How to Advertise on LinkedIn: Formats, Floors & Where the Reach Actually Is

Three facts govern every LinkedIn plan, and a buyer who holds all three at once will make better decisions than one who holds any two. LinkedIn is the only major platform where the targeting data is a professional identity rather than an inferred behavioural profile. It is the most expensive mainstream platform per click by a wide margin. And its reach is so unevenly distributed across the world that the same campaign brief is a mass-market buy in one country and a rounding error in another.

The third fact is the one most plans get wrong, so it goes first. DataReportal’s Digital 2026 country reports, which read LinkedIn’s own advertising reach figures out of Campaign Manager, put LinkedIn at 98.3 percent of the adult population in the United States and 93.1 percent in the Netherlands in late 2025. The same series puts it at 5.1 percent of Japanese adults. That is not a difference of degree. It is a different medium.

Where LinkedIn has reach, and where it does not

LinkedIn reports potential audience as registered members rather than active users, a caveat DataReportal states explicitly, so every figure below is a ceiling and not an engaged audience. Read them comparatively, which is what they are good for.

The near-census markets. The United States at 270 million members, 98.3 percent of adults. The Netherlands at 14.0 million, 93.1 percent of adults and 76.2 percent of the entire population including children. The United Kingdom at 48.0 million, 86.8 percent of adults. In markets like these LinkedIn functions as the professional register: if a person holds a white-collar job, there is a profile, and the targeting will resolve against it. This is where LinkedIn’s premium is arguably worth paying, because the alternative channels cannot isolate a job function at all. The market context behind those numbers is in the United States guide, the United Kingdom guide and the Netherlands guide.

The fast-growing giant. India reached 170 million members in late 2025, up 30.0 million or 21.4 percent in a single year, the largest absolute increase in the DataReportal series. Yet that is 16.4 percent of Indian adults, so LinkedIn in India is simultaneously one of its biggest member bases and a minority channel. The implication for a buyer is specific: the audience is large enough to build real scale against, the cost per click is far below the US or UK, and the accuracy of self-reported seniority is the thing to interrogate rather than the volume. See the India guide for the surrounding media landscape.

The structurally weak markets. Japan at 5.40 million members and South Korea at 5.00 million are the clearest cases. Both are wealthy, digitally advanced, heavily B2B economies, and in both the professional networking behaviour LinkedIn depends on sits elsewhere: in Japan inside established recruitment intermediaries and company relationships, and in Korea inside domestic platforms. A B2B campaign that needs Japanese or Korean decision-makers at volume does not get there through LinkedIn. It gets there through search and messaging, which in Korea means Naver for intent and Kakao for reach, and in Japan means search plus LINE. The market detail is in the Japan guide and the South Korea guide.

China, which is not a market here at all. LinkedIn announced on 14 October 2021 that it would sunset the localised version of LinkedIn in China, citing a significantly more challenging operating environment and greater compliance requirements, and replaced it with a standalone jobs app with no social feed. That app, InCareer, was discontinued on 9 August 2023. LinkedIn’s help documentation now describes its China focus as helping Chinese Mainland companies hire international talent and market their products and services overseas. Reach into Chinese professionals runs through domestic platforms instead, principally WeChat, and the China guide covers what a foreign B2B advertiser can and cannot do there.

The cost floor, stated honestly

LinkedIn is the platform buyers most often describe as having a punitive minimum bid. The mechanics are not what the folklore says, and the truth is more useful.

What LinkedIn actually publishes is a low floor. Its manual bidding documentation states that the lowest bid that can be placed is $0.01, or the equivalent in the local currency, per result. Its targeting documentation sets the smallest runnable audience at 300 member accounts. Its budget pages no longer print a fixed minimum daily figure at all: they say the minimum lifetime budget is determined by the duration of the campaign or ad set and will be displayed if the budget entered is too low. Campaign Manager suggests a competitive bid range rather than enforcing a high one.

So there is no published rate card forcing a high price. The floor is structural, and it comes from three things multiplying together.

First, the auction. LinkedIn describes it as a competition among advertisers with similar formats and target audiences, priced by audience desirability, bidding strategy, objective and an ad relevance score built from click-through rate, comments, likes and shares. The audiences B2B advertisers want, senior decision-makers in large companies in wealthy markets, are the most contested inventory on the platform and are contested almost exclusively by other advertisers with high customer values. There is no long tail of cheap impressions to fall back into.

Second, the audience minimum interacts badly with tight targeting. A precise account-based audience can sit near the 300-member threshold, which means the campaign is buying a very small inventory pool at whatever the pool costs.

Third, B2B conversion rates are low by nature. A platform where clicks cost several times what they cost elsewhere, combined with a funnel where a small fraction of clicks become qualified leads, means the spend required to observe a statistically readable result is large in absolute terms.

Ad formats and what each is actually for

Sponsored Content is the feed, and it is the default. It covers single image ads, video ads, carousel ads, document ads, thought leader ads, and event and article formats. Everything else on the platform is a variation on or an alternative to this.

Document Ads are the format with no clean equivalent elsewhere. A PDF, DOC, DOCX, PPT or PPTX is rendered natively in the feed so a member can page through it without leaving LinkedIn, and it can be gated behind a Lead Gen Form or left open. LinkedIn’s specifications allow files up to 100MB and recommend under 10 pages against a hard ceiling of 300. For a category where the buying committee genuinely reads a research document, this is the strongest unit on the platform, because the value is delivered before the form.

Lead Gen Forms are the reason many B2B advertisers are on LinkedIn at all. The form opens inside LinkedIn and prefills from the member’s profile: name, contact information, company name, seniority, job title and location. You can select up to 12 fields, including up to three custom questions, which count toward the 12. Gender is not prefilled. The trade is well understood and worth restating: the friction removed raises volume and lowers the average intent per lead, so a prefilled LinkedIn lead and a lead from a typed-out website form are not the same object and should not share a target cost.

Thought Leader Ads sponsor a post from a person rather than a company page. They support brand awareness and engagement objectives across single image, video, event and article or newsletter formats, and video views for video. Permission is explicit: LinkedIn requires you to request approval from the author, who receives an email and can approve, reject or enable auto-approval. Celebrations, documents, polls, multi-image posts and reposts are ineligible, and single image or video thought leader ads carry no call-to-action button, so the link has to live in the intro text. That last restriction is the one that surprises performance teams.

Sponsored Messaging, covering message ads and conversation ads, delivers into the LinkedIn inbox. Conversation ads carry multiple call-to-action buttons that can open landing pages or a Lead Gen Form; message ads carry one. They are billed on cost per send, which is a distinct economic model: you pay for delivery, not for attention, so a poorly targeted send burns budget with no signal at all.

The European position on Sponsored Messaging

This is the single most frequently misreported thing about LinkedIn Ads, and the general claim that Message Ads are banned in the EU is no longer correct.

LinkedIn’s conversation ads help documentation states that starting mid-October 2024, LinkedIn supports European Union targeting for Sponsored Messaging conversation and message ad sets. The restriction that remains is on the member side, and LinkedIn’s Sponsored Messaging overview states it plainly: only EEA and Switzerland members who opted in to see Sponsored Messaging ads in their LinkedIn inbox can receive these ads. The format can be bought against European audiences; the deliverable pool is whoever opted in.

For a buyer the practical consequences are three. Forecast European Sponsored Messaging on delivered sends rather than on nominal audience size, because the gap between the two is much larger than in the United States. Never assume parity between a European and a North American send plan at equal budget. And because the billing is cost per send, an over-optimistic European volume forecast fails quietly as underdelivery rather than loudly as overspend.

Separate from messaging, LinkedIn has narrowed European targeting under the Digital Services Act. Its targeting documentation records that from 15 May 2024, Member Groups ceased to be available as a targeting attribute for the EEA and Switzerland, and it directs advertisers using that attribute to find substitutes. LinkedIn also runs an Ad Library covering ads that ran after 1 June 2023, which shows additional impression, targeting and date information for ads targeted into the European Union, and retains ads for one year after their last impression. That library is a legitimate research asset: it is the only place where a competitor’s European targeting parameters are disclosed by the platform itself.

Targeting on self-reported professional data

LinkedIn’s targeting attributes fall into seven families: company attributes including industry, size, revenue, growth rate and named accounts; demographics; device; education; job experience covering title, function, seniority, skills and years; interests and traits; and location, which is mandatory. Minimum audience is 300 member accounts.

The interesting question is not what the attributes are but where the data comes from, and LinkedIn’s documentation distinguishes. Job title targeting is described as primarily member account input with some business logic and inference, and LinkedIn states no third-party data is used. Skills come from the member-entered skills section. Interests are derived from actions and engagement with content. Gender is inferred from profile information rather than declared. Where a member enters a free-text company name, LinkedIn maps it to a company entity with a deep learning model.

That construction is the platform’s strength and its geographic weakness in one. A profile maintained because it is a person’s working identity, which is the norm in the near-census markets, produces targeting other platforms cannot match. A profile created once for a job search five years ago and never updated, which is closer to the norm in markets with low penetration, produces a job title that is simply wrong. Seniority and job function decay fastest, because they change most often. When you are buying in a market with thin penetration, weight company-level attributes, which are more stable, over individual seniority claims.

Measurement, and the attribution window problem

Conversion tracking runs on the LinkedIn Insight Tag, a client-side tag, and on the Conversions API, a server-to-server integration that carries both online and offline events: sales completed by phone, leads collected at an event, CRM outcomes. LinkedIn deduplicates events sent through both paths so a single action is counted once, which means running both is additive rather than double counting. Email addresses are hashed with SHA-256, event data is retained 180 days while aggregate reporting persists longer, and LinkedIn puts a typical direct integration at two to four weeks.

The window settings matter more here than on any consumer platform. LinkedIn allows conversion windows of 1, 7, 30 or 90 days, defaulting to 30 days post-click and 7 days view-through, and uses last-touch attribution for most conversion types, counting repeat conversions inside a window once. For its own Conversions API, LinkedIn recommends 90-day click and 90-day view for lead or lower-funnel conversions, reserving the 30-day click and 7-day view combination for website conversions.

Read that recommendation as what it is: the platform telling you its default is too short for the thing most of its advertisers are actually buying. A considered B2B purchase with a multi-month committee cycle will not resolve inside 30 days, so a LinkedIn campaign measured on the default will show a cost per acquisition that is wrong in a consistent direction. The fix is not a longer window alone. It is deciding before launch which intermediate event you will actually optimise toward, and accepting that the revenue event will be attributed in the CRM and fed back through the Conversions API weeks later.

What it costs, and why we will not print a number here

Every competing page answers the cost question with a range attributed to nobody. We will not, because the number would be unfalsifiable and the mechanism is more useful than a fabricated midpoint.

LinkedIn itself publishes no rates. Its cost and pricing documentation describes only the determinants: how desirable the target audience is, which of maximum delivery, cost cap or manual bidding you choose, the campaign objective, and an ad relevance score that lowers your price as relevance rises. Manual bids run in CPC, CPM, CPV or CPS depending on objective and format, with a stated floor of $0.01 per result. Nothing in that set produces a single national average, and the same brief priced in two markets will diverge by a multiple rather than a margin.

What can be said without inventing anything: LinkedIn’s clearing prices are the highest among mainstream self-serve platforms because the audience is scarce, professionally qualified and contested by advertisers with high deal values, and the per-click price is the wrong comparison anyway. The comparison that decides the buy is cost per qualified opportunity, which requires your own closed-loop data and cannot be read off any benchmark table, ours included. Sourced rows are published per market as they are verified, with their source and verification date attached, at the LinkedIn benchmark hub.