The three audience types
Cold is attribute targeting — title, function, industry, company size, seniority — with no list. Uploaded list is a company or contact list matched to LinkedIn members. Retargeting is website visitors, video viewers, lead-form openers and people who engaged with your ads or company page. Every number below is from Kiin Labs 2026 — 1,588 sponsored messaging campaigns, 263 accounts, $7.7M — and every comparison is the same advertiser running both audience types in the same window.
Cross-panel: what each audience type does
| Audience | Campaigns | Accounts | Spend | Open rate | Lead rate | Cost per send | Cost per lead |
|---|---|---|---|---|---|---|---|
| Cold | 512 | 108 | $2.30M | 43.7% | 0.15% | $0.33 | $126 |
| Uploaded list | 372 | 93 | $2.86M | 38.1% | 0.11% | $0.61 | $340 |
| Retargeting | 206 | 114 | $0.88M | 47.6% | 0.14% | $0.47 | $213 |
Campaign medians, lead gen form campaigns. Read this table carefully, because it says cold is cheapest per lead — and that is true across the panel, but it is a statement about which advertisers run cold (big audiences, low bids) rather than about the audience. The uploaded list, where 37% of the money goes, is the dearest by a distance.
Paired within account: retargeting wins on everything that is response
| Same account, running both | Cost per lead | Lead rate | Open rate | Cost per send |
|---|---|---|---|---|
| Retargeting vs cold (n=49 opens; 40 cost per lead) | −23% [−49%, −3%] | +71% [+15%, +116%] | +18% [+11%, +26%] | +6% (ns) |
| Retargeting vs uploaded list (n=49; 32) | −36% [−49%, −11%] | +52% [+19%, +72%] | +21% [+14%, +35%] | −10% (ns) |
| Uploaded list vs cold (n=35; 27) | +18% (ns) | −9% (ns) | +1% | +21% [+4%, +44%] |
Retargeting was cheaper per lead in most accounts, opened better in 44 of 49 against cold and 43 of 49 against lists, and cost the same per send. That last point matters: the warm audience is not a dearer audience to reach — cost per send is set by your bid and the audience size, not by who is in it (how conversation ads are priced). It is simply an audience that answers.
The uploaded list result: a list is not a warm audience
This is the finding clients push back on. Most treat their target account list as the warmest audience they have — these are the companies we want, sales already knows them, surely a message to them lands better than a message to strangers. In the data it does not. Paired against cold in 35 accounts, an uploaded list opens 1% better, converts 9% worse per send (not significant either way), and costs 21% more per send, which is significant. Against retargeting it loses on every metric.
The reason is simple once you see it. A retargeting pool is people who have already done something — visited the site, watched the video, opened a form. A list is people you have selected. The message ad cannot tell the difference between a selected person and a stranger, and neither can the person. What a list buys you is precision on who, at a premium on cost per send because matched audiences are smaller and dearer to fill. That is worth paying for when sales will only work named accounts. It is not a way to make conversation ads perform.
If you only have a list, run it as a cold campaign with a cold target, and get the list into the retargeting pool first: thought leader ads to the list, then conversation ads to the engagers. The mechanics of that layering are in LinkedIn retargeting for B2B.
Cold conversation ads: a different campaign with a different target
Cold works. The ten cheapest accounts in the panel, at $13 a lead, are 98% cold. But look at how they win:
| Cheapest 10 accounts | Most expensive 10 | |
|---|---|---|
| Cost per lead | $13 | $1,776 |
| Cost per send | $0.107 | $1.192 |
| Bid per send | $0.35 | $1.03 |
| Audience size | 97,000 | 17,000 |
| Cold share of spend | 98% | 26% |
| Uploaded list share | 0% | 31% |
They win on cost per send and audience size, not on response. Cold conversation ads are a volume play on cheap sends to big audiences: bid at the floor, target 50,000+ people, and judge the campaign on cost per send and lead volume. Retargeting conversation ads are a response play on warm people: the sends cost what they cost (median $0.47 — the pools are small), and you judge it on cost per lead. Running them in one campaign with one target is how accounts end up at $340 a lead and blaming the format. Kiin's floor for a cold conversation ad is a 15,000-person audience, 50,000 preferred, because cost per send halves from the smallest audience band to the largest.
Booking meetings: nobody does it cold
For the no-form version — a message ad sending to a calendar link — the audience effect is starker. Of the 84 campaigns in the study that recorded booked meetings, 83 ran on a list or a retargeting pool. Retargeting pools booked 55% of landing page clicks at a campaign median of $145 per meeting; uploaded lists booked 28% at $232. The no-form chapter, including cost per booked meeting across all seven tracking accounts, is here; the standalone page is what a booked meeting costs from LinkedIn Ads.
How Kiin runs it
- Build the pool first. If the retargeting audience is under a few thousand people, conversation ads are not the next campaign. Thought leader ads and video views build it; 180-day website and engagement audiences hold it.
- Retargeting conversation ads at the bottom of the funnel, aimed at that pool, lead gen form attached, three fields, bid at the floor. Judge on cost per lead. This is where the $59 leads come from.
- Cold conversation ads as a separate campaign on 15,000–50,000+ people, bid at the floor, judged on cost per send and volume. Refresh when cost per send is 15% above launch.
- Lists go into the cold campaign with the cold target — never labelled "retargeting" in the reporting, because that is how the $340-a-lead expectation gap opens.
- Measure lead quality in the CRM, by campaign. The panel cannot see whether a $13 lead and a $1,776 lead are worth the same; your stage-two rate can. How to judge that is in how much a LinkedIn SQL costs.
The 30% of accounts that got zero leads from conversation ads in twelve months — and spent $1.16M doing it — are mostly the other combination: a cold uploaded list, a high bid, a landing page and a small audience. The full study is LinkedIn Conversation Ads Benchmarks 2026.