A B2B data enrichment service is measured by rows that reach a person

Which sources actually publish contact data

Enrichment starts earlier than most vendors admit: before a field can be appended to a row, the row has to come from a source that carries the field. Hiring threads give you the company. Directories give you the mailbox. The table is the yield of every family in the file.

Two readings. 40% of the file comes from hiring threads, the family that gives the most companies and almost no mailboxes (304 addresses in 2 792 rows) — that is what an enrichment step is for. And the only family with near-total contact data is agency directories (492 of 492), because agencies put an address on their own site on purpose. A file built only from job feeds is a file of companies; enrichment is the work that turns the rest into something sendable.

What public sources can enrich — and what they cannot

This is the whole reason the offer is priced per working arrangement rather than per record: the fields worth money cannot be bought in bulk, they have to be observed, one row at a time.

The three checks every row passes before it is enriched

Deduplication sits underneath all three: 45 sources produced 6955 rows and 6955 unique root domains, no duplicates and no row without a domain. Subdomains are folded into their root (app., shop., blog.) unless the root is a platform, in which case the candidate is dropped entirely.

What is delivered, and in what shape

• A file with its own provenance. Every row keeps the source it came from and the live check it passed, so another tool, another sender or another agency can continue the work.
• Plain format. CSV and JSON, not a locked export from a tool you have to keep paying for.
• The verified slice separated from the raw slice. Rows with a checked address are marked, rows with only a company are kept in the same file, so you see exactly what was verified and what was not.
• Fields you asked for, not a package of 40. The enrichment runs are scoped to the segment and to the fields that decision actually uses.

The segment work, the three checks and the first sends are scoped as one job: $900 Sprint one-off, $1 900/month Engine for the running loop, $2 900 Full Build when the offer, the page and the distribution have to be built as well. Full scope and what each one stops at: pricing.

Start with the 100-contact test

Enrichment that pays for itself, or not

The honest version: for a segment of a few hundred companies, enrichment done by hand is affordable and accurate; for tens of thousands of rows it is a tooling problem, not a service. The work here is on the side where the row count is small enough that every row can be checked — which is also the side where a wrong field costs a reply. A strict two-variant test needs 13 914 contacts per arm before a rate can be read, so the first test is 100 verified contacts: it decides whether the segment is worth the volume before a month is committed to.

Related: B2B lead list building — where the 6955 domains and their source yield table came from · Appointment setting — the same rows used for their outcome · Outsourced SDR — the sending run as a service · for SaaS, for agencies, for local B2B.

What you can open before paying

The proof page lists the artifacts other people’s repositories accepted, the machine-readable offer carries the same scopes for agents, and the engagement page states what happens in the first week. Nothing on this page is a projection: the numbers are read out of the file it describes.

Questions asked about data enrichment

What does B2B data enrichment cost?

Two markets, two prices. Per-record enrichment runs from fractions of a cent to a few dollars per row, and covers appended fields, not outcomes. Working arrangements are priced per engagement: $900 Sprint (segment, checked rows, first touches), $1 900/month Engine (the loop keeps running), $2 900 Full Build (offer, page, tool, distribution). There is no per-1 000-records price here, because the row is the input.

Can you enrich a list I already own?

Yes, and the first thing that happens is a quality pass rather than an append: domains are folded to their root, rows without a domain are dropped, duplicates are collapsed, then the live check and the address-domain match run. Usually a third of a bought list does not survive that pass, and that is the finding, not a failure.

Do you guarantee a bounce rate?

No. Bounces depend on the sending domain, the volume per day and the age of the row. What is guaranteed is the method: SMTP check before the send, addresses matched to their own domain, and every bounce fed back into the file. A published number for a promise nobody controls is worth nothing.

Where do the 1 040 published addresses come from?

From the companies themselves: role and generic mailboxes on their own pages and in their own listings. No scraped personal mailboxes, no guessed patterns, no purchased databases. That is why the yield is 15% of the file rather than 100% — and why the rows that are there hold up.

Does the enriched file stay mine?

Yes. It is delivered in CSV and JSON with the source column intact from the first delivery. Data and mailbox costs are never inside the fee, and the same list is not resold to a second buyer.

Related reading: Marketing engineer vs agency — what enrichment work costs when it is inside an agency retainer · the Freeman framework — the order these checks run in.