LinkedIn Ads for SaaS

A LinkedIn click costs roughly what a sandwich costs. That is the entire problem with the channel, and it explains why so many SaaS teams run it for one quarter, stare at a 340 dollar cost per lead, and move the money to Google.

Teams that keep LinkedIn are not getting cheaper clicks. They accepted the price and changed what they buy with it. This page is the build order, the numbers to plan against, and the specific settings that separate a campaign that teaches you something from one that donates 6,000 dollars to Microsoft.

What does LinkedIn actually cost for a SaaS advertiser?

Plan for 8 to 20 dollars a click, 50 to 110 dollars per thousand impressions, and 120 to 600 dollars a lead. Those are the ranges I see across B2B software accounts, and where you sit inside them is decided almost entirely by three inputs: seniority, geography and how much the offer asks for.

Ask for a demo from a VP of Engineering in San Francisco and you are at the ceiling. Offer a benchmark report to marketing managers in Spain and you are at the floor.

Treat those as planning ranges from working accounts rather than a published study. The segmented version, with sample sizes and the split between demo requests and content downloads, sits in the LinkedIn Ads cost benchmarks for SaaS research page.

Two costs never appear in Campaign Manager and both are bigger than people expect. Creative production runs 2,000 to 6,000 dollars a month once you are refreshing properly. Sales time spent disqualifying content leads can quietly exceed the media cost itself.

Prices also move with the calendar. The auction tightens in the last six weeks of the year as enterprise advertisers dump unspent budget, and it loosens through July and August. A 12 dollar click in March can be a 17 dollar click in late November against the same audience, which matters if you build an annual plan on a single quarter of data.

Build the audience from function, seniority and headcount

Job title targeting is the single most common way SaaS teams waste LinkedIn budget. Titles are self-reported, they vary enormously between a 40-person startup and Salesforce, and LinkedIn has to map thousands of freeform strings into standardised buckets. Every mapping miss is a buyer you cannot reach and a bill you still pay.

The construction that holds up across accounts is three filters stacked: job function, seniority, company headcount. Add industry only if your product genuinely does not sell outside a few verticals.

• Job function: Engineering, Information Technology, Marketing, Finance, Operations
• Seniority: Manager, Director, VP, CXO, Owner
• Company headcount: 201 to 500, 501 to 1,000, 1,001 to 5,000

That stack produces a bigger audience than a title list, prices better because the auction is less thin, and survives LinkedIn’s data changes. Titles then belong in one place only: exclusions. Strip out students, recruiters, consultants and anyone at an agency in your category if you do not sell to them.

Three other audience sources do real work once the basics run.

Company lists. Upload a target account list into Matched Audiences and combine it with the function and seniority filters. A 1,200-account list crossed with three functions and four seniority levels usually lands in the 40,000 to 90,000 range, which is workable. Account-list mechanics, tiering and the message ladder are covered properly in the LinkedIn ABM advertising playbook.

Website and engagement retargeting. The Insight Tag builds page-level pools, and LinkedIn also lets you retarget video viewers by completion percentage, lead form openers who did not submit, and people who engaged with a single image ad. Video-viewer retargeting at 50 percent completion is one of the better warm pools available on the platform.

Contact lists from the CRM. Closed-lost opportunities from 6 to 18 months ago are an underused audience. They know the category, they evaluated you once, and match rates on business email are tolerable if you also include any personal emails you legitimately hold.

How big should a LinkedIn audience be?

Between 50,000 and 300,000 members for prospecting. Below 20,000 you pay a visible CPM premium, frequency climbs within days, and any budget above a few hundred dollars a week burns the pool. Above 500,000 you are paying to reach people who cannot buy, and the reporting stops telling you anything about who responds.

Retargeting is the exception. Those pools are small by nature, and the 300-member floor LinkedIn enforces is usually the binding constraint. Accept high CPMs there, because the people are worth more.

Audience size	What happens	Spend it supports	Verdict
Under 20,000	CPM premium, frequency hits 8 plus per week inside a fortnight	Under $1,500 a month	Only for named-account ABM with matching creative
20,000 to 50,000	Workable but fatigues fast, needs more creative variants	$1,500 to $5,000 a month	Fine for a tight ICP, plan refreshes every four weeks
50,000 to 300,000	Stable delivery, readable segment reporting	$5,000 to $60,000 a month	The window to aim for on prospecting
Over 500,000	Cheap clicks, unreadable quality, delivery skews junior	Any	Too broad. Add a headcount or seniority filter

A practical trick: build the audience, then check the forecast panel for the seniority split. If more than 40 percent of the forecast sits at entry level, your filters are not doing what you think they are.

Which ad format belongs at which stage?

Format choice is a stage decision, not a taste decision. Cold audiences need something that earns a second of attention and costs little to produce. Warm audiences need something that closes a loop.

Format	Best stage	What it does well	Watch out for
Single image	Cold prospecting	Cheapest to produce, fastest to test, works at volume	Fatigues fastest. Needs 4 to 6 live variants
Document ads	Cold prospecting	In-feed preview of a PDF, strong dwell time, gated or ungated	Ungated versions produce no leads, only attention
Video	Mid-funnel	Builds retargeting pools by completion percentage	Production cost. Under 30 seconds or nobody finishes
Thought leader ads	Mid-funnel	Lower CPC and higher engagement than page ads	No lead gen forms attached. Optimise for clicks
Conversation ads	Warm retargeting	High open rates on people who know you	Not available to EU members. Feels intrusive if cold
Text and spotlight ads	Warm retargeting	Pennies per impression, keeps the brand present	Tiny click volume. Support act, never the lead

Document ads are the most underused option in B2B SaaS. A six-page teardown of how 40 companies structure their pricing page, previewed in the feed with no form, will out-dwell any single image you produce this quarter. Run one gated version for pipeline and one ungated version for reach, then retarget everyone who opened the ungated one.

Thought leader ads deserve their own treatment because the mechanics are different: the post lives on a person’s profile, the permission flow runs through Campaign Manager, and you cannot attach a lead form. The full setup is in LinkedIn Thought Leader Ads for SaaS.

On copy, the pattern that keeps winning is boringly consistent: name the audience in the first line, state the problem in the second, promise a mechanism in the third. Brand adjectives lose to specificity every single time. There are 60 tested variants in the SaaS ad copy templates if you want a starting point rather than a blank editor.

What should you actually offer a cold LinkedIn audience?

Something worth 20 seconds, not something worth 30 minutes. The size of the ask has to match how much the person knows about you, and on a cold audience that is precisely nothing.

The closed-lost row is the one teams skip and should not. An ad that says you shipped SSO, SOC 2 and a Snowflake connector since they last looked beats any clever cold concept, because it answers the specific objection that lost the deal. Pull the loss reasons from the CRM, group them, and write one ad per reason.

Live sessions need an honest correction applied to the numbers. Registration might cost 220 dollars, but attendance on a paid-acquisition webinar typically runs 25 to 40 percent, so your real cost per attending human is closer to 600 dollars. Report that number, not the registration number, or you will scale something that looks three times better than it is.

Bidding: manual CPC first, cost cap later

Start on manual CPC and stay there for 60 days. Maximum delivery hands LinkedIn permission to spend your full daily budget regardless of what inventory costs that day, and on a new campaign with no conversion history it will find the most expensive impressions available.

Set your manual bid around 60 to 70 percent of the low end of LinkedIn’s suggested range, then raise it in 10 percent steps until the campaign delivers roughly its daily budget. You will usually land well under the suggested bid. That gap is the money most accounts hand back.

Move to cost cap once a campaign produces 15 or more conversions a week for three consecutive weeks. Below that volume the algorithm cannot hold a target and you get erratic delivery with a confident-looking dashboard.

What a LinkedIn program looks like at three budget levels

Five thousand dollars a month buys you an answer, not a program. That is the honest floor: two audiences, one offer, eight weeks, enough clicks to know whether your ICP responds at all.

Below 5,000 dollars, the honest recommendation is to spend it somewhere else. Search capture and a good retargeting setup will produce more for a small budget, and Google Ads vs LinkedIn Ads for B2B SaaS walks through where the crossover point sits for different ACV bands. If you want cheap reach against technical audiences instead, Reddit ads for SaaS covers a channel where 3,000 dollars actually buys something.

Frequency, fatigue and the creative refresh cadence

LinkedIn audiences fatigue faster than most advertisers plan for. Between four and six impressions a week, click-through rate starts falling and cost per click starts climbing, and the two together can double your effective CPL inside a fortnight without anything else changing.

Check the average frequency figure in the demographics tab weekly. If it passes six, one of three things is true: the audience is too small, the budget is too high for the pool, or you have too few live variants. Fix it in that order.

Plan a creative refresh every four to six weeks on any always-on campaign. That means four to six new concepts a month at 15,000 dollars of spend, not four a quarter. Refresh means a genuinely different angle, not a new background colour on the same headline.

How to measure LinkedIn without cancelling it by accident

Last-click cost per lead is the wrong primary metric for this channel, and reporting it as the primary metric is how good LinkedIn programs get killed. The channel’s job above 25,000 dollars ACV is to make your category and your name familiar to a buying committee that will search for you later.

Four measurements that survive scrutiny from a CFO:

• Self-reported attribution on the demo form. A single open field asking how they heard about you, compared monthly against platform-reported conversions. LinkedIn is consistently under-credited in platform data and over-represented in self-report.
• Branded search volume trend, pulled from Search Console and segmented by the regions you advertise in. A demand-creation channel should move it within 90 days.
• Pipeline created within 90 days of first ad exposure, matched at the account level rather than the person level. The person who clicked is rarely the person who signs.
• Win rate on accounts exposed to ads versus matched unexposed accounts. Crude, but it is the comparison a board understands.

Self-reported attribution needs one design rule to work: make the field open text, not a dropdown. Dropdowns force people into whichever option sits at the top, and the top option on most SaaS forms is Google. An open field produces messier data and truer data, and you can bucket the answers in a spreadsheet in ten minutes a month.

Two reporting traps worth naming. The default view-through window inflates conversions, so set it to one day and compare. And Campaign Manager will happily report a lead form fill as a conversion while your CRM records nothing, because the form fired but the integration silently failed. Check both numbers monthly. That failure and several like it are catalogued in SaaS PPC mistakes that waste budget.

The pre-launch audit that saves the first 3,000 dollars

Most of the money wasted on LinkedIn is wasted in week one, on settings nobody checked. Run this list before the campaign goes live, and again whenever someone new touches the account.

Half of those take under a minute. The two that reliably cost real money are audience expansion and the seven-day view window, and both are on by default.

Where LinkedIn belongs in your paid mix

For a product above 25,000 dollars ACV selling to a committee, LinkedIn reasonably takes 25 to 40 percent of paid budget, weighted toward demand creation and account coverage rather than form fills. For a product between 10,000 and 25,000 dollars, cap it near 15 percent and run retargeting plus a narrow ABM overlay only. Below that, use search, review sites and retargeting, and keep LinkedIn for hiring.

The comparison against every other paid option, including the marketplaces that often beat both Google and LinkedIn for late-stage intent, is laid out in SaaS advertising platforms compared, and the broader sequencing question sits in the SaaS PPC and Paid Ads hub.

Start here this week. Rebuild one prospecting audience on function plus seniority plus headcount, check that it forecasts between 50,000 and 300,000 members, switch the campaign to manual CPC at 60 percent of the suggested floor, and turn off audience expansion and the Audience Network. Then set a calendar reminder for day 30 to pull cost per qualified opportunity from the CRM. Those four changes are worth more than any creative you could write this month.