LinkedIn audience targeting: 2026 benchmarks

The industry teaches the opposite of what the data says

Every B2B marketer is taught the same thing about LinkedIn: get the audience tight. Name the titles, stack the filters, keep it small, pay the premium because the people are worth it. It is the most repeated advice in the category and almost nobody has tested it at scale, because testing it needs cost data from hundreds of advertisers rather than one.

On this panel it is wrong, with one important exception. Narrowing a cold audience makes every impression and every click cost about a third more and does not improve the click rate at all. But narrowing a known audience, by layering a job filter onto an account list you already trust, makes the click 26% cheaper. The difference is not how tight the audience is. It is whether the narrowing is adding information or just removing people.

Does narrow targeting actually perform better?

It costs more. It does not click better. Those two facts together are the finding, and they hold at every audience size, on every objective and inside the same advertiser account.

A target audience on LinkedIn is the number of members matched by a campaign's include rules. The platform reports it before launch, and it is the single strongest predictor of what an impression will cost. Across 9,297 campaigns with a reported audience size, cost falls steadily as the audience grows, bottoms out between 300,000 and 2 million people, and turns back up above that.

Medians, not averages. Audience Network excluded. US dollar accounts, 12 months to October 2026.

The cost spread is wide: a thousand impressions costs 2.2 times more against an audience under 2,000 people than against one of 300,000 to 2 million. Cost per landing page click runs 74% higher at the narrow end.

The defence of tight targeting is that the people are more relevant, so they engage more and the higher price buys better traffic. The click-through column is where that argument has to show up, and it does not. The entire range runs from 0.306% to 0.363%, with the best figure belonging to the 80,000 to 300,000 band rather than the tightest one.

Cross-account comparisons invite an objection: perhaps small advertisers run small audiences, and the cost gap is really a budget or sophistication gap. So we paired it. Every advertiser that ran both a narrow and a broad audience in the same year contributes one ratio, and the median of those ratios is the result.

245 advertisers ran both. Bootstrap intervals, 4,000 resamples, median of per-account ratios.

How much should you spend against an audience of a given size?

The platform never asks this question and almost no advertiser answers it. Spend per thousand people in the audience predicts cost better than audience size alone, and the relationship is close to perfectly ordered.

Audience size on its own is incomplete. What matters is how hard a budget is pushed through it. Budget sizing against an audience is covered practically in penetration rate and minimum budget and what a LinkedIn Ads monthly budget should be. We divided campaign spend by the audience size in thousands, which gives a clean intensity measure containing no cost term, and sorted the panel by it.

Roughly 300 advertisers appear in every band, so this is not a handful of accounts driving the curve.

CPM rises three times over and cost per landing page click 2.6 times, in order, with no reversal. Click-through is flat across a sixty-fold range of spend intensity. Holding spend constant and varying audience instead gives the same answer, and the penalty widens as budgets grow: at under $500 of spend a small audience costs 1.56 times a large one, and at $10,000 and above it costs 2.31 times, with the landing page click at $161.96 against $32.56.

Paired within advertiser, the traffic result holds and a second result appears that runs the other way.

165 advertisers on the traffic measures, 44 on cost per lead.

Read the last row against the rest. Across accounts, cost per lead rises with spend intensity, from $111 to $601. Paired inside the same account it falls by half. Both are correctly calculated and they disagree, which is the signature of a confound: the advertisers who push hard into small audiences are not a random sample, and the small audiences they push into are warm lists and retargeting pools that convert far better than anything cold.

Which targeting facet should you actually use?

Job titles are the most used and among the most expensive. Skills, seniorities, industries and years of experience are cheaper and barely touched. Most of that gap does not survive pairing, but the title premium does.

LinkedIn exposes roughly twenty targeting facets. Advertisers concentrate on four. Titles appear in campaigns from 383 of the 491 advertisers on the panel; skills appear for 151. Seniority and function cuts priced on their own are in the benchmarks by audience series, and the C-suite premium specifically in what it costs to reach the C-suite.

Cross-account medians. A facet counts if it appears anywhere in the campaign's include rules.

That table is a map of where the money goes, not a ranking of what works. Paired within advertiser, most of the differences disappear: titles against skills comes out at +19% CPM with an interval spanning zero on 60 advertisers, and titles against function-plus-seniority at +10.5%, also not significant. The advertisers who use skills may simply be better operators.

One difference survives everything. Comparing titles-only campaigns against functions-only campaigns inside the same audience-size band, so that narrowness is held constant, the premium is large and consistent.

Median audience within each band is near-identical, so this is a like-for-like comparison. Paired within advertiser the same comparison gives +13.0% CPM [+0.8%, +34.2%], significant on 142 advertisers.

The premium runs +28%, +35% and +38% across the three bands, and the click rate is the same to three decimal places in the middle band. Title targeting is the most crowded inventory on the platform and it prices accordingly.

Two smaller results. Adding seniority on top of job function lowers landing page click-through by 23.9% on 37 paired advertisers, significant. And campaigns that use exclusions run 19.2% lower CPM and 60.0% lower cost per lead than campaigns in the same account that use none, both significant, at a 24.4% lower click rate.

Should you layer job filters on top of an uploaded list?

Yes, and this is the clearest result in the report. The same list, with a person filter on top, costs 30% more per impression and delivers a landing page click 26% cheaper.

Account lists are the backbone of B2B targeting and the layering question is the one practitioners actually argue about. You upload 500 companies. Do you then add job titles, or let LinkedIn serve anyone at those companies? If you have not built the list yet, start with how to build a B2B account list for LinkedIn Ads.

Comparing layered campaigns against bare ones across the panel is misleading, because the lists are different sizes in each group. Instead we identified the individual audience segments by their LinkedIn identifiers, found the 362 segments that were run both ways, and compared the same list against itself.

317 segment pairs. A layer means job titles, seniorities or job functions added to the list's include rules.

This is the exact opposite of Chapter 1 and both results are correct. Narrowing a cold audience by shrinking it buys nothing. Narrowing a known audience by filtering to the right people inside it pays for itself twice over. The list establishes which companies matter; the layer establishes who inside them is worth reaching.

For completeness: lists themselves are not cheap. Paired within 165 advertisers, uploaded lists run +18.2% CPM and +19.7% cost per landing page click against cold native targeting, both significant, with click-through indistinguishable. Named-company targeting runs +23% on the same measure across 72 advertisers.

Is retargeting cheaper than cold targeting?

Across accounts it looks much cheaper. Inside the same advertiser the difference disappears entirely, on every measure we can test.

Retargeting carries the best headline numbers on the panel: the highest landing page click-through of any facet at 0.406%, and a landing page click at $16.42 against a panel median of $22.76. Those figures are real and they are also misleading. The practitioner setup is in LinkedIn retargeting for B2B, and the paired cold-versus-retargeting result for the inbox in conversation ads: retargeting against cold.

156 advertisers ran both. Every interval spans zero.

Nothing is significant. The cross-account advantage is a selection effect: advertisers who build and maintain retargeting audiences are, on average, running better accounts in general, and that quality shows up in their cold campaigns too.

Retargeting does have one genuine structural property. Its click rate is flat across every audience size, from under 2,000 people to over 50,000, where cold targeting's click rate moves with size. Familiarity substitutes for precision.

Click-through varies by less than three hundredths of a percentage point across a 25-fold range of audience size.

The cost per lead column should be read with care. Only 17.5% of retargeting campaigns record any in-platform lead at all, because retargeting on this panel is a traffic and conversion play: 30.9% of its campaigns run on website visits and 12.2% on website conversions, against 4.1% for lists and 5.2% for cold. Judging it on lead form completions measures the wrong thing.

Benchmarks: what the narrowness penalty costs on each objective

The penalty is universal and it is worst precisely where advertisers narrow hardest.

Upper funnel is brand awareness, engagement and video views combined. Landing page metrics are omitted for lead gen forms because the form opens in-platform and produces almost no landing page clicks.

Lead gen form campaigns carry the largest penalty and are also the campaigns advertisers narrow most aggressively, on the reasoning that a form submission from the wrong person is worse than no submission. The panel says that reasoning costs 72% on the price of an impression before a single form is opened.

Ten numbers to quote, and what each rests on

Every figure below is median-based, computed on US dollar accounts with LinkedIn Audience Network excluded, over the twelve months to October 2026.

Cite as: Kiin Labs, LinkedIn audience targeting 2026 benchmarks, kiin.co/research/linkedin-audience-targeting-benchmarks

Six things the industry believes about LinkedIn targeting

Each verdict is the result of the test described, not an opinion.

What this is built on, and what it cannot answer

10,248 campaigns from 491 advertisers, $34.8M of spend, 714 million impressions, 1.39 million landing page clicks and 52,372 form completions, in the twelve months to October 2026.

Inclusion. US dollar accounts only, so no currency conversion distorts a cost figure. LinkedIn Audience Network excluded throughout, because off-platform delivery prices differently and would flatter broad audiences. Campaigns need at least 1,000 impressions and non-zero spend. Target audience size is the figure LinkedIn reported for the campaign; 9,297 of the 10,248 campaigns carry one.

Medians, not averages. Every headline figure is a median, and paired tests take the median of per-advertiser ratios rather than pooling spend. One large advertiser cannot move a result.

Pairing. A paired test uses only advertisers that ran both things being compared, in the same year. Intervals are 95% bootstrap, 3,000 to 4,000 resamples. A result is called significant only when the interval excludes zero, and non-significant results are published with the same prominence as significant ones.

Targeting structure. LinkedIn's include rules are an AND of OR-groups. Facets inside one group are united; separate groups are intersected. Where a list and job titles both appear, 10,536 campaigns intersect them against 804 that unite them, so layering in this report means a genuine narrowing in 93% of cases.

What we cannot answer. Whether LinkedIn's standardised job title facet captures all, half or most of the members who hold a given title. That requires member-side data. Whether the extra clicks from including small companies convert worse, because only one advertiser ran both configurations with measurable lead volume. And anything about pipeline or revenue quality, because this panel sees platform outcomes, not CRM outcomes.

One analysis we discarded. Impressions divided by audience size looks like a frequency measure and produces a clean relationship with CPM. It is circular: impressions are spend divided by CPM, so the measure contains CPM in its denominator and the relationship is guaranteed. Spend per thousand audience members, used in Chapter 2, has no such problem.

Frequently asked questions about LinkedIn audience targeting

Does a smaller LinkedIn audience cost more?

Yes. Across 9,297 campaigns, a target audience under 2,000 people costs a median $107.63 CPM against $49.63 for an audience of 300,000 to 2 million. Paired inside the same advertiser, audiences under 30,000 people cost 32.1% more per thousand impressions and 32.2% more per landing page click than audiences above 30,000.

Do narrow LinkedIn audiences get better click-through rates?

No. Landing page click-through runs between 0.306% and 0.363% across the entire size range, and the best figure belongs to the 80,000 to 300,000 band rather than the narrowest. Paired within advertiser the difference is +0.8%, not statistically significant.

What is the ideal LinkedIn ads audience size?

On cost per landing page click the panel bottoms out between 80,000 and 2 million people, at $16.40 to $18.19. On cost per lead the best band is 30,000 to 300,000, at $190 to $199. Below 10,000 people both measures deteriorate sharply.

How much should I spend against a LinkedIn audience?

Keep spend under roughly $15 per thousand people in the audience. At that intensity the median landing page click costs $15.18 to $15.84. Above $300 per thousand it costs $39.63, and CPM is three times higher.

Is job title targeting better than job function targeting on LinkedIn?

It costs 28% to 38% more per thousand impressions at the same audience size, with no measured improvement in click-through. Paired within advertiser the premium is 13.0% and statistically significant. Titles are used by 383 of 491 advertisers on the panel, which is why the inventory is priced the way it is.

Should I layer job titles on top of an uploaded account list?

Yes. Looking at 317 audience segments run both ways, adding a title, seniority or function filter to the same list raised CPM by 30.4% but raised click-through by 38.6%, making the landing page click 26.2% cheaper. All three results are statistically significant.

Is LinkedIn retargeting cheaper than cold targeting?

Not inside the same advertiser account. Across 156 advertisers running both, CPM, click-through, cost per click, cost per lead and engagement rate were all statistically indistinguishable. Retargeting looks cheaper across accounts because advertisers who run it tend to run better accounts overall.

Should I exclude companies with 1 to 10 employees?

Excluding them costs click-through. Campaigns that include micro companies show a 31.9% higher landing page click rate, paired within advertiser and significant across 27 advertisers. Whether those extra clicks convert worse cannot be answered from this panel, because only one advertiser ran both configurations with measurable lead volume.

Do exclusions improve LinkedIn ad performance?

Campaigns using exclusions show 19.2% lower CPM and 60.0% lower cost per lead than campaigns in the same account using none, both significant, alongside a 24.4% lower click-through rate.

What is the narrowest audience worth running on LinkedIn?

If the audience is cold, the data gives no reason to go below about 30,000 people. If it is a known account list or a retargeting pool, size stops predicting cost: retargeting click-through varies by less than three hundredths of a percentage point from under 2,000 people to over 50,000.