LinkedIn targeting combines a required location filter with professional facets like company, title, seniority, skills, and industry, plus Matched and Predictive Audiences for higher-intent reach. Set your geo first, add one or two key facets, then check your audience counts stay at or above 300 before you launch a small test.
TL;DR:
- LinkedIn reports estimates below 300 as zero, so check active members as well as total reach and repeat the count after exclusions.
- CRM and retargeting lists suit warm campaigns, but verify segments are BUILDING, UPDATING, READY, or ARCHIVED; CSV uploads cannot exceed 300,000 hashed email addresses.
- Predictive Audiences accept only a location filter, so use them to extend reach after exhausting known lists, not to target job titles or industries.
- Test one variable at a time, then enable audience expansion only after the core segment performs, monitoring acquisition costs as broader reach can reduce precision.
LinkedIn’s ad targeting system runs on two linked concepts: facets and entities. A facet is the category, like job title, skills, or industry. An entity is the specific value inside that category, such as “Product Marketing Manager” under the job title facet. Many facets support typeahead search and a similar-entities finder, which surfaces related values once you pick one, useful when you’re not sure every relevant title or skill to include.
Most facets cap out around 100 entities, though employer and geography facets allow more room to work with. At the campaign level, you can include or exclude any facet, and that logic applies across every creative running under the campaign, not per ad. There’s one regional exception worth flagging: Member Groups targeting was disabled across the EEA and Switzerland starting in May 2024, so facet availability can shift by region. Build your targeting plan with that in mind if you run campaigns in multiple markets.
Think of facets as a mapping layer between your ideal customer profile and LinkedIn’s available filters. If your ICP includes “mid-market companies with fast headcount growth,” that maps to the company size and growth rate facets, not a single filter you’ll find labeled that way.
Location comes first in every campaign, and it’s not optional. LinkedIn defaults to a member’s self-reported profile location rather than device GPS, which matters for B2B targeting since a traveling executive’s device location would otherwise skew your reach. Get the geo right before touching anything else.
From there, the facets marketers lean on most include:
For lead generation campaigns, a common combo is job function plus seniority plus company size: it narrows the audience to actual buyers without shrinking it past viability. Event promotion often works better with industry plus job title, since the goal is topical relevance rather than tight account precision. ABM pilots typically start with a named employer list plus seniority, sometimes layered with a Matched Audience built from a CRM export for the same accounts.
The mistake we see most often is marketers stacking five or six micro-filters at once (a specific title, a specific skill, a specific school, a specific company size band) until the resulting audience is too small to deliver impressions at any meaningful volume. Two or three well-chosen facets almost always outperform a long filter chain, both for reach and for the learning speed of your tests.
If you’re running parallel prospecting through Sales Navigator, mapping your ICP to Sales Navigator’s filters before you build your Campaign Manager audience keeps your paid and outbound targeting consistent, which matters when you’re trying to read results across channels.
Matched Audiences let you target people you already know: CSV contact lists, website visitors through a retargeting pixel, or CRM and account-based lists synced directly. They’re the closest thing LinkedIn offers to warm-audience targeting, and they consistently outperform cold professional-facet targeting for consideration and conversion stages.
A few platform limits shape how you use them:
Predictive Audiences work differently: instead of a list you upload, LinkedIn builds a lookalike-style audience from your existing one. The catch is that predictive targeting currently accepts only a geo filter as its targeting criterion, no job titles, no industries layered on top. That makes predictive audiences a reach-extension tool rather than a precision one.
In practice, Matched Audiences do the heavy lifting for CRM reactivation and ABM account lists, while predictive audiences earn their place when you’ve exhausted a known list and want incremental reach without starting a cold facet build from zero.
LinkedIn requires a minimum of 300 matched members to run a campaign, and any audience estimate below that threshold returns as 0 rather than a precise number, a privacy safeguard built into the Audience Counts system rather than a bug. If your targeting combination comes back at 0, it almost never means nobody matches. It means the real number is too small to disclose.
The Audience Counts API returns two figures worth distinguishing: “active” audience, meaning members more likely to be on LinkedIn and see your ads soon, and “total” audience, the full matched population regardless of recent activity. A campaign built on total count alone can look larger on paper than it performs in delivered impressions, since active count is the better predictor of near-term reach.
Technically, queries to the Audience Counts API require properly encoded URNs. Unencoded URNs are a common cause of failed automated checks, especially for teams building targeting through the API rather than the Campaign Manager interface.
Once your audience count clears 300 with healthy active numbers, the same targeting criteria feed into Ad Budget Pricing, which returns minimum, maximum, and suggested bid ranges for your chosen daily budget, giving you a planning number before spend actually starts.
Setting up targeting in the order below avoids most of the rework we see when marketers build facets first and geography as an afterthought.
Conseil : Run your audience count check twice, once before excludes and once after, since excludes can quietly drop a viable audience below the 300 threshold without an obvious warning.
Once the audience is saved and counted, pull a bid estimate through Ad Budget Pricing or the Campaign Manager’s suggested bid range before you set your daily budget. That number should inform your budget decision, not the other way around. Campaign structure in LinkedIn runs Sponsored Accounts above Campaign Groups, which sit above individual campaigns and their creatives, and each level carries its own caps, so check current creative and campaign limits if you’re running several concurrent tests under one account.
A clean test changes one variable at a time. Run the same creative against two targeting setups, or the same targeting against two creatives, and let each version run long enough to reach a result you can actually trust rather than calling a winner after a few dozen clicks.
Start narrower than feels comfortable, confirm the audience performs, then broaden deliberately:
Conseil : Treat a falling relevance score as an early signal to refresh creative or tighten targeting, since delivery and cost both get harder once members start tuning an ad out.
The agencies and in-house teams that waste the least budget are the ones that resist the urge to “optimize” by adding filters. Broadening a facet and watching performance almost always teaches you more than narrowing one further. If your tracking setup makes it hard to tell which targeting change actually drove a result, that’s worth fixing before you run another test; fixing tracking before testing creative is the more common bottleneck than the targeting itself.
Our workflow starts by mapping a client’s ideal customer profile to LinkedIn’s facet structure before a single campaign goes live: company attributes, job-based filters, and skills get translated into the exact facets Campaign Manager supports. From there, we run a pilot campaign against a defined budget, track conversion events against CRM data, and rotate creative on a set cadence rather than letting one ad run until performance quietly decays.
Our checklist covers tracking setup, CRM list matching for Matched Audiences, creative rotation schedules, and conversion event verification, the same groundwork we walk through for client engagements across industries. Measurement comes before scale, every time.
The common assumption is that more filters mean a smarter audience. In practice, the opposite usually holds: the campaigns that perform best tend to combine two or three well-chosen professional facets and lean on Matched Audiences for warm reach, rather than chasing precision through a long filter chain that collapses delivery volume. Treat your first targeting setup as a hypothesis, not a final answer, and let the audience counts and early delivery data tell you whether to widen or narrow before you commit real budget to it.
— Xpert
Getting LinkedIn targeting right takes more than reading the facet list. It means mapping your actual buyers to the right combination, checking audience viability before spend, and rotating creative on a schedule instead of letting campaigns run unattended. That’s the work our Planification et achat média team handles day to day, alongside Marketing digital support for tracking, creative production, and conversion setup.
If your team wants hands-on exposure to the platform itself rather than full outsourcing, our Formations cover campaign setup and optimization directly. For a full review of your current targeting, audience structure, and bidding approach, reach out through our services page and request a campaign audit.
Yes. LinkedIn ads require a location filter and let you layer professional facets like company, job title, seniority, skills, and industry on top, plus Matched Audiences for people you already know and Predictive Audiences for lookalike-style reach based on geo.
LinkedIn doesn’t publish a fixed rate since cost depends on your targeting criteria, competition for that audience, and daily budget. The Ad Budget Pricing tool returns minimum, maximum, and suggested bid ranges once you’ve built your targeting and set a budget, which gives a realistic planning figure for your specific campaign.
This isn’t an official LinkedIn targeting rule, and definitions of it vary across marketing sources rather than coming from LinkedIn itself. Readers are better served following LinkedIn’s documented practice of combining two or three professional facets with location, since that’s the structure LinkedIn’s own targeting documentation supports.
LinkedIn requires a minimum of 300 matched members to run a campaign. Audience estimates below that threshold return as 0 rather than an exact number, which protects member privacy rather than signaling an empty audience.
