> ## Documentation Index
> Fetch the complete documentation index at: https://docs.revenuebase.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# Release notes
> Recent updates to the RevenueBase platform, APIs, and data.
export const RBUpdate = ({label, children}) =>
}
{tag && {tag}}
;
};
## How to read our quality stats
Here is how we measure the quality of our data
* **Email re-verification:** % of emails verified (last 60 days)
* **Contact freshness:** % of job experiences verified on LinkedIn (last 90 days)
* **Company accuracy:** % of company data verified by trusted sources (last 90 days)
**98.4%****77.5%****89%**
## What's new
The June release brings the dataset to **403M total contacts** (no change), with **66.5M verified emails** (+1.2M) and **53.2M phone numbers** (-0.2M).
The job\_order\_in\_profile field is used to indicate
the sequence of a person's job history entries as they appear on their LinkedIn profile. In some cases, this
value did not correctly reflect the true LinkedIn display order — job history entries could appear out of
sequence when reconstructed from this field.
Roughly 3M companies saw updates to their websites and domains.
**97%****84%****89%**
## What's new
The June release brings the dataset to **403M total contacts** (+15M), with **65.3M verified emails** (+0.3M) and **53.4M phone numbers** (+0.6M).
AI-classified fields recovered coverage that was missing from the May release, with substantial gains across all three dimensions:
* **Persona:** 71% → 84%
* **Job function:** 66% → 79%
* **Job level:** 73% → 87%
Additional fill-rate gains across several enrichment fields:
* **ERP tech org:** +11pp
* **Education:** +5pp
* **LinkedIn connections count:** +4pp
* **LinkedIn headline:** +1pp
**98%****83%****89%**
## What's new
The May release brings the dataset to **388M total contacts**, with approximately **65M verified emails** and **52.8M cellphone numbers**.
Significant fill-rate gains across multiple enrichment fields in this release:
* **Persona:** 62.2% → 71.3% *(expected to grow further in the next release)*
* **Job function:** 60.0% → 66.7%
* **Job level:** 70.1% → 72.6%
* **Education:** 41.4% → 44.8%
A new `linkedin_num_id` field is now exposed on person records with a fill-rate of \~95%. This is the permanent numeric LinkedIn Member ID (also referred to as Member Number, LinkedIn ID, or Person ID in API contexts) — a fixed identifier that cannot be changed, unlike vanity profile URLs which users can update at any time.
Because the numeric ID is stable and permanent, it's the preferred identifier for deduplication and cross-system matching. Customers frequently use it to reliably link RevenueBase records with their own CRM data or third-party enrichment sources.
Fixed an issue where the same person could appear multiple times in the dataset with different LinkedIn profile URLs due to URL encoding variations. Approximately 5M affected records have been deduplicated, improving overall dataset consistency.
**97%****84%****89%**
The Contact Freshness is less than our usual number due to brand new contacts we have imported from a new vendor updated at the end of 2025.
## What's new
Integrated an additional data source for collecting social profiles, broadening the foundation of our profile coverage. This expansion will directly improve experience data coverage and increase the total number of profiles available in the dataset.
Major improvement to cellphone data coverage (7.5M cellphone number increase globally, 6.9M in the US), substantially increasing the fill-rate for cellphone numbers across profiles. This enhancement makes mobile contact data more accessible and reliable for outreach workflows.
Cleaned up job title fields that contained meaningless values — records where `job_title` consisted entirely of special characters such as `&`, `;`, `/`, and similar symbols rather than actual job title text. These values have been removed or corrected, improving the overall quality of the data.
**98%****92.3%****88.7%**
## Data updates
* Removed 0.9M contacts who recently left their jobs
* Removed 1.2M undeliverable email addresses, sustaining >95% deliverability
* Removed 0.1M invalid mobile numbers
## What's new
RevenueBase launched our dedicated documentation site covering quickstart guides, full API reference with live playground, and data feed schema documentation.
Two new fields sourced from LinkedIn have been added to the Company dataset:
* **`logo_url`** — direct link to the company's LinkedIn profile logo
* **`company_headline`** — headline text displayed on the company's LinkedIn page
LinkedIn periodically rotates logo URLs as part of their asset protection mechanisms. For long-term use cases, cache a local copy of logos at retrieval time rather than relying on the URL remaining stable.
The Gigasheet dataset now includes the `linkedin_connections_count` column, surfacing the number of LinkedIn connections per profile directly in the dataset. No additional joins required.
Resolved an issue where LinkedIn's data protection mechanisms caused company names to appear in place of job title values. Approximately 90% of affected records have been corrected via a targeted backfill. Remaining cases will be resolved in upcoming job data updates.
Resolved a data gap where last funding dates were not updated for approximately two months due to a third-party provider issue. Funding date fields in the Insight dataset are now current.
**97%****93%** last 90 days**88%**
## Data updates
* Removed 0.9M contacts who recently left their jobs
* Removed 1.5M undeliverable email addresses, sustaining >95% deliverability
* Removed 0.3M invalid mobile numbers
## What's new
A dedicated `company_name` field has been added directly to Person records. Previously, retrieving company names required joining Person data with Company data. This field is now available inline, removing that join dependency.
Person location data now includes full support for Indian state codes and state names, extending existing coverage beyond the USA and Canada. Contacts in India can be searched and filtered using standardized state codes.
**Impact:** 7% increase in fill-rate for `state_code` and `state_name` fields for Indian contacts.
Resolved an issue where company headquarters addresses formatted as "City, State" could be incorrectly assigned to the wrong country. US state codes were in some cases misinterpreted as country codes (e.g., `MD` for Maryland parsed as Moldova).
**Impact:** 2–5% improvement in headquarters address fill-rate and accuracy across countries.
**98%****95%****89%**
## Data updates
* **Email re-verification:** 98% verified within the last 60 days
* **Contact freshness:** 95% of job experiences re-verified on LinkedIn within the last 90 days
* **Company accuracy:** 89% of company data verified by trusted sources within the last 90 days
* Removed 3.2M contacts who recently left their jobs
* Removed 2.9M undeliverable email addresses, sustaining >95% deliverability
* Removed 0.7M invalid mobile numbers
## What's new
LinkedIn has tightened its data protection measures, making data collection significantly harder. A new workaround approach has been introduced to prevent loss of most experience and education data from LinkedIn profiles.
In the January release, all fields maintained approximately the same fill rate as the previous month, with the exception of experience and education data, which saw a minor drop of 1–2%. These protections may also extend the rescraping period going forward.