Lusha product usage data, measured since the buying group tool became available through MCP. Figures cover accounts using the Decision Makers and buying group capability and are aggregate; no individual contact details are included or derivable from them.
A third of the accounts using Lusha’s buying group data have never called the API.
They did not integrate anything. They connected an MCP client, asked a question in conversation, and the data arrived. For a category built on the assumption that B2B data reaches a team through an integration a developer builds, that is a different consumption pattern, and it showed up without a campaign behind it.
The numbers
That works out to about 149 contacts per account and 6.6 contacts per request. Buying group mapping returns a set of people, not a single record, which is why the request count is the smaller number.
Where the credits went
The account split is the headline. The consumption split is the part that changes how you read it.
Accounts touching MCP are about 39% of the total and account for roughly 64% of the credits. Per account, that is roughly 2.8 times the consumption of API only accounts.
One caution on reading that. Credit consumption reflects how each method is priced and how each one is typically used, not only how much value a team gets from it. A conversational request that returns a full buying group consumes differently from a targeted API call for one record. The ratio is real; the explanation for it is not something this data settles.
The growth curve
Weekly active accounts on the buying group tool moved from 18 in the first week to about 2,200, with the steepest part of the climb in the months after MCP access became available.
That growth happened without a campaign behind it. No paid promotion, no launch sequence, no outbound. Teams building agent workflows found the data layer, connected it, and used it.
What this suggests
1. Integration is no longer the gate
The traditional path to B2B data runs through a developer: read the documentation, write the integration, handle authentication, ship it. That path filters out anyone without engineering capacity or time.
30.9% of accounts here skipped it entirely. A revenue operations lead who can connect an MCP client can now get buying group data in a conversation. The technical gate that shaped who could use this kind of data has moved.
2. Conversational access is not lighter usage
The reasonable expectation would be that MCP serves exploratory, low volume work while the API carries production traffic. The credit split points the other way in this dataset: the accounts touching MCP consume more per account, not less. Treat that as an observation on this tool over this period rather than a general rule.
3. Discovery happened without marketing
Teams found this, connected it and used it at volume with no campaign pointing them there. For anyone building a data product, that is the useful part: the demand for agent accessible data existed before anybody went looking for it.
What this data cannot tell you
Outcomes. These are usage figures. Nothing here measures pipeline, meetings, conversion or retention, and higher credit consumption is not evidence of better results.
Cohort composition. Accounts that adopted MCP may differ systematically from API only accounts, by size, by recency, or by how they were acquired. We have not controlled for that here.
One tool, one window. This covers the buying group capability since its launch. It is not a general statement about how all Lusha data is accessed, and other tools may behave differently.
Pricing effects. The credit split reflects pricing design as well as usage behaviour. Those cannot be separated from these figures.
Methodology and data notes
Data source: Lusha internal product usage logs for the Decision Makers and buying group capability.
Window: from the tool’s launch to the reporting date. Figures are cumulative unless stated otherwise.
Account definition: a billing account rather than an individual user, so one account can cover several people.
Access method: accounts are classified by which interfaces they used in the window. MCP only means no API requests were recorded. Both means requests were recorded through each.
Credit share: the proportion of credits consumed against the buying group capability by each group, not total account spend across all Lusha products.
Privacy and compliance: all figures are aggregate product usage. No individual contact details are included or derivable from them, and no customer is identified. Lusha processes business contact data under legitimate interest. Current certifications and full privacy documentation are on the trust center.
Related resources
- Build a verified buying group for a target account — the play behind these numbers
- Best MCP for sales data — what each server does once connected
- Best data provider for Claude
- See every connector
Connect Lusha where your team already works. Add the Lusha plugin to Claude
