The Cost of a Demo on LinkedIn: 2026 benchmarks

Everyone asks what a demo costs. The answer is: which quarter are you in?

The question we get most from B2B marketers is some version of "what should a demo request cost on LinkedIn?" The honest answer has never been a number. It is a range so wide that the number in the middle tells you almost nothing: at the median account a demo lead costs $350, but a quarter of accounts get them for under $161 and a quarter pay more than $754. Same platform, same objective, same kind of form.

So this report is not really about the $350. It is about the 5x spread and what sits inside it. We split the 155 accounts with meaningful demo volume into four tiers by cost and by volume, and the first thing that fell out was that the accounts paying $146 a demo lead and the accounts paying $926 spend the same per month, $2,168 against $2,058. Budget is not what separates them. Six things are, and the biggest by a distance is where the form sits: a demo ask in the inbox costs 57% less per lead than the same ask in the feed, in the same account, because the form completes 42% of the time instead of 3.9%.

It is the fourth Kiin Labs report and it is built the way the first, the second and the third were: medians across accounts, never pooled totals; every "cheaper than" paired inside one account; a confidence interval on every claim, and a star key that says how many accounts it rests on. Two findings in it cut against things we have said ourselves. Retargeting is not the cheap demo audience people assume, because the pools are too small and get exhausted. And running demand generation alongside demo capture does not make the demo lead cheaper; it makes the pool bigger. Both are in the report as found.

What we still cannot say is what happens after the form: whether a $146 demo lead and a $926 demo lead are the same quality. That needs a CRM join, and it is the next chapter, not this one.

Use the tables. Cite the numbers. Tell us where we are wrong.

Phil Ilic
Co-founder and CEO, Kiin · London, September 2026

The panel, the four tiers, and how to read the numbers

1,644 demo campaigns, 268 accounts, $8.25M, 20,277 demo leads. Split at the median cost and the median volume, the panel falls into four kinds of account — and the two that matter spend the same money.

Every lead-generation-objective campaign with delivery across accounts connected to Kiin Intelligence, 8 Sep 2025 to 7 Sep 2026, whose lead gen form asks for a demo, a call, an audit, a quote or a trial. Campaigns under $100 and LinkedIn Audience Network delivery are excluded. Demo asks are 38% of lead-generation-objective spend; content offers are another 38%; the remaining 24% could not be classified or has no form synced and is out. 268 accounts ran a demo ask; 155 got five or more demo leads and are the base for every account-level figure below. Seventeen accounts spent $1,000 or more and got no demo lead at all ($48,719, 1% of demo spend).

1. The distribution

155 accounts with 5+ demo leads. The pooled figure (spend ÷ leads) is $407 and is dominated by the biggest spenders; it is not quoted. If you want a "good" number: the top quarter of accounts is under $161 a demo lead and over 17 a month.

2. The four tiers

Split the 155 accounts at the median cost ($350) and the median volume (6.6 a month) and you get four kinds of demo account. Tier A gets cheap leads at volume; tier D pays most and gets least. The whole report is about what A does that D does not.

Shares are spend-weighted within the tier; audience is the spend-weighted geometric mean. Tier B is inbox-only accounts on tiny budgets ($477 a month). Tier C is uploaded lists in the inbox at $8,428 a month: they bought volume and paid the list premium for it (51% of their spend is on lists).

Mann-Whitney. Winners have fewer creatives per campaign because their inbox campaigns are single-message; read chapter 05 before concluding fewer is better.

3. How to read the numbers

Every benchmark is the median across advertiser accounts. Every comparison marked paired is the same account running both settings, reported as the median of each account's own ratio with a 95% bootstrap interval; ✱ means the interval excludes zero, "ns" means it does not. ★ is 5+ paired accounts, ★★ 25+, ★★★ 100+. Cross-account tables are benchmarks, confounded by market, offer and CPM; paired and fixed-effects results are the ones to quote as rules. No cell with fewer than seven accounts is published.

The inbox is where demo asks convert: −57% per lead in the same account

A demo form in a conversation ad completes 42% of the time. The same form on a feed ad completes 3.9% of the time. The inbox costs five times more to reach, and it is still 57% cheaper per demo lead.

This is the one finding that survives every test in the study. Paired across the 23 accounts running demo asks in both placements, inbox is cheaper in 20, median −57% [−69%, −29%]. With account fixed effects across 661 campaigns, −61%. Inside a single account, the best demo campaign is inbox and the worst is feed in 14 of the 15 cases where they differ (sign test p = 0.001), and 11 of 11 on the second-best against second-worst check, the only set-up feature that survives that check. At every robustness threshold tried, from 1 lead per side up to 10, the paired effect is −36% to −57%, n = 21 to 48.

Open rate is the same in both placements; the entire effect is what happens after the open. A form inside the inbox is pre-filled and two taps from done; a form behind a feed ad has already lost the reader to the scroll.

Two cautions. Inbox delivery is one send per person per 30 to 45 days, so an inbox campaign cannot be scaled by raising the budget; it scales by adding people to the pool (chapter 04 on how big that pool should be). And inbox copy, sender and bid amounts are not in the data, so this chapter cannot say which message wins; it says only that the placement does.

Friction in the feed, cost per send and fit in the inbox

Cost per lead is cost per thousand, divided by how many open the form, divided by how many finish it. Which of the three matters depends entirely on the placement.

Take the log of cost per lead and split its variance between accounts into the three parts. In the feed, form completion (friction) explains 51%, CPM 39%, and form-open rate (intent) only 12%. In the inbox, completion is a solved problem at 42% and explains 10%; cost per send explains 46% and whether the right people open the form explains 45%. The feed is a friction-and-CPM game. The inbox is a targeting game.

209 / 144 feed campaigns; 263 / 99 inbox campaigns. In the inbox, winners and losers pay the same per send and get the same open rate. Winners get nearly four times the leads per thousand sends. The whole difference is what happens after the open: the offer, and whether the person opening it is a buyer.

Why retargeting accounts lose: the pool runs out

At the account level, cold audiences produce demo leads at $210 and retargeting at $467. Paired inside one account, retargeting is 10 to 20% cheaper than cold. Both are true, and the second explains the first.

Grouped by the audience type carrying most of the account's demo spend. Cross-account model, holding the rest constant: retargeting share +203% ✱ on CPL, list share +108% ✱.

Paired, it flips

Retargeting is not intrinsically worse per lead. Warm people open the form 23% more often and, campaign for campaign, convert a little cheaper. The accounts that depend on retargeting lose for a different reason.

The reason: the pool is 4,600 people and it gets hammered

The median retargeting audience used for a demo campaign is 4,600 people (p25 1,600, p75 15,000). Lists are 20,000; cold 52,000. Feed retargeting demo campaigns deliver 4 to 16 times their audience size over their life; small ones (under 5,200 people) deliver 15.7x. Every campaign that runs long enough drains its pool, and the cost of the next lead rises as it does. Indexed to each campaign's own first quarter-pass through its audience, cost per lead is 134 at one to two passes, 151 at two to four, and 222 past four. Lists saturate hardest: 323 past four passes. The form-open rate, the intent signal, falls to 66 over the same run.

That is the whole "full-funnel" story in one table. The losers are retargeting-and-list-heavy accounts hammering an 11,000-person pool. The winners run cold at 32,000 to 52,000, or lists in the inbox. Retargeting converts best per person; it just has the fewest people. The fix is not to abandon it; it is to rotate the pool before the fourth pass, and to build a bigger one, which is what chapter 07 says demand generation is actually for.

Opposite rules for feed and inbox

In the feed, bigger audiences are cheaper per demo lead. In the inbox, bigger audiences are dearer. Both hold within account. The reason is the pricing model.

Feed, cold: go broad

Inbox, cold: go tight

The cheapest demo leads in the study are inbox retargeting on pools under 2,300 people: $150 to $214. They are cheap because everyone in the pool has already met the brand; they are few because the pool is tiny. ρ +0.30 ✱ cross-account, +0.35 ✱ within.

Scale, bids, the ask, the format, seniority, region and industry

Diminishing returns are real but not a wall. The word "demo" is the dearest ask. Enhanced conversions are the dearest bid. And the senior floor everyone fears is free.

1. Scale

Within account, with fixed effects, cost per demo lead rises 25% for every e-fold (2.7x) of spend, about 17% per doubling ✱. Within a campaign, month to month, the correlation between spend and CPL is +0.30 ✱. Across accounts the slope is steeper (+66% per e-fold) but confounded: big accounts also run dearer set-ups, enterprise lists, US markets, healthcare. Read it as: doubling demo budget in the same set-up costs 17 to 20% more per lead, which is the in-market pool being drained, not a cliff. The volume tier (C) pays $533 at $8,428 a month; they bought volume with lists in the inbox and accepted the price. Volume comes from adding pools, not from adding spend to one.

2. Bid

3. The ask

4. Format, inside the feed

If it has to be feed, the form goes on a document or a thought leader ad. Winners have fewer creatives per campaign (3.3 vs 5.0) only because their inbox campaigns are single-message; within account, more creatives is cheaper.

5. Seniority, region, industry

A Director-and-above seniority floor is not a penalty. Accounts that set one pay 36% less per demo lead in the cross-account model ✱; within account the paired effect is +17% ns on CPL and +13% ns on CPM. Target senior for demo asks without fear. On region, the US carries a 48% premium ✱ against the rest of the panel (median $431, 49 accounts); the UK reads $667 on eight accounts and is directional only.

6. Free trials, briefly

88 campaigns, 15 accounts, $203,944, 641 trials, 2.5% of demo-type spend. Cost per trial at the 11 accounts with 3+ trials: p25 $73, median $362, p75 $897. Feed trial forms complete at 8.0%. Half the spend is on uploaded lists and the targeted industries skew construction, real estate, manufacturing and financial services: not a SaaS product-led sample. Directional.

The best demo campaign vs the worst, in the same account

Take every account with three or more demo campaigns and compare its best campaign to its worst. The worst costs 4.8x the best. Then ask what the best one has that the worst does not.

73 accounts qualify. Placement, audience type and bid are choices made before the outcome, so reading them off the best and worst campaigns is safe from regression to the mean; the metric gaps (form open +68%, completion 40% vs 22%) are partly mechanical and are not the finding. The set-up shares are.

The second-best/second-worst pass is the check that the pattern is not carried by one lucky and one unlucky campaign per account. Placement is the only feature that survives it. Retargeting (8 vs 4) and cold (7 vs 3) do not.

Note what this table says about retargeting: inside one account, the retargeting campaign is more often the best than the worst. Retargeting converts. The accounts in chapter 03 that lose on it are the ones that run only retargeting on a pool too small to feed the volume they want.

Copy, form design, "intent", and the demand-gen question

Four things people tune that the data does not reward, and one belief of ours the data reframed.

Messaging

641 feed demo campaigns have their post text. Thirteen copy features were tested: hook length, a question in the hook, "you", first person, a number, an emoji, a link, a percentage, social-proof words, pain-point words, whether it says "demo", whether it says "call". Every one is significant across accounts. Not one survives within account (n = 214, all ns). Different advertisers write differently, and their CPL differs for other reasons; the copy is a marker of the advertiser, not a lever. The same pattern as the thought leader ad report. Inbox copy was never synced, so message-ad copy is untested.

Form design

Question count, a phone field, a calendar on the thank-you page, an incentive: nothing significant on cost per lead at this n (phone −28% completion, ns, 12 of 20; short forms −62% completion, ns, 7 of 9; incentive −21% CPL, ns, 8 of 12). Completion is driven by placement, not by the field list. The one finding in Report No. 3 that a phone field costs two thirds of completions was on content forms; on demo forms the effect is smaller and not significant.

"In-market share"

Form-open rate per impression, the closest thing to an intent signal, barely separates winners from losers in the feed (0.48% vs 0.40%) and explains 12% of feed CPL variance. The in-market fraction is real, but it shows up as saturation (the pool drains as you repeat) and as audience size, not as a static property of a segment you can target for.

Demand generation and the price of a demo lead

The full-funnel claim, ours included, is that awareness and engagement spend makes demand capture cheaper. The panel cannot confirm it. 80% of winners, 92% of losers and 100% of the volume tier run $1,000 or more of demand generation (engagement, brand awareness, video views, content lead gen) alongside demo capture, so the only testable question is whether more of it makes the demo lead cheaper. Matched by total-spend band the sign favours demand gen (n = 7–8 per cell). Within account, comparing months after demand-gen spend to months without, demo CPL is +9% [−7%, +65%], cheaper in 11 of 29. No evidence either way at this n.

Accounts that run retargeting demo campaigns. Demand-gen spend correlates with retargeting demo volume (ρ +0.23 ✱) and not, favourably, with retargeting CPL (ρ +0.40, the big accounts run dearer set-ups). Influenced pipeline is not observable here; that needs a CRM join.

So the reframe: demand generation does not lower the price of a demo lead. It raises the ceiling on how many the warm pool can yield, which, after chapter 03, is the constraint that actually binds. Buy it for volume, not for price.

Eleven accounts that get demo leads at volume, at or under the median cost

The top accounts by demo leads at ≤ $350, de-identified. One pattern, and one counter-example worth a case study.

De-identified; a twelfth account at $10 a lead was excluded as an almost certainly misclassified form. Above-median volume months inside an account run at a CPL index of 87 against the account's median; below-median months at 122. Volume and price are not in tension inside a working set-up.

The pattern: conversation ads on a list or a warm pool of 5,000 to 40,000, manual CPM at $0.30 to $0.70 a send, 30 to 50% completion, 10 to 60 demo leads a month at $60 to $250. Eight of the eleven fit it. The counter-example: W2, thought leader ads and video, cold, DACH, asking for a call, $71 a lead on 1,000 leads. It is the feed done the way chapter 05 says the feed has to be done: a person not a logo, a broad cold audience, and the word "call". It is worth a case study of its own.

Twenty numbers to quote, with what they rest on

Each one a full sentence, free to cite with a link. The base line says what it rests on: accounts, whether it is paired, and the star.

Twelve demo-request myths, with the verdict

Terms used in this report

How it was built, how robust it is, what it cannot answer, how to cite

Frequently asked questions about demo requests on LinkedIn

The questions people ask, answered with a number, the n and the method.

How much does a demo request cost on LinkedIn?

$350 at the median advertiser account, across 155 accounts with five or more demo leads in the 12 months to September 2026. The useful answer is the range: a quarter of accounts pay under $161 and a quarter pay over $754. The pooled figure (all spend ÷ all leads) is $407 and is dominated by the largest advertisers.

What is a good cost per demo on LinkedIn?

Under $161 puts you in the top quarter of accounts; under $71 in the top tenth. The accounts that get there run 86% of demo spend in the inbox, half of it to cold audiences of 30,000 to 50,000, and get 17.6 demo leads a month on $2,168.

How many demo leads a month should a LinkedIn campaign produce?

6.6 at the median account on $2,142 a month; 17 at the 75th percentile; 44 at the 90th. Volume comes from the number and size of pools, not from budget: doubling spend in one set-up costs 17 to 20% more per lead.

Are conversation ads or feed ads better for demo requests?

Conversation ads, by 57% per lead in the same account (23 accounts running both, inbox cheaper in 20). The inbox costs five times more per thousand but the form completes 42% of the time against 3.9% in the feed. If the form has to be in the feed, put it on a document ad (−80%) or a thought leader ad (−32%), not a single image.

Is retargeting cheaper than cold for demo leads?

Per lead inside one account, slightly: −10% to −20%, not significant. At the account level, accounts that depend on retargeting pay $467 against $210 for cold, because the median retargeting pool is 4,600 people and gets shown ads 4 to 16 times over; past four passes cost per lead is 2.2x where it started. Use retargeting, and rotate the pool.

How big should the audience be for a LinkedIn demo campaign?

It depends on the placement. Feed: 100,000 or more; 250,000+ audiences get a demo lead for $76 to $87, under 30,000 for $232. Inbox: under 40,000; under 10,000 gets a lead for $173, over 170,000 for $635. Inbox retargeting on pools under 2,300 people is the cheapest demo lead in the study at $150 to $214.

Does "request a demo" or "book a call" convert better on LinkedIn?

Book a call: $342 a lead at the median account against $611 for a demo request; an audit or assessment $260; a free trial $362. Cross-account benchmark on 41 / 104 / 22 / 11 accounts, not significant within account, so treat it as a strong hint.

Should I use enhanced conversions on LinkedIn lead gen?

No. Accounts on enhanced-conversion bidding pay $515 a demo lead at the median against $342 on manual CPM, and 47% of them sit in the expensive-and-low-volume tier. Max delivery vs manual, paired, is +189% on seven accounts.

Does running brand awareness or engagement campaigns make demo leads cheaper?

Not in this data. Within account, demo CPL in the months after demand-gen spend is +9% [−7%, +65%], cheaper in 11 of 29 accounts. Demand-gen spend correlates with how many demo leads the retargeting pool yields (ρ +0.23), not with what they cost. It raises the ceiling; it does not lower the price.

Does senior targeting make demo leads more expensive?

No. Accounts with a Director-and-above floor pay 36% less per demo lead in the cross-account model; within account the difference is +17%, not significant. The senior filter costs nothing in CPM (+13%, ns).

What does a demo lead cost in the US vs the UK?

US $431 at the median of 49 accounts, a 48% premium against the rest of the panel. UK $667 on eight accounts, directional only. By targeted industry: financial services $283, manufacturing $308, software development $429, hospitals and health care $829.

How is this different from the lead gen forms report?

Report No. 3 covers every lead gen form, content and demo side by side, across 489 advertisers. This report takes only the demo-request half (268 advertisers, $8.25M), goes deeper on placement, pool size and saturation, and adds the within-account best-vs-worst and demand-gen tests. Where they overlap the numbers agree; where they differ in scope, cite each for its own.