Most guides to LinkedIn account-based marketing stop at "upload your list and target decision makers". That is the first ten minutes. The decisions that set your cost per meeting come after it: how many people the list resolves to once a job filter is on, which format each funnel layer gets, how much budget the audience can absorb, and what counts as an "engaged account" when LinkedIn hides every individual. Below are nine steps, each tied to a published Kiin Labs finding and its advertiser count.
The data comes from Kiin Intelligence, the analytics layer behind Kiin: $60M+ of LinkedIn spend across 1,000+ advertiser accounts, 22,942 campaigns, 12 months to 7 September 2026, LinkedIn Audience Network excluded. The targeting results come from the US-dollar subset in Report No. 7, LinkedIn audience targeting (491 advertisers, $34.8M). The named-account and audience-size cuts were first published in the best ABM agencies in the UK; this page turns them into build instructions.
What LinkedIn ABM is, mechanically
At the targeting level, LinkedIn ABM means restricting a campaign to employees of companies you chose, rather than everyone matching a title, industry and company size. There are two ways to do it.
- An uploaded company list, a Matched Audience. You upload a CSV of company names, domains and LinkedIn Page URLs, up to 300,000 companies, and LinkedIn matches them to its company pages. Expect 60 to 90% of the companies to match. The audience needs 300 matched members before it serves, and delivery is erratic below about 1,000 (Campaign Manager guide).
- Named-company targeting, the company name facet inside the targeting builder. You type the employers in. It is the tool most people reach for when the list is short and important.
Both cost more than attribute targeting and both resolve to fewer people than marketers expect. Across the panel, next to retargeting for scale:
| ABM targeting method | Median audience | CPM | LP CTR | Cost per LP click | Advertisers |
|---|---|---|---|---|---|
| Uploaded or matched company list | 23,000 | $91.66 | 0.335% | $26.56 | 359 |
| Named companies (company name facet) | 4,200 | $91.66 | 0.328% | $27.77 | 100 |
| Retargeting segments, for comparison | 5,350 | $76.55 | 0.406% | $16.42 | 310 |
Cross-account medians, Report No. 7: 10,248 campaigns, US dollar accounts, 12 months to October 2026, Audience Network excluded. LP = landing page. A facet counts if it appears anywhere in the campaign's include rules.
Two things in that table should shape your build. First, the two ABM methods price identically per thousand impressions; the choice between them is about audience size and workflow, not cost. Second, the median named-company campaign reaches 4,200 people, which is already below the size band where LinkedIn ABM performs worst (step 1).
Step 1: Build and size the account list
Start with the number of companies, because it decides almost everything else. Kiin's working rule, from how to build a B2B account list: size the list to about three months of ad budget and outbound capacity. For a B2B SaaS company spending $5,000 to $25,000 a month on LinkedIn, that is usually 3,000 to 5,000 companies. Tightly focused ABM runs on 1,000 to 2,000. Lists of 10,000 and above need bigger budgets, or a top-of-funnel layer that builds retargeting pools first.
Then translate companies into people, because LinkedIn prices people. The number that matters is the audience LinkedIn forecasts once the job filter from step 3 is on, not the account count on your spreadsheet. Write that forecast down before you build creative and check it against these bands.
| If the layered audience is | LP CTR | Cost per LP click | What to do |
|---|---|---|---|
| Under 5,000 people | 0.292% (lowest) | $27.60 (highest) | Widen first: add accounts from the next tier, or relax the person filter one notch. If the list cannot grow, move the budget to the inbox, where tight is an advantage (step 5). |
| 5,000 to 20,000 | 0.420% (highest) | $25.79 | The sweet spot for click rate. Build the core feed programme here. |
| 20,000 to 80,000 | 0.401% | $24.50 | Good for a one-to-many tier. Split by segment only if each half stays above 5,000. |
| 80,000 and above | 0.370% | $22.03 (lowest) | Cheapest click, but you can no longer see frequency building on any one account. Segment by tier or region. |
Named-company campaigns only, split by the audience size LinkedIn reported, campaign-level medians across advertiser accounts (376, 145, 81 and 63 campaigns). Source: best ABM agencies UK. Directional: cross-account, not paired.
Under roughly 5,000 people the auction has too little inventory, the same members see the ad again and again, and each visit costs more while fewer people take it. One-to-one ABM aimed at five or ten strategic logos belongs in sales-led channels with LinkedIn as air cover, not in a standalone feed campaign.
Step 2: Upload it, or name the companies
Within the same advertiser, naming the companies does improve the audience. Across 72 advertisers who ran both named-company and attribute campaigns in the same year, the named-company campaigns earned 28% more landing page clicks per impression. They also paid a 75% higher CPM, so each landing page click still cost 23% more. All three results are significant.
That gives a simple choice rule:
- Use an uploaded list when you have more than a few dozen accounts. It scales to thousands, it feeds the Companies tab reporting in step 9, and you can sync it from a CRM rather than retyping. The uploaded list is also the version the layering result in step 3 was measured on.
- Use named-company targeting for a short tier-one list you will combine with others. Put the 30 accounts sales care most about into the same campaign as the wider tier-two list, so the campaign clears 5,000 people, rather than giving them a campaign of their own that lands in the worst band.
- Never run either as if it were warm. A list tells LinkedIn who you want. It does not make those people familiar with you. On thought leader ads an uploaded list performs like cold targeting, with every paired interval spanning zero (Report No. 1), and across all formats it costs 19.7% more per landing page click than cold native targeting in the same account (165 advertisers).
That last point is where most programmes go wrong: a list run on brand awareness, at retargeting-style frequency, with a demo ask, is a cold audience treated as warm.
Step 3: Layer the people on top
Whether to put a job filter on an account list has the cleanest answer in the panel. Kiin identified 362 individual audience segments, by their LinkedIn identifiers, that advertisers ran both bare and with a person filter, so the list itself is held constant. With the filter on, CPM rose 30.4%, landing page click-through rose 38.6%, and the cost per landing page click fell 26.2%. All three are significant across 317 pairs.
A bare list is half an audience: it picks the companies, then serves anyone who works there. The person filter is the half that finds a buyer. Phil Ilic put the distinction this way in Report No. 7: "Choosing who matters is strategy and it should be tight. Choosing how few people to put the budget in front of is a media decision and on cold inventory it should be loose."
Which filter you use still matters for cost:
| Layering choice | Measured effect | Base | Rule |
|---|---|---|---|
| Any person filter on the same list | LP click −26.2%, click-through +38.6%, CPM +30.4% | 317 segment pairs, significant | Always layer. A bare list is unfinished. |
| Job titles vs job functions | Titles +28% to +38% CPM at the same audience size, same click-through | 6,950 campaigns in matched size bands; +13.0% paired on 142 advertisers | Start with function. Add titles only where the function taxonomy mis-files your buyer. |
| Seniority added on top of function | Click-through −23.9% | 37 advertisers, paired, significant | Test it, do not assume it. One of seniority or function is often enough on a list. |
| Using exclusions vs none | CPM −19.2%, cost per lead −60.0%, click-through −24.4% | Same account, significant | Always exclude customers and open opportunities from the cold layer. |
All rows from Report No. 7, US dollar accounts, Audience Network excluded. Layering means the filter sits in a separate AND group from the list, which is how 93% of panel campaigns that combine them are built.
One build detail catches people out. In the targeting builder, facets inside one group are joined with OR and separate groups are joined with AND. If you add job functions to the same box as the company list, you have widened the audience to "anyone at these companies, or anyone in this function anywhere", which is the opposite of ABM. Put the list in one group and the person filter in another, then confirm the forecast audience went down, not up.
Step 4: Untick the two settings that undo the list
Two boxes under the targeting block are ticked by default and both work against an account list. Untick them on every campaign and check again after saving.
- LinkedIn Audience Network: off. It serves ads in third-party apps where you cannot verify who saw them. In Kiin's audit of 107 accounts, $274,000 of spend had gone to the Audience Network, and in the worst account 99% of a $50,000-a-month budget was served off LinkedIn (Campaign Manager guide). Every figure on this page excludes it.
- Audience Expansion: off. It adds members LinkedIn considers similar, which spends your ABM budget on companies you deliberately left out.
Accelerate and auto-targeting campaigns switch the Audience Network on by default and hand the bid to LinkedIn, which is why Kiin does not use them for account-based programmes.
Step 5: Pick formats by funnel job
An ABM programme on LinkedIn has three jobs, and each wants a different format and audience size.
| Job | Format and objective | Audience | What the panel says |
|---|---|---|---|
| Warm the list | Thought leader ads, engagement objective | The layered account list, above 5,000 people | On an uploaded list, brand awareness costs 49% more per engagement than engagement, and on lists under 30,000 delivers 77% fewer conversions per dollar (Report No. 1). Engagement TLAs run a $13.51 median landing page click. |
| Convert the warm accounts | Conversation or message ads, often with no form, linking to a calendar | Retargeting pools from step 6, kept tight | Retargeting earns 71% more leads per send than cold and is 36% cheaper per lead than a list (Report No. 2). Retargeting pools book 55% of clicks at $145 a meeting. |
| Capture the demo ask | Lead gen form in the inbox rather than the feed | Tight and relevant in the inbox; broad in the feed | The same demo ask costs 57% less per lead in the inbox than the feed, paired in 23 accounts; the inbox form completes 42% of the time against 3.9% (Report No. 4). |
Sources: Report No. 1, thought leader ads (464 accounts, $8.4M of TLA spend); Report No. 2, conversation ads (263 accounts, $7.7M); Report No. 4, the cost of a demo (268 advertisers, $8.25M).
The audience-size rule flips between feed and inbox, which suits ABM. In the feed you pay per impression, so a broad audience lets the auction find cheap inventory: cold feed demo campaigns on 250,000+ people got a lead for $76 to $87, against $232 under 30,000. In the inbox you pay per send whatever the relevance, so relevance is everything: cold inbox campaigns under 10,000 people produced demo leads at $173, against $635 above 170,000. A tight named-account audience that struggles in the feed is exactly what the inbox wants.
Two cautions. A raw list in the inbox performs like cold and costs 21% more per send, so message the engaged pool. And refresh thought leader creative by audience size: under 10,000 people see a creative once within 14 days; 100,000 take 53.
Step 6: Build the retargeting ladder
The ladder turns a cold named-account list warm. Each rung is a Matched Audience built from what people at target accounts did, and each sees a different message. Retargeting is not cheaper than cold traffic: paired inside the same advertiser across 156 advertisers, CPM, click-through and cost per landing page click were all statistically indistinguishable. What retargeting has that cold does not is a click rate that holds steady at any pool size, from under 2,000 people to over 50,000 (0.375% to 0.402%). So the rungs can be small without the step 1 penalty.
| Rung | Built from | What they see next | Watch for |
|---|---|---|---|
| 1. Engaged with a thought leader ad | Ad engagers from the warm-up campaigns | A second thought leader ad from a different author, or a case study | The pool fills slowly on a small list; run it once it passes 300 members |
| 2. Watched most of a video | Video viewers at 50% or 75% | Proof: a customer result or a benchmark | Viewers are anonymous; spend against the pool, do not try to export it |
| 3. Visited the site | Insight Tag, segmented by page (pricing, case studies) | A message ad that links to a calendar | One campaign for all visitors repeats the first touch; segment by page |
| 4. Opened a lead form and left | Lead form openers who did not submit | The same offer again, one field shorter, or the inbox version | 88% of form openers abandon; an open costs $18.21, a submission $140.49 |
| 5. Engaged account | Companies tab, engaged companies exported as a company list | Outbound from sales to the roles you chose in step 3 | This is the hand-off; see step 9 |
Retargeting results: Report No. 7. Form abandonment: LinkedIn lead form abandonment, 11.6% median completion. Pool mechanics: LinkedIn retargeting for B2B.
Two things kill a ladder. The first is draining it. Retargeting pools are small, the median pool behind a demo campaign is 4,600 people, and a feed campaign on a pool that size will deliver several times its audience: past four times, cost per demo lead was 2.2 times where it started (Report No. 4). Keep the daily budget per pool under $2 per 1,000 members and move conversion asks into the inbox, where a small pool is not a problem. The second is pausing the warm-up layer to save money: the pools stop refilling and every rung goes quiet.
Step 7: Bid manually, then fix the page
An account list is expensive inventory, and the bidding default makes it more expensive. Maximum delivery, where LinkedIn sets the bid to spend your whole daily budget, cost 103% more per website click than a manual bid in the same account and objective, across 90 accounts (Report No. 5). Bidding high manually is a separate mistake: in the inbox, the highest-bid campaign in an account paid 80% more per send and 44% more per lead than the lowest, with no change in open rate (119 accounts). Bid manually, start low, raise only if delivery stalls.
Then stop tuning the bid. Across 119 accounts measured from impression to booked meeting, cost per booking varied 121 times from cheapest to dearest. Traffic price explained 1.5 times of that spread; landing page and form conversion explained 72 times (best B2B marketing agencies, LinkedIn bidding strategy). Having paid a premium to reach the right companies, sending them to a generic homepage wastes it. Build a page, or a calendar link, for each tier.
Step 8: Do the budget maths before launch
The ABM budget question is how much the audience can absorb. In Report No. 7, spend per 1,000 people in the audience predicted cost better than size alone: below $5 per thousand, the median landing page click cost $15.18 and CPM $43.75; above $300 per thousand, $39.63 and $129.04. Click-through stayed flat at 0.31% to 0.35%, so pushing harder only buys dearer traffic.
The working rule for cold feed campaigns is $1 to $2 a day per 1,000 people in the audience (LinkedIn ads cost), with LinkedIn's own floor of $10 a day per campaign. Applied to the layered audience from step 1:
| Layered audience | Daily budget at $1 to $2 per 1,000 | Monthly media, all feed campaigns on it | Read |
|---|---|---|---|
| 5,000 | $5 to $10 | $150 to $300 | Below LinkedIn's $10 floor at the low end. One campaign at most; this audience belongs in the inbox. |
| 15,000 | $15 to $30 | $450 to $900 | Room for one warm-up campaign. The rest of the budget goes to the ladder. |
| 30,000 | $30 to $60 | $900 to $1,800 | Warm-up plus a second author or format. |
| 60,000 | $60 to $120 | $1,800 to $3,600 | A full programme on one tier. |
| 100,000 | $100 to $200 | $3,000 to $6,000 | Split into tiers or regions so you can see frequency per segment. |
Arithmetic, monthly = daily × 30. The rule is a ceiling on cold feed spend per audience, not a target; retargeting and inbox budgets sit on top. Report No. 7 measures spend per thousand over a campaign's life, so three months at $1 a day per thousand lands near $90 per thousand, inside its $40 to $100 band ($22.57 median landing page click).
So the audience, not the ambition, caps the warm-up budget: a programme on 15,000 people cannot usefully spend $10,000 a month in the feed. The cap moves once accounts warm up. Paired within advertiser, pushing $100+ per thousand into a small audience, against under $15, raised the cost per landing page click 40.7% but cut cost per lead 48.7%, because the small, heavily funded audiences were warm lists and retargeting pools. Overspending a cold list is waste; concentrating budget on a warm pool is the point of the ladder.
For scale: $4,000 a month of media at the $26.56 median uploaded-list landing page click buys about 150 visits from target accounts. That is why Kiin judges ABM on meetings and engaged accounts, not lead volume.
Step 9: Measure accounts, not people
LinkedIn does not tell you which people at a target account saw, clicked or engaged with your ads; no report, export or API field returns it (can you see who clicked your LinkedIn ads?). Only a completed lead form and a reply to a message ad produce a named person. Everything else resolves to the company, or to an anonymous retargeting pool.
So the company is the unit of measurement, which suits ABM. The tool is the Companies tab in Campaign Manager, which replaced the company engagement report in November 2024. It shows each company on your matched lists with an engagement level, paid and organic engagement and website visits over 7 to 180 day windows, and builds a company list from engaged accounts in one click.
| ABM metric | Where it comes from | Honest definition | Do not claim |
|---|---|---|---|
| Accounts reached | Companies tab, paid impressions | List companies served impressions in the window | "Buying committee reached": reach per person inside a company is modelled, never reported |
| Accounts engaged | Companies tab, engagement level rising | Companies whose paid or organic engagement is rising | Who engaged. The report names the company, not the people |
| Accounts on site | Insight Tag website visits in the Companies tab, or a site identification tool | Companies identified as visiting | That a visitor was the buyer |
| Named hand-raisers | Completed lead forms, message ad replies | The only person-level signals LinkedIn gives you | That a form open is a lead; 88% of opens never submit |
| Frequency | Reach API, deduplicated across campaigns | Account-level frequency, about 2.7 times the campaign figure | Campaign frequency as the real number: 3.45 reported, 9.48 actual (71 advertisers) |
Companies tab: Campaign Manager guide. Frequency: LinkedIn ad frequency and fatigue, 154 ninety-day windows. 2.7 is 9.48 ÷ 3.45.
An engaged account is the sales trigger; outbound goes to the roles chosen in step 3, assuming the company has seen you, not that a named person clicked. Pull the tab monthly, export engaged companies as a list, move them from the warm-up layer to the inbox rung, and add a "how did you hear about us" question to the booking form.
What the data says: decision rules for LinkedIn ABM
The findings above, plus the enterprise cost most ABM briefs face, reduce to seven rules. The named-account premium, the size bands and the enterprise comparison were first published in our UK ABM agency comparison; the layering, retargeting and spend-intensity results come from Report No. 7.
| If you are deciding | The rule | The evidence | Strength |
|---|---|---|---|
| Whether ABM targeting saves money | It does not. Budget for a premium and judge it on meetings | Named companies vs attributes: CPM +75%, LP CTR +28%, LP click +23% | 72 advertisers, paired |
| How small a campaign can go | Keep feed campaigns above about 5,000 people | Under 5,000: lowest LP CTR (0.292%), dearest LP click ($27.60) | Cross-account, directional |
| Whether to filter people on a list | Always, in a separate AND group | Same list layered: LP click −26.2% | 317 pairs, significant |
| Whether a list counts as warm | No. Run it as cold until people engage | List vs cold: LP click +19.7%, click-through no better | 165 advertisers, paired |
| How hard to push budget | $1 to $2 a day per 1,000 people, cold | Under $5 per 1,000: $15.18 LP click. $300+: $39.63 | 9,297 campaigns |
| What enterprise accounts cost | Expect a dearer click, not a better one | 5,001+ employees vs 11 to 500: LP click $43.15 vs $26.24, LP CTR 0.214% vs 0.210%; +27% paired | 24 advertisers paired |
| Where to ask for the meeting | The inbox, to a warm pool | Demo lead −57% in the inbox vs the feed | 23 accounts, paired |
Read the first and last rows together: you pay more to reach the right companies, you do not recover it at the click, and you recover it at the conversion, in the inbox, on a warm pool, with a page built for the account. The enterprise ABM agency guide goes deeper on what large-account lists cost to cover.
Kiin runs this for you
Kiin runs LinkedIn-led ABM programmes for B2B companies: named-account targeting on LinkedIn, retargeting across LinkedIn, Google and Meta, and outbound triggered by account engagement, for a flat management fee from $2,500 a month with ad spend paid directly to LinkedIn and never marked up. Details are on our ABM agency page, and a 30-minute call to look at your list is at kiin.co/book.
From the pricing page: LinkedIn Ads $2,500 a month, LinkedIn plus Google $4,500, full paid media (LinkedIn, Google, Meta, Reddit) $5,500, and GTM outbound, which turns engaged accounts into conversations, as an add-on from $4,000. We ask for an initial three months and recommend about $4,000 a month in LinkedIn media. Never a percentage of spend.
What it does not include: Kiin is a LinkedIn-led paid media agency with its own research panel, not a full-service ABM consultancy. No intent data resale, direct mail or events, and not the right choice for one-to-one programmes on five strategic logos. For that brief, the agencies in the best ABM agencies, the best enterprise ABM agencies and the best ABM agencies in the UK include firms built for it. More on our LinkedIn Ads service.
Panel: $60M+ of LinkedIn spend across 1,000+ advertiser accounts, 22,942 campaigns, 12 months to 7 September 2026, LinkedIn Audience Network excluded. Targeting results are from the US-dollar subset in Report No. 7 (491 advertisers, $34.8M, 10,248 campaigns, 12 months to October 2026). Medians across advertiser accounts; "paired" means the same advertiser ran both options in the same year. LinkedIn's own documentation on ad targeting and conversation ads covers the settings referred to here. Cite as: Ilic, P. (2026). LinkedIn ABM: how to run account-based marketing on LinkedIn Ads. Kiin. kiin.co/linkedin-abm.