B2B data enrichment completes a company or contact record (work email, phone number, job title, company details) by looking it up in external data sources. Waterfall enrichment is a way of doing that lookup: it asks one provider after another and stops as soon as one of them returns the missing field. Because every provider has different gaps, a waterfall usually fills more records than any single source, and the common billing model charges only for the data that comes back.
What is B2B data enrichment?
B2B data enrichment is the process of adding missing attributes to the records you already have about companies and the people who work there, using data from outside your own systems. A record that starts as a name and a company can end up with a verified work email, a direct phone number, a current job title and the company's size and industry.
Enrichment sits between building a list and contacting it. In the five stages of prospecting, it is the step that turns a list of names into people you can actually reach.
The terms overlap in practice. Lead enrichment and contact enrichment both mean enriching people; company or account enrichment means enriching the organisation itself. Data enrichment tools are the products that run those lookups against one or more data sources.
What is waterfall enrichment and how does it work?
Waterfall enrichment is defined as enriching a record by querying several data providers in a fixed order, moving to the next provider only when the previous one returns nothing. The name comes from the way each lookup that comes back empty falls through to the next level.
A waterfall for a work email runs in five steps:
- Send the record to the first provider. The input is usually a full name plus the company domain or the person's LinkedIn profile URL.
- Stop if the field comes back. When the first provider returns an email, the waterfall ends for that record and no other provider is asked.
- Fall through on an empty result. If the first provider has nothing, the same record goes to the second provider, then the third, in the order you set.
- Check the result. The email that comes back is checked for deliverability before anyone writes to it.
- Mark records that no provider could fill. A record that falls through every level stays empty and is flagged, rather than filled with a guess.
The order of the providers matters. Teams usually put first the provider that is strongest in their target market or cheapest per result, and leave the slower or costlier ones for the records nobody else could fill.
Waterfall enrichment vs a single data provider
Waterfall enrichment beats a single data provider on coverage, because it combines the data of several sources instead of stopping at the limits of one. A single provider is simpler to run and easier to audit. The trade-off looks like this:
| Attribute | Single data provider | Waterfall enrichment |
|---|---|---|
| Coverage | Capped by what one provider has for your market | The combined coverage of every provider in the sequence |
| Cost per data point | One provider's rate, sometimes charged even when nothing is found | Depends on which provider fills the field; often charged only when data is found |
| Conflicting data | Rare: there is only one answer per field | Possible: the first answer wins, so provider order decides which value you keep |
| Maintenance | One contract, one integration | Several providers to order, monitor and replace, unless a tool runs the sequence for you |
| Best for | Small volumes in a market one provider covers well | Large or international lists where any single source leaves gaps |
Neither option makes a list accurate on its own. Both need the deliverability check of the next section before a campaign starts.
What data gets enriched, and which is hardest to find
The data that B2B enrichment adds falls into three groups: work emails, phone numbers and company attributes. Company attributes are the easiest to fill, and direct phone numbers are the hardest, because few public or commercial sources hold them.
Work email addresses
A work email is the field most enrichment runs are built around, and its result should always carry a deliverability status. The usual statuses are deliverable (the mailbox exists), catch-all (the company's mail server accepts any address, so the specific mailbox cannot be confirmed) and undeliverable. Writing only to deliverable addresses, and treating catch-all ones with care, protects the reputation of the sending domain.
Phone numbers
A direct or mobile phone number is the hardest field to enrich, because it is rarely published and coverage changes a great deal by country and seniority. Waterfall enrichment is the method that scales for phone numbers; for a single person, the manual routes are in our guide to finding one person's phone number.
Company data
Company data such as industry, headcount, location, website and description is the easiest to enrich, because companies publish most of it themselves. Where that data comes from, and how to filter a list by it, is explained in our guide to B2B company databases.
How data enrichment is usually charged
Data enrichment is usually charged in one of two ways: per lookup, where every query costs a credit whether or not it returns data, or per data found, where an empty result costs nothing. The second model fits a waterfall better, because most lookups in a long sequence come back empty.
HappySales charges enrichment only when the email or phone number is found, and Clay works the same way: its pricing FAQ says that you are charged Data Credits only when Clay successfully returns data (Clay FAQ, checked on 9 November 2026). Per-seat licences and the price of each plan are a separate question, compared tool by tool in our guide to cheaper ZoomInfo alternatives.
Before running a large batch on any tool, check three things: what counts as a billable result, whether a field that comes back unverified is still billed, and whether you can see the cost of the run before it starts.
When is waterfall enrichment worth it?
Waterfall enrichment is worth it when one provider leaves too many records empty for your market. A single provider is enough when its coverage of your target list is already good and your volume is small.
Four signs point to a waterfall:
- International lists: an SDR prospecting across Spain, Germany and Mexico at once will hit markets where one provider is thin.
- Phone numbers matter: a team that calls before it emails needs the field with the lowest coverage filled as often as possible.
- Volume: when a team enriches thousands of contacts a month, every gap in coverage becomes a list of people nobody can reach.
- Billing per data found: when empty lookups cost nothing, adding more providers to the sequence adds coverage without adding waste.
A simple test settles it. Take 50 real accounts from your target list, run them through your current provider and count the empty fields. That count shows how much a waterfall could still fill.
How HappySales does waterfall enrichment
HappySales enriches contacts with work email and phone through a waterfall over several data providers, up to 100 contacts per run, and bills only the data it finds. Each email comes back marked as deliverable, catch-all or undeliverable, so undeliverable addresses can be left out before a sequence starts.
The enrichment runs on contacts that HappySales has already found. It searches LinkedIn for the companies, finds the people in them by job title, function or seniority (up to 10,000 people per search) and saves them to contact lists. From Claude or ChatGPT, where HappySales is added as an MCP connector, a Head of Sales Development can ask in plain language: "get the work email and mobile for the operations managers in my logistics list in Amsterdam".
In HappySales, the cost of every paid step is shown first as a preview and only then does the run begin; a daily spending limit for each organisation keeps an agent from overspending. For company records, each organisation can also create its own fields, such as "does this company sell to retail?", which the user's AI assistant fills in for every company on the list; HappySales stores the answer and does not run a model of its own.
What HappySales does not do is just as clear: it is not a CRM and it has no reporting or analytics. Once a list is enriched, the same conversation can launch an outbound campaign from Claude or ChatGPT on LinkedIn and email.