Skip to guide

See Distill in action

Request Demo

Tell us a little about yourself and we’ll follow up to schedule time.

Browse guides

Targeting groups

On this page 12 sections

Distill helps people find a relevant business audience, research why each person fits, and prepare personal outreach through LinkedIn and email. A targeting group organizes one audience inside a project: the searches that find suitable companies or people, the criteria used to assess them, and the lists that hold the results.

A project can have several groups, such as potential partners, enterprise buyers, and people attending an event. Each can have different company requirements, roles, and research questions. This supports focused audiences for relevant outreach, as well as audience research before any campaign starts.

How groups, searches, lists, and campaigns connect#

PartWhat it does
ProjectHolds the business context, audiences, campaigns, and project-level outreach rules.
Targeting groupOrganizes one audience and its associated searches and lists.
SearchDefines where to look and which criteria, preferences, and research questions to apply.
ListHolds saved companies or people, including search results or imported records.
CampaignUses selected people lists for outreach, with its own sender, messages, channels, and operating settings.

A common company-first arrangement is:

Company search → company list → people search at those companies → people list → campaign.

The company search establishes which businesses fit. The people search identifies relevant people at those businesses. A group can instead start with companies already supplied, search directly for people, or use an existing people list.

Each group can have up to one Companies search and one People search, plus up to one list of each type created for that group. It can also reference existing lists shared from the same project. A search definition, its current results, and a campaign's active audience are separate records. Having a campaign in the project does not automatically connect every group to it.

Groups can be renamed or removed. Removing a group unlinks its lists rather than deleting their records. Shared lists keep their own names when a group is renamed.

Ways to build an audience#

  • Discover companies, then people. Company criteria identify relevant businesses; person criteria identify the intended roles within them.
  • Start with known companies. An existing company list or company CSV can supply the businesses, with a people search finding the relevant contacts.
  • Search directly for people. Role, location, work experience, and other person criteria can define an audience across companies.
  • Start with existing people. A people list, CSV, or available connection list can be the starting audience. Further research can help qualify those records.
  • Use existing relationships. Searches can be limited to available network data. A connection-distance trait checks whether someone is directly connected to a project member or connected through another person.
  • Describe an audience in words. An audience description written in everyday language, with example profiles mentioned in it, can help generate criteria and research questions. Those generated criteria remain visible and editable.

The Starting Set determines the pool being searched. Searching within a list or network does not establish the size of the wider market. Broader searches can use Distill data and supported web or LinkedIn discovery sources, depending on the search path and access. These sources work differently. Finding people at companies in a selected list uses its own discovery process. A source can be selected in the saved search without having returned any results.

Imported records still need to be matched to the right person or company. Creating an import list is not the same as finishing the import or finding every missing identifier. Specialized project types, including Conference and Inbound Screening, can use manual search and list management instead of the targeting-group setup flow.

Company criteria#

The company search editor supports the following criteria. Data coverage varies by company.

AreaAvailable criteria
Company identity and classificationCompany name, company type, LinkedIn industry, startup sector, Distill tags, and global list membership.
Companies related to a people listEmployers of people in a selected list. This looks at those people's current employers.
HeadquartersCompany country. This describes the company, separately from the location of an employee.
SizeTotal employees and known employees by function.
Stage and fundingRevenue bands, total capital raised, last funding-round amount, last round type and year, investors, and founded year.
Other company attributesDomain rank, customer segments, and whether the company is Software as a Service (SaaS).
Text and relevancePhrases in profile text and description similarity.

Known employees by function counts people in Distill's searchable data rather than every employee. If several functions are selected, a company can match when any selected function meets the condition; the counts are not added together. Missing coverage is different from a confirmed zero.

SaaS status and customer-segment classifications can be based on saved AI assessments. They are useful targeting attributes, but their presence is not independent verification of every aspect of a company's business.

Person criteria#

AreaAvailable criteria
RoleJob Title Contains, seniority, years of experience, time in the current role, and time at the current company.
EmployerSpecific companies or company names, companies in a selected list, company tags, industry, headcount, and latest funding.
Location and profile attributesThe person's location, name, age, and pronoun fields where data is available.
EducationSchool and degree.
Text and relevancePhrases in profile text, role similarity, and description similarity.

People searches normally require a current position and exclude profiles that only show retirement. These defaults can be changed. A historical employer condition does not automatically turn off the current-position requirement.

Some work-experience fields can apply to current or past roles, depending on where the criteria are edited. Title and employer conditions that both apply to current roles, or both to past roles, are checked against the same position. A former role at one company does not by itself establish that the person holds that role at a different current employer.

In the common company-first arrangement, the people search is tied to the group's company list. Company fit is established in that list; the people search then focuses on the desired roles and person attributes.

Job title and seniority answer different questions#

Job Title Contains matches phrases, such as Partnerships, Sales, or Engineering. Seniority describes the level. Its supported selections are Founder, Head, Director, Vice President, Chief / CxO, Senior Vice President / SVP, Manager, and Senior Manager.

For example, heads of partnerships can be represented by Partnerships in Job Title Contains and Head in Seniority. This avoids relying on one full title spelling. A Vice President selection does not automatically include Senior Vice President / SVP; distinct levels are separate selections.

Several roles can be represented in separate groups when their requirements differ. For example, one group might target finance chiefs while another targets operations directors. Combining every title and every level into one broad selection can include combinations that were not intended.

Company geography is not person geography#

A company headquartered in the United States can employ someone in Brazil. A US company-country condition and a US person-location condition therefore describe different audiences. A request for US-headquartered companies does not inherently require their employees to live in the US.

Requirements, preferences, and exclusions#

Required structured criteria define the conditions a result must satisfy. Optional positive criteria influence ranking rather than requiring a match. Exclude criteria define which matching records to omit from that search, even when the associated positive criterion is optional.

Multiple values within a criterion commonly represent alternatives. Separate required criteria combine to narrow the audience. Some requests, such as “at least 150 engineers OR at least 500 total employees,” cannot be entered as that exact combination of conditions across different fields. A simpler substitute changes the audience definition and should not be mistaken for the original condition.

Similarity ranks results by how closely their meaning matches example text or roles. It affects ranking without requiring an exact match. It does not establish that a prospect shares every characteristic in the example.

Creating or attaching a standard search to a group can add default geography and minimum company size. Standard defaults include the US plus the known country of the user adding the search, and a minimum of 10 employees for a company search. These are editable defaults, not restrictions on which markets or smaller companies Distill supports. Saved searches and specialized project setups can have different settings. Other search defaults can also affect coverage, including a company profile-photo requirement that can be disabled.

AI traits and additional research#

AI traits assess research questions that standard filters cannot fully answer, such as whether a company publicly describes an integration partner program or whether a person's experience is relevant to a particular business problem.

The search editor offers:

  • AI scoring: evaluates available profile and company context against the trait's description.
  • Web research: investigates a question using available public sources and returns an assessment with evidence and citations where available.
  • Connection distance: evaluates first- or second-degree connections to one selected project member using relationship data, rather than AI scoring.

For AI-scored and web-research traits, results distinguish supporting evidence, unclear evidence, and evidence of a non-match. A required trait can reject a result when there is concrete contrary evidence. Missing or ambiguous evidence stays unclear, including for a required trait. Appearing in the results therefore does not mean every required research question has been answered with a confirmed yes.

Optional traits contribute to ranking without rejecting a result merely because it does not match. A pending or failed assessment is also different from a non-match. Some results can be visible while research or scoring is incomplete.

Existing searches may contain older LinkedIn-insights traits. These use stored function headcount and percentage thresholds rather than the same AI uncertainty rules, and are not offered as a new trait in the current search editor.

Public evidence has limits. A job title alone does not establish purchasing authority, budget, active buying intent, or private software use. A company publishing an article about a technology does not prove its employees use that technology internally. A trait can assess the available evidence; it cannot make a private fact publicly observable.

Connection distance and network scope#

A connection-distance trait selects a project member and allows first-degree, second-degree, or both. A second-degree result represents a route through another person. Shared employment or a shared school alone is not the same as a direct connection for this calculation.

Network coverage depends on available connection data. If Distill cannot determine someone's connection distance, the result remains unscored. That does not prove there is no relationship. Relationship strength and willingness to introduce require separate context.

When the connection distance is missing, Distill can check through the selected member's connected LinkedIn session, subject to availability and limits. These connection checks are separate from finding people at selected companies.

What the counts mean#

Count or controlMeaning
Matching search poolRecords matching the search's required filters and starting pool. It can include records already saved or hidden. It is neither the number of new records waiting to be added nor a count of fully AI-assessed or outreach-ready people. A + means there are at least that many results, and there may be more.
People discoveredPeople found at selected companies, before later checks determine which results qualify and are saved.
Target count in Find people in rolesA target for how many people to find per company. It is not a promised number of qualified people.
Max per companyA cap on people selected per company across that search's results. It is not a universal cap on every record already in a list. Newly added employer-list criteria default to five; the setting can be changed or removed.
Saved companies or peopleRecords currently held in the associated lists. These can differ from the matching pool and from the number discovered.
Campaign outreach limit per companyA separate limit on concurrent outreach to people at one company across the project's groups. It is not a lifetime total, and a search result cap does not replace it.

For example, a target of ten people per company and a result cap of five can be used together. The search looks for a larger pool of people; checks for audience fit and the result cap can produce fewer saved people. Neither setting means five or ten messages will be sent.

The standard targeting-group view requests results in batches of 250. This is a retrieval size, not a limit on the total market. Load more can retrieve additional results when available. Automatic replenishment can top up connected live campaigns according to project settings and each campaign's auto-replenish setting.

Reviewing and refining an audience#

Searches expose their criteria and traits, result rows, supporting evidence where available, and feedback on result quality or missing coverage. In the fuller search results view, saved results and hidden or reviewable results are separate states. A quality review reflects the search run it evaluated. Its date and the results it covered matter when the search has since changed.

Common reasons for unexpected results include:

  • A broad starting pool or an important criterion marked optional.
  • An industry category that is broader than the intended business model.
  • Title phrases or seniority choices that miss relevant variants or admit adjacent roles.
  • Company geography being used where person geography was intended, or vice versa.
  • A list or network that does not contain the intended audience.
  • Incomplete research, missing data, or unclear traits.
  • A per-company cap, limited source coverage, or work that has not finished.

Search data can lag behind changes to a profile or employer.

What changes when targeting is edited#

In the targeting interface, inline edits are applied with Update and run search. Saving or replacing a definition and completing its new search are separate events. For example, pasting criteria onto an existing search updates its definition without immediately rerunning its saved results.

A refreshed search can recheck results it saved into its connected list and remove entries that no longer qualify. Manually added or imported entries are not automatically reassessed against that search in the same way. A criteria edit should therefore not be understood as instantly rebuilding every list or cancelling every conversation already in progress.

Connected campaigns can receive newly eligible people from their source lists. Project exclusions, screening rules, and campaign settings still determine what happens next. Researching an audience or saving a group does not by itself start outreach.

Targeting and outreach exclusions#

An exclusion inside a search changes that search's audience. The app also adds the project's Do Not Contact exclusion when creating a saved search within that project. The project's Do Not Contact settings and configured CRM screening, which checks records from a connected customer relationship management system, serve a different purpose: controlling who may be contacted. Excluding a title in one search is not a project-wide prohibition on contacting a particular person or company.

Rules also have scope. Dividing an audience into additional groups does not bypass project-level company limits or exclusions. Customer and opportunity screening depends on the configured rules and available records; a search result is not proof that someone has passed every outreach check. The project's Screening History can help explain recorded screening runs, rule changes, and decisions. Reporting and results explains access and the limits of that history.

Example audience#

Example configuration:

Goal: identify people responsible for partnerships at US-headquartered software companies with 50 to 500 employees.

PartExample
Company criteriaCompany country: United States; Employees: 50 to 500; an appropriate software industry selection.
Person criteriaJob Title Contains: Partnerships; Seniority: Head; current employer in the selected company list.
Person locationUnrestricted unless the outreach goal also requires a particular employee location.
Additional researchAn optional web-research trait assessing whether the company publicly describes an integration partner program.
Existing customersCovered by the intended project's configured Do Not Contact or CRM rules, rather than assumed from the audience description.
OutputA company list, a people list, and the associated criteria and available research. Campaign connection and sending settings remain separate.

Two other starting points can serve the same goal: a supplied list of relevant software companies, or an existing people/network list narrowed to partnerships roles. An event list can also be a source, but appearing on a previous attendee list does not establish attendance at a future event.

Working through an AI assistant#

A connected assistant can inspect targeting groups, saved criteria, lists, counts, and available quality feedback, and use the targeting actions exposed by its current connection.

The current MCP (Model Context Protocol) create-group action creates a People search; it does not reproduce every company-first setup option in the interface. Supported updates can change saved criteria and start search work.

Related pages: Company search and People search, Research and Enrichment, Networks and Introductions, Campaigns and personalization, Using an assistant.

All guidesBack to top
Distill product guidesAll pages as text