Guide · LinkedIn Ads
LinkedIn Ads targeting: the 2026 guide
Who sees your ad is decided before the creative ever loads. The attributes, the AND/OR logic that narrows an audience down to your ICP, matched audiences, and how far to take it before delivery starves.
Last updated July 2026 · Sourced from LinkedIn Marketing Solutions
The short answer
Targeting on LinkedIn is choosing which professionals enter the auction for your ad. You start from a required location, then layer professional attributes, company, job, industry, seniority, or bring your own data through Matched Audiences. Every impression is priced on that identity data, which is why targeting is your single biggest lever on both cost and lead quality.
Our recommendation: build the audience yourself with a Classic, build-your-own ad set, then combine attributes with AND and OR logic to focus tightly on your ICP. LinkedIn's automated ad sets (Auto-Targeting and Accelerate) hand that control to its AI, which is the opposite of what precise B2B targeting needs. This guide covers the attributes, how they combine, matched audiences, and how narrow to go without starving delivery.
The targeting attributes
Location is required on every ad set. Everything else is optional and stacks on top. The art is choosing the few facets that genuinely define fit, not selecting all of them.
| Attribute group | What it covers |
|---|---|
| Location (required) | Permanent, recent, or both, from city to country. Every ad set needs one. |
| Company | Industry, name, size, page followers, connections, growth rate, and category (funding stage, business type, Fortune lists) |
| Job experience | Job titles, job function, seniority, years of experience, skills, and member groups |
| Education | Degrees, fields of study, and member schools |
| Demographics | Age and gender, both inferred. Use sparingly, they narrow reach fast |
| Interests & traits | Professional and product interests, plus member traits |
| Matched & predictive | Your own company and contact lists, retargeting segments, and AI-modeled high-intent audiences |
Combining attributes: AND and OR
This is how you actually narrow an audience to your ICP. LinkedIn joins your selections with two different logics, and knowing which is which is the difference between a focused audience and a bloated one.
OR, within a single attribute
- Add several values to the same facet, three job titles or four industries, and LinkedIn matches a member who fits any of them.
- Use it to widen one dimension: cover every title that describes your buyer without missing variants.
AND, across different attributes
- Each new facet you add, job function then seniority then industry, is joined with AND: a member must match all of them.
- This is the real narrowing lever. Every AND layer tightens the audience toward your ICP.
In Campaign Manager, "Narrow audience further" adds the next AND layer and "Exclude" removes people who match. A clean ICP audience is usually a short OR list inside 2 to 3 AND'd facets, for example (VP OR Director OR Head of) AND (Marketing function) AND (SaaS industry). Stack more than that and reach collapses.
Three ways to build an audience
There are three ways to build a LinkedIn audience. For precise B2B targeting, only one of them is the right default.
| Ad set type | How targeting works | Best for |
|---|---|---|
| Classic, build your own | Full manual control over every attribute and the AND/OR logic below. You define the audience yourself. | Recommended for B2B. Defined ICPs and ABM where you control exactly who is reached |
| Classic, Auto-Targeting | LinkedIn combines its professional audience, platform signals, and your inputs to build the audience for you. | When speed beats precision. You give up control, so not ideal for tight ICP targeting |
| Accelerate | LinkedIn's AI-powered ad set. Uses your source URL, product description, chosen signals, and account history to reach members it predicts will convert. Not all facets are available. | Hands targeting to AI. Hard to keep tightly on-ICP, so not our recommendation for B2B |
For B2B, use Classic, build your own.
It is the only ad set type that gives you full control over the attribute logic and exclusions above. Auto-Targeting and Accelerate hand that control to LinkedIn's AI, which makes tight, on-ICP targeting hard to guarantee. Reach for them only when speed matters more than precision.
Matched audiences & retargeting
Targeting built from your own data almost always outperforms cold attribute targeting, because the intent is already there. These are the four ways to bring your data to LinkedIn.
Company targeting (ABM)
- Upload account lists to reach your target companies and their buying committees.
- The core of any account-based campaign on LinkedIn.
Contact targeting
- Upload contact lists or connect your CRM or marketing platform directly.
- Reach known prospects and existing pipeline by email or contact ID.
Retargeting
- Segment by engagement: website visitors (Insight Tag), video viewers, form openers, and page or event engagers.
- Warm audiences convert several times better than cold.
Predictive Audiences
- Seed with a company list, contact list, conversions, or a lead form, and LinkedIn's AI models high-intent lookalike-style members.
- The successor to lookalikes, which LinkedIn has discontinued.
Audience size and the saturation trap
A campaign needs at least 300 members to run at all. Above that floor, LinkedIn recommends 50,000+ members for Sponsored Content and 15,000+ for Message ads. A larger pool gives the auction and the delivery algorithm room to optimize toward the people most likely to act.
Here is the trap most B2B accounts fall into: their real addressable audience is under 20,000 to 50,000 members. At that scale you cycle through the same people fast, frequency climbs, and CTR decays within weeks. Once you are that small, creative rotation and frequency management matter more than another targeting tweak. Rotate creative every 2 to 4 weeks, cap frequency, and refresh the offer before fatigue sets in, rather than slicing an already-small audience thinner.
How narrow should you go?
Narrowing to your ICP with AND layers is the goal, but there is a floor. Slice too thin and you drop below a viable audience size and saturate fast. The skill is being focused without being starved.
Ease off the layers when
- Your audience has dropped below LinkedIn's recommended size (50k for Sponsored Content, 15k for Message ads).
- Frequency is climbing and CTR is decaying, the signs of saturation.
- Your ICP is genuinely large and another AND layer would cut real buyers.
- A single facet already gates fit well on its own.
Add AND layers when
- Impressions are leaking to out-of-ICP companies, titles, or industries.
- You are running tight ABM on named accounts.
- You are targeting expensive, senior-only buying committees.
- You must control exactly who sees a specific message.
Rule of thumb: define the audience by the 2 to 3 AND'd facets that truly gate fit, usually function plus seniority, sometimes industry, then stop. Over-layering shrinks reach and raises CPMs; under-layering wastes budget on the wrong people. Negate the rest as you go.
Negating: prune the bad-fit audience
Targeting is not set-and-forget. Open each ad set's Demographics report, where LinkedIn breaks delivery down by company, job title, function, industry, seniority, and more, and negate what is a bad fit. We call it Company / Job Title / Industry negating, and it is worth doing every 2 weeks.
Company negating
- Exclude companies that keep showing up but never convert: competitors, students, agencies, or clearly out-of-ICP logos.
- Add them to an excluded company list so they stop eating impressions across every ad set.
Job title negating
- Broad functions pull in adjacent titles you never meant to reach, interns, assistants, or unrelated roles.
- Exclude the specific titles that are spending budget without matching the buyer or influencer you want.
Industry negating
- Watch for industries delivering impressions and clicks but no pipeline, a common leak on function-based targeting.
- Exclude the verticals that are a poor fit so delivery stays concentrated on the industries inside your ICP.
Each negation compounds. Every bad-fit company, title, or industry you remove tightens delivery back toward your ICP, so a 2-week review keeps the audience as focused as possible instead of drifting wider over time.
Targeting best practices
Six habits that keep reach healthy and cost down, whether you build audiences by hand or lean on LinkedIn's AI.
Start broad, then read the data
- Launch wider than instinct says and let the ad set gather data.
- Then exclude what underperforms. It is easier to trim than to expand a starved audience.
Do not stack every facet
- Job title plus seniority plus function plus skills at once collapses your reach.
- Pick the 2 to 3 attributes that genuinely gate fit and stop there.
Always exclude the obvious
- Exclude competitors, your own employees, current customers, and irrelevant functions.
- Every wasted impression is paid for at full price.
Test Audience Network off first
- Turn off Audience Expansion and the Audience Network on the first run.
- You learn your core audience's true performance before letting LinkedIn widen it.
Separate ABM from demand gen
- They need different audiences, budgets, and creative.
- Splitting them keeps reporting clean and stops one from cannibalizing the other.
Layer, do not just target
- Run a brand or awareness campaign first, then retarget the engagers with a conversion offer.
- Members exposed to brand ads are far likelier to convert on the follow-up.
LinkedIn ads targeting FAQ
- A campaign needs at least 300 members to run. Beyond that floor, LinkedIn recommends 50,000+ members for Sponsored Content and 15,000+ for Message ads, because a larger pool gives the auction and the delivery algorithm room to optimize.
Sources
All specs verified against LinkedIn's official documentation.
Need help turning B2B ads into pipeline?
B2B Ads is brought to you by D.I.G.S. Marketing. They run LinkedIn Ads, Google Ads, and Meta Ads, plus email and HubSpot operations for early-stage B2B SaaS. If you want a partner to make your ads actually drive pipeline, book a call.
Get help with your ads