Skip to main content

​
General Questions

Yes, our on-demand scraping system supports scheduled refreshes, so we can ensure that selected profiles are updated on a weekly or monthly basis.
Yes, each profile you fetch from the fetch endpoint is counted as 1 request against your quota of 1 million per month. You can fetch profiles in bulk up to 1000 profiles per call (it will count as 1000 requests). Search calls are not counted against the quota, regardless of how many calls you make or how many IDs you get back.
Null values are omitted by default.
The data in our database is sourced from various vendors, each with their own update schedules. As a result, the refresh frequency varies depending on the vendor and data type.
Sarah was working as a Client Success Manager at Company A. She moved to Company B in January 2024 with the same job title. In March 2024 she updated her public profiles and added the information about the job change. In July she gets promoted and her title changes to Senior Client Success Manager. This information immediately gets published.Below is how Sarah’s employment changes will affect the date fields in her profile:current_job_updated_at - timestamp of the last update of the job_title field in The Swarm database. This is when the change was detected by The Swarm so in the case of Sarah it will be:
  • 2024-03-08T15:12:33.191068Z between March and July
  • 2023-07-15T12:22:17.030094Z since mid-July
latest_company_change_at - the date (rounded to full months) of the latest change of the job_company_name. In 2024, Sarah changed the company once, in January, but the information about it went public in March. Hence, the latest_company_change_at equals 2024-01-01 since March 2024.latest_role_change_at - the date (rounded to full months) of the latest change of the job_title. Sarah got promoted and her job title changed, so this field takes the value of 2024-07-01job_start_date - the start date (rounded to full months) of the latest employment, which starts with a new title or company. You can think about this field as the greatest of latest_company_change_at and latest_role_change_at.updated_at - technical timestamp of the latest profile update. We are constantly working on improving the data in our database. We might change the values of any other field in the profile if we learn something about the person. It gets reflected in this field’s value.
Network: google → connections from Gmail integration
Origin: google → connections imported from email contacts
Origin: google-calendar → imported from Calendar eventsNetwork: linkedin → linkedin connections
Origin: plugin → imported from LinkedIn using our Chrome Extension
Origin: csv → imported from Linkedin using CSV
Origin: overlaps → imported using algorithms based on work history data from LinkedIn profiles

​
Data Questions

We can provide fresh datasets daily. Job changes typically range from 10k-50k profiles per day.
We use internal deduplication algorithm to merge duplicate profiles, but minor inconsistencies may still exist.
The estimate value is around 1%. We are currently in the process of reducing this value even further.
No, phone numbers are not currently on our roadmap.
Yes, we ensure persistent IDs for both people and companies across different data loads.
No, this field is not available in the default public profile, so it has very limited availability in the ethical data sources.
Yes, Linkedin Urls are generated based on username, ensuring consistency.
Yes, we support data enrichment requests. The process is automated: if you share a list of profiles or identifiers, we can deliver enriched data accordingly.
No, we do not track gender, ethnicity, or other diversity-related data.
We comply with industry standards to anonymize and secure data while maintaining compliance with privacy regulations.
Yes, we typically provide exports in batches of 100,000 records.
Yes, if a person has a job description listed on their LinkedIn profile, it is also available in our data.

​
Search & API Queries

Responses are in json format.
You can filter by network using the inNetworkOnly parameter in search queries.
Not directly. You need to first fetch profiles with list and connection data and apply filtering in downstream processing.
Yes, a partner-level API key enables programmatic creation of new The Swarm accounts and isolated teams.
Connectors can be added via API using a partner-level key. Removal is not yet automated but is supported by our team.
It identifies and enriches people connected to a target company based on work history, shared networks, and investments.
Not currently, but we can provide insights on request.
You are charged per fetched profile, not per API call. Search is free and unlimited.
No, tenure-specific filtering is not available at this time.
No, revenue data is not currently planned.
Yes, you can access your data via API, including connection strength scores, intro paths, relationship details, as well as enriched information about people and companies.

​
Data Quality & Formatting

Not always. Smaller companies may lack persistent IDs.
Names are user-generated on LinkedIn, so variations exist. Full names may include middle names or initials.
No, but timestamps indicate when updates occurred, with a specific timestamp for job changes.
Currently, we don’t detect or filter out fake profiles. We detect and dismiss only empty profiles (profiles for which we don’t have basic information).
We don’t have such parameters, as the Name and Company are enough for profile to be identified and saved. The amount of information we have for each profile may be very different but there is always a chance that it will be enriched in the future. Additionally, a field with the number of LinkedIn job experiences is available and can help assess completeness.
We only keep companies with high-quality extracted data. Work is ongoing to expand and improve company records.
Yes, we’re actively working on a project aimed at increasing work email availability to up to 80%.
Some people may have multiple degrees and this is how we currently store that information. In most cases, the array should contain a single value.
Our job change tracking system is new, and gaps exist in historical data. We are improving our methods to reduce stale profiles.