TL;DR
- HubSpot data enrichment is the first option to evaluate if the company records already live in HubSpot and the available properties meet the job. HubSpot says its standard contact and company enrichment does not consume HubSpot Credits, though eligible subscriptions and permissions still apply. Smart properties are a separate, credit-consuming feature.
- People Data Labs suits an API-led product that needs a company record keyed by domain and is prepared to own field mapping, matching and permitted use. Its self-serve company pricing uses credits per successful result, with separate monthly and annual tiers.
- Apollo is a practical option when the same team also wants its prospecting and sales workspace. API enrichment, endpoint credit costs and any rights to expose the resulting data in another product must be assessed separately.
- Crustdata deserves a pilot when company and people data need to feed an API-driven application. Its credits vary by endpoint and field; indexed results and real-time requests should not be priced as the same operation.
- Dropcontact is a narrower candidate where contact enrichment and company context are intertwined, particularly where French company identifiers matter. It should not be assumed to replace every company-only Clearbit endpoint or field.
- The hard part of replacing Clearbit is preserving the field contract and write rules. A company name, employee band or technology field that looks similar may use a different entity, source or refresh cycle. Test the output before allowing it to overwrite the CRM.
What is actually replacing Clearbit?
A Clearbit integration may sit in several places: an inbound form, a CRM enrichment rule, a product's account record or an internal data pipeline. The visible outcome may be one company name and employee range, but the application underneath can depend on a domain match, a specific response schema, an overwrite rule and a source contract.
Clearbit became part of HubSpot's enrichment offering. That does not make every legacy Clearbit API integration interchangeable with a HubSpot CRM feature. Identify the exact endpoint, fields, contracts and downstream systems your organisation uses before assuming an old workflow will continue or end on a particular date.
For inbound qualification, the first question is whether the CRM is the system of record. If a form creates a HubSpot company and the team only needs supported HubSpot properties, native enrichment may remove an integration. If a software product needs to return company data to its own customers through an external API, a CRM property update is a different architecture and the data-use agreement matters.
Our Clay, Clearbit and Apollo comparison examines the broader platform decision. This guide begins with a live replacement problem: what must happen to the existing fields, rules and data flows?
Five routes at a glance
| Option | Starting input and output | Buying unit or entitlement | Strongest fit | Limit to investigate |
|---|---|---|---|---|
| HubSpot | Company domain into CRM properties | Eligible subscription and permissions; standard enrichment documented without HubSpot Credits | HubSpot-owned company records | Does not by itself establish external API or redistribution parity |
| People Data Labs | Domain or other company identifiers into a structured API record | Successful enrichment credits and purchased tier | An application owning an API integration | Fields and data licence must match the product's use |
| Apollo | Company and contact data in a sales workspace and API | Plan, endpoint and credits | Prospecting plus CRM workflow | API credits and customer-facing rights vary |
| Crustdata | Company API and data fields | Endpoint, fields and credits | API-first company research | Indexed versus live data and contract eligibility |
| Dropcontact | Contact plus company context; limited company-only paths | Credits and selected plan | Contact workflows and French entity data | Not a universal company-only Clearbit clone |
These are different routes to a business outcome, not interchangeable records. We did not run a vendor coverage or accuracy benchmark. Use the same field acceptance test and comparable cost boundary before making a purchase.
1. HubSpot: enrich the CRM record
If HubSpot already owns the company, start with its native data enrichment rather than automatically buying another API. HubSpot's documentation says company enrichment uses the company domain to match business information from its commercial dataset, third-party sources and public information. An administrator can configure when enrichment occurs and which users have access.
The current documentation states that standard data enrichment does not consume HubSpot Credits. That is not the same as “every HubSpot enrichment-related feature is free”. HubSpot lists Starter or higher for several Hubs, but Professional or Enterprise for Smart CRM and Revenue Hub. A super admin must configure the enrichment settings and manage access; individual users need the appropriate enrichment permissions. Smart properties are a separate credit-consuming feature, with different subscription availability. Confirm the account's edition and the actual feature before estimating a cost.
The default write policy matters as much as the source. HubSpot says existing values are retained unless a user chooses an overwrite during manual enrichment or enables the setting to overwrite incorrect enrichment values. Do not turn on broader overwrites merely to make the test show more filled properties. First decide which seller-confirmed values should be protected.
For an exit or privacy review, distinguish stopping enrichment from removing previously enriched data. HubSpot says turning the feature off stops future activity but does not remove information already added to the CRM. Inventory those properties and agree any required retention or removal process separately.
HubSpot is strongest where the output is a CRM property and the team wants record ownership and enrichment in one system. It is a less direct substitute for a legacy Clearbit API called by an independent application. Confirm supported API access, rate limits and the permitted downstream use of the CRM data for that design.
2. People Data Labs: own the company API contract
People Data Labs is relevant when the application wants a structured company record through an API and can own the mapping into its own schema. A domain is an important matching input, but the application should still validate that the returned entity is the intended operating company rather than a parent or a similarly named business.
Its pricing and credits guide describes company enrichment and search as separate API operations. The guide says company enrichment typically uses one credit for a successful response. Its published self-serve monthly company tier starts at 1,000 credits for $100, with volume tiers above that. Annual commitments use different rates and quantities.
The advantage is a clear API boundary. You can version the adapter, retain your internal account ID and decide exactly which fields the CRM may receive. That flexibility comes with responsibility for retries, data freshness, reconciliation and licensing. Buying API access does not automatically permit redistribution or indefinite retention of every field.
Run a representative sample against the exact Clearbit fields being replaced. If your routing rule depends on company size band, compare ranges and evidence dates rather than assuming two providers calculate headcount in the same way. Preserve an unknown result as unknown, not zero.
3. Apollo: company data in a prospecting workflow
Apollo is worth considering when enrichment is one step in a wider prospecting and sales process. The same account may need a company record, people at that company, a list and an engagement workflow. That can make a consolidated sales workspace attractive, provided its data and export terms fit the application's job.
Apollo's API documentation says API access is available across plans, while endpoint rate limits and credit use vary. Its credit guidance distinguishes enrichment, exports and other actions. Ask which company endpoint, field set and plan apply to your exact workload; do not use a generic per-contact price for a company-only API job.
Apollo is a good candidate for a revenue team that wants account research and action in the same environment. It is not automatically a like-for-like substitute for a legacy company data API embedded in another product. Compare object coverage, permissions, customer-facing display and export rights before treating the API as a replacement.
A useful trial follows a single inbound form through identification, company matching and CRM writeback. Check which system remains authoritative when the domain already belongs to a customer, and prevent a newly returned estimate from replacing a verified account field without a rule.
4. Crustdata: API-first company research
Crustdata offers company and people data through API endpoints. It is a candidate when enrichment is part of an application or research pipeline and the team wants to inspect the specific endpoint and field costs. Its company API documentation provides a starting point for the response contract; its enrichment overview discusses company and people endpoints and credit-based pricing. Confirm current indexed and live endpoint availability with the vendor.
Do not collapse those modes into a single “cost per company”. Determine whether the application needs an indexed record that may be refreshed periodically or a live research operation, and whether the required fields are available on the selected plan. Search, base enrichment and extra fields can consume different numbers of credits. Obtain the dollar price and expiry for the actual credit pack before modelling the workload.
Crustdata is strongest as a pilot when your team owns an integration and needs company context alongside people or event data. It is less direct for a nontechnical HubSpot operator who only needs a few native company properties. Budget the adapter and review process needed to turn the response into a safe CRM write.
5. Dropcontact: contact and company context
Dropcontact's enrichment flow begins with contact and company context. Its API documentation describes submitting batches of up to 250 contacts and retrieving the result asynchronously by request ID. Its quickstart shows contact output alongside company fields, including French company identifiers where available. A 5,000-contact submission would therefore need at least 20 full batches, before retries or smaller batches; that is a request-count calculation, not a price estimate.
That makes it useful when the inbound workflow starts with a named person and work email, and the team wants both contact and company context. It is not a general promise that a bare domain produces the same broad company record as Clearbit. Dropcontact's documentation describes limitations for organisation-only inputs and notes that enrichment outcomes depend on identifiable data. Test the exact country, input and required fields.
For a French operating entity, SIREN or SIRET can help distinguish the legal entity being enriched. For an international account, validate that the available fields support your routing decision before assuming the same match behaviour. If a request is asynchronous, keep the request ID and retrieve the existing result rather than starting a duplicate paid job after a timeout.
The plan, credits, applicable field eligibility and API rate limits need a current quote or documented configuration for the intended volume. Do not treat one found email as a complete replacement for Clearbit company enrichment.
What should a 5,000-form workflow cost?
Consider a constructed B2B funnel with 5,000 form submissions in a month. Suppose those submissions represent 2,000 unique company domains, because several contacts can come from the same company. A team that enriches on every form submission sends 5,000 company lookups. A team that safely reuses a sufficiently fresh company result may need 2,000 lookups before exceptions and refreshes.
| Stage | No reuse | Illustrative reuse case | What to check |
|---|---|---|---|
| Form submissions | 5,000 | 5,000 | Count people separately from companies |
| Company lookups | 5,000 | 2,000 | Only reuse a result where its freshness and licence permit it |
| Hypothetical successful lookups | 4,000 | 1,600 | Assumes an 80% match rate for arithmetic, not a measured vendor outcome; repeated lookups need not represent distinct companies |
| PDL matched-result credits | 4,000 | 1,600 | One company enrichment credit per successful match under the documented model |
| Illustrative PDL cost at $0.10/credit | $400 | $160 | Before subscription details, tax, staff time or other services |
The $0.10 rate uses PDL's published self-serve monthly company tier for 1,000–10,000 purchased credits. Under the example's assumptions, 4,000 credits cost $400 and 1,600 cost $160. A different match rate, purchase quantity, annual commitment or contract changes those amounts. The $240 difference is an arithmetic scenario, not a claim about actual savings or licence rights.
HubSpot's documented standard enrichment does not consume HubSpot Credits, but the applicable subscription and implementation are not free merely because the incremental credit count is zero. If the company already pays for an eligible edition, distinguish the existing subscription allocation from any new cost required to run the workflow. Smart properties and other usage-based features can have separate credit costs.
For Apollo, Crustdata and Dropcontact, obtain the current plan, credit price, field charges, minimum commitment and permitted reuse terms for this exact job. The inputs may differ: a domain-only company request is not necessarily equivalent to a named-contact enrichment. Include any integration, exception review and ongoing operations cost before comparing the result with the PDL model.
How do you replace the fields safely?
Create a field map from the old payload to the destination before choosing a vendor. The example below is an implementation design, not a claim that every provider returns each field.
| Existing business need | New response or destination | Rule to preserve |
|---|---|---|
| Company identity | Domain, provider company ID and internal CRM ID | Retain the internal ID; do not merge parent and subsidiary solely by a similar name |
| Company name | CRM company-name property | Keep a seller-confirmed value unless an authorised correction is reviewed |
| Employee size | Headcount or employee band | Store the unit and observation date; do not mix an exact count with a broad range silently |
| Industry | Provider category mapped to internal taxonomy | Preserve the original value alongside the mapped segment |
| Location | Headquarters and operating-region fields | Do not substitute headquarters for a buyer's operating office without a rule |
| Technology | Technology list or observation | Treat a missing detection as unknown, not proof that the software is absent |
| Match status | Provider result and internal decision | Keep unmatched, ambiguous and accepted states separate |
A form submission should resolve to an existing company where appropriate. If the domain maps to a parent but the opportunity belongs to a subsidiary, send it for review rather than overwriting the account automatically. If the CRM already has a customer-confirmed size, store the enrichment as a proposed update with its source and date.
Treat the domain as a lookup key, not as the account's permanent identity. Normalise the submitted domain, check it against existing company aliases and keep the original input. A shared domain can represent several operating entities; a domain change can also leave the same customer with two records. When identity remains ambiguous, attach the submission to an exception queue instead of creating or merging a company automatically.
For repeated submissions arriving close together, give the lookup an internal request key and reuse an in-progress result where the provider's terms permit it. Record whether a lookup was attempted, returned no match, failed or produced a result awaiting review. A timeout is not evidence of an empty company record: check the request status where supported before retrying, and keep the existing CRM fields until a valid response passes the write rules. Measure exceptions and duplicate requests alongside credit consumption during the pilot.
The Clay waterfall enrichment guide is useful for designing fallback. The same principle applies without Clay: request a second source only when the field remains unresolved and its licence and budget permit the additional lookup.
A 20-record migration fixture
Before cutover, prepare 20 synthetic or authorised test records and apply the same acceptance rules to the old and candidate workflows. This is a proposed test, not an executed vendor comparison.
- Five known domains: existing customer accounts with verified company identity and a protected CRM value.
- Five repeated domains: multiple contacts from one company, used to test caching, duplicate company creation and contact-to-company association.
- Four parent/subsidiary cases: related brands or domains where the operating entity matters to routing.
- Three unknown or personal domains: no reliable company match, where the correct result may be an explicit unknown.
- Three stale or conflicting records: an old employer or company size conflicting with a customer-confirmed CRM field.
For each case, record input, returned entity, required-field completeness, source and observation date when available, proposed write, actual test outcome and cost. Keep the original CRM value and the decision. If the provider does not supply a field, mark it missing; do not fill the gap with a guessed value just to demonstrate parity.
Then rehearse a rollback: disable automatic writes, restore any synthetic changes and reconcile which records were touched. Do not declare the migration complete until the exception queue, ownership rules and data-use terms have owners.
Which route should you choose?
Start with HubSpot if the supported CRM properties, subscription and write controls cover the requirement. Evaluate People Data Labs or Crustdata when an owned application needs an API and the team can maintain the field contract. Test each against the same domain and entity cases, not a vendor's global database count.
Choose Apollo when the value includes prospecting and sales execution, and its API and data-use terms match the workflow. Consider Dropcontact for named-contact enrichment with useful company context, while treating a broad company-only replacement as a separate proof point.
If the current Clearbit setup still meets the contract, price and support requirements, the lowest-risk choice may be to retain it while testing alternatives. A move is justified by a specific operational or commercial need, not merely by the availability of a new API.
FAQ
Is HubSpot enrichment the same thing as the old Clearbit API?
HubSpot offers native CRM enrichment and Clearbit became part of HubSpot. A CRM property workflow and a legacy API integration can have different fields, entitlements and permitted uses. Inventory the exact integration before assuming they are interchangeable.
Does HubSpot data enrichment use credits?
HubSpot's current documentation says standard contact and company data enrichment does not consume HubSpot Credits. It separately identifies smart properties as credit-consuming, and subscription and permission requirements still apply.
Should every form submission trigger another company lookup?
Not necessarily. Several contacts may share an existing company, and a recent trusted result may already satisfy the decision. Reuse only where your agreement allows it and your freshness rule is met. Re-enrich when a material field is missing, stale or disputed.
What if a provider returns a parent company instead of the subsidiary?
Keep the response as an observation and send the identity match for review. Do not merge or overwrite accounts based solely on a similar name or shared domain. The right entity depends on the product, geography and sales ownership.
Is a 20-record test enough to pick the most accurate vendor?
No. It can uncover schema, matching and writeback failures, but it is too small to support a general accuracy ranking. Expand the test across the countries, segments and fields that matter before committing to a large contract.





