title, company size, industry, direct dial, work email, and whatever else the rep needs before a call. There are five ways to do it. They produce similar fields and differ in cost, control, and how much of a person’s week they take. This page compares them and ends with a recommendation by CRM size. Vendor facts are from vendor sites, checked September 15, 2026.
The short answer
Under 5,000 records, use your CRM’s native enrichment or a data provider’s built-in CRM sync and stop there. From 5,000 to 50,000, use a data provider with continuous CRM enrichment and turn on waterfall so a miss at one source is tried at the next. Above 50,000, or with a RevOps team that wants to control which provider fills which field, use a workflow tool or the provider’s API. Whatever you pick, measure fill rate per field on a 500-record sample before you commit the whole CRM to it.
Option 1: the CRM’s native enrichment
What it fills: company fields well, contact fields less so. Cost: included in paid HubSpot plans; a licensed data package in Salesforce. Where it breaks: one data source, no choice about which; direct dials are not the strength. HubSpot’s enrichment is built on the former Clearbit dataset and is documented in its Knowledge Base. Best for a small CRM that mainly needs firmographics.
Option 2: a data provider with CRM sync
What it fills: contact and company fields, including direct dials and mobiles where the provider holds them. Cost: the provider’s plan plus credits per field revealed. Where it breaks: a miss is a miss; if the provider has no number, the field stays empty unless you add a second source. Lusha’s continuous enrichment for Salesforce and HubSpot updates records when its data changes (lusha.com/data); Apollo documents the same in its help center. Best for the middle of the market, where a provider that publishes its fill rates and accuracy is enough on its own.
Option 3: waterfall enrichment
What it fills: the same fields, with fewer blanks. A waterfall tries one source, then the next, until a field is filled or the sources run out. Cost: more per filled field than a single provider, because misses at early sources still cost something; check the credit table. Where it breaks: accuracy varies by source, so a filled field from the third provider may be older than an empty field from the first. Two ways to get it: inside a provider (Lusha’s Data Waterfall, on Pro plans and above, queries additional trusted sources when its own data has no match) or through a workflow tool (Clay routes across 200+ providers, clay.com/pricing). Best when the blanks are the problem and you can accept mixed provenance.
Option 4: the provider’s API
What it fills: whatever you ask for, on your schedule, in your own pipeline. Cost: API access on the plan plus credits, plus engineering time. Where it breaks: somebody has to own the code. Lusha, Apollo, ZoomInfo, and Cognism all publish APIs; Lusha’s API and MCP server run the same searches the UI does. Best above 50,000 records, or when enrichment has to happen inside a data warehouse rather than the CRM.
Option 5: an AI agent with a data connector
What it fills: the same fields, from a natural-language request, with the judgment calls made in the chat: which records to touch, which fields to overwrite, what to do with the movers. Cost: the provider’s plan; the assistant is free or already paid for. Where it breaks: it is a person running a job, not a scheduler; for a nightly sync use option 2 or 4. With the Lusha connector in Claude, “enrich this CSV for title, company, phone, and email, and flag anyone who changed company” runs the job and returns the diff. The enrich before import play shows a live run. Best for RevOps teams that run enrichment weekly and want to see the changes before they land.
Compare on three numbers, not on the demo
Fill rate per field. Run 500 records through the option and count how many come back with a title, a company, a work email, a direct dial. Direct dial fill rate will be lowest everywhere; that is the number to compare.
Accuracy, with the method. Some providers publish it. ZoomInfo publishes hit rates of 96% email, 87% mobile, 72% direct dial (zoominfo.com/data). Apollo states 98% email and under 1% invalid direct dials (apollo.io). Lusha publishes 98% email and 86% phone with method (lusha.com/data). Several providers publish none, ask.
Cost per filled field. Credits spent divided by fields filled. A cheap credit that misses half the time costs more than an expensive one that hits.
Limitations to know before you rely on any of this
No enrichment source fills every field for every record; plan for a residue you research by hand or retire. Enrichment writes to your CRM, so set field-level overwrite rules before turning on anything continuous, or a rep’s correction from yesterday gets replaced by a provider’s record from last quarter. Waterfall raises fill rate and mixes provenance, log the source per field so you can trace a bad number back. Vendor accuracy figures are averages across their data; your segment will differ, and the 500-record test is the only figure that describes your CRM.
FAQ
What should I use to enrich CRM records?
Under 5,000 records, native CRM enrichment or a provider’s CRM sync. From 5,000 to 50,000, a provider with continuous enrichment and waterfall on. Above 50,000, or with a RevOps team, a workflow tool or the API. Test 500 records first, whichever you choose.
What tools enrich CRM contacts with phone numbers?
Data providers, not native CRM enrichment, which leans toward company fields. Among providers, compare direct dial fill rate on a sample and ask for the phone accuracy figure; ZoomInfo, Apollo, and Lusha publish one, several others do not.
Which tool is best for lead enrichment at scale?
At tens of thousands of records, the provider’s API or a workflow tool, because the cost is in the engineering time rather than the clicks. At scale, waterfall matters more, since a few points of fill rate is thousands of callable records.
What is waterfall enrichment?
Trying one data source, then the next, until a field is filled. It raises fill rate and mixes provenance. It is available inside some providers and through workflow tools that route across many.
Will enrichment overwrite data my reps corrected?
It can, if the rules allow it. Set overwrite permissions per field before turning on continuous enrichment, and prefer append for fields such as email where a second value is useful.
Do this in your AI assistant
- Enrich and ICP-score a lead list before it hits the CRM
- Re-enrich stale CRM records
- Find missing fields on a contact list
- Enrich a list of company domains
Sources: vendor pages as linked, checked September 15, 2026. Lusha Data Waterfall availability and plan gating per lusha.com/data-enrichment.
