Every Guide Starts With a Step You Cannot Take
Open any article about LinkedIn retargeting and the first instruction is the same. Install the Insight Tag. Put LinkedIn's tracking code on your website, wait for visitors, and once enough of them have come through you can start showing ads to people who already know who you are.
That advice was written for a company with traffic. If your site sees a few hundred visits a month, and a fair number of those are you, your bookkeeper, and somebody's cousin checking whether you are a real business, the instruction quietly fails.
It fails in a way that is easy to miss, too. Website audiences do not work retroactively. The tag does not reach back and gather up the people who visited last year. It begins counting the day you install it, and from that day it fills at exactly the speed strangers happen to find you. For most small firms, that speed is close to nothing.
So the honest version of the standard advice reads more like this. Install this code, wait several months, and you may still not have enough people to run a single ad.
There is a better route. It has been sitting in the same menu the entire time.
The One Number That Decides Whether Any of This Works
Before the strategy, the constraint.
A LinkedIn retargeting audience cannot be used to target a campaign until it holds 300 member accounts. Under 300, the audience exists, the campaign exists, and nothing serves. That floor is the same whichever kind of audience you build.
Which means every discussion about retargeting on a small budget is really one discussion. How do you get to 300?
Website traffic is one way to get there. It is simply the slowest one available to you.
Eight of LinkedIn's Nine Retargeting Audiences Never Touch Your Website
Here is the part nobody puts in front of a small business.
LinkedIn builds retargeting audiences from nine different sources, and exactly one of them has anything to do with your website.
You can build an audience from people who engaged with a single image ad. From people who opened or paged through a document ad. From people who watched a quarter, a half, three quarters, or essentially all of a video. From people who opened your lead gen form, and separately from the people who actually submitted it. From people who opened a conversation ad or clicked the button inside it. From people who viewed your LinkedIn company page or clicked the button in its header. From people who registered for or attended a LinkedIn event. And from server-side signals sent through the Conversions API.
One item on that list needs the Insight Tag. The website one. LinkedIn's own documentation puts it plainly, saying you will need to install the Insight Tag on your site to enable website retargeting. It says that about website retargeting because website retargeting is the only place it applies.
Every other audience is assembled from things people did on LinkedIn, in the feed, where LinkedIn was already keeping count. No tag. No developer. No website at all, if that is where you are starting from.
You Are Not Waiting for an Audience. You Are Buying One.
This is the shift the whole article rests on, and it changes what that 300 floor means.
A website audience fills at the rate people find you. You do not control that rate. You can improve it over quarters with content, search and referrals, but you cannot decide on Monday that you would like more of it by Thursday.
An engagement audience fills at the rate you pay for impressions. That rate you do control. Every time your video plays and somebody watches half of it, you have added a member. Every time somebody taps to read the rest of your post, you have added a member. The audience stops being a thing you wait for and becomes a thing you buy, at a price you can look up in your own reporting.
AJ Wilcox, who has run LinkedIn ads through his agency since 2014, gave that price a name. Cost per retargetable member. Take what a campaign spent, divide it by the number of people it made retargetable, and you have the real cost of building your warm audience. Not cost per click, which measures something you may not want. Cost per person you have earned the right to speak to again.
Once you can measure that, picking an ad format stops being a matter of taste.
What a Warm Audience Actually Costs to Build
Wilcox publishes the figures his agency sees, and they need reporting carefully. These are one agency's results across their own accounts, not a rate LinkedIn promises you. Your account will tell you your own number within two weeks, which is the whole point of having the metric.
His best all-round performer is a video thought leader ad, meaning a video published from a real person's profile and promoted with ad budget. He puts the cost per retargetable member between fifty cents and a dollar. The reason it wins is arithmetic rather than magic. Video is cheap to serve, and somebody who watched half of a forty-second video has given you the better part of twenty seconds of attention.
A static thought leader ad comes in around a dollar, because he bids roughly a dollar per engagement and an engagement is the thing that makes someone retargetable.
The cheapest item on his list, by a wide margin, is a document ad run on the website visits objective, where he reports getting the cost down to two or three cents a person. That happens because of a quirk worth understanding. On that objective you are charged when somebody clicks through to your site, but somebody becomes retargetable when they flip to the next slide. Documents are engaging and the link underneath them is not, so a lot of people take the action that costs you nothing and very few take the action that costs you money.
Put the arithmetic together at the top of that range. At fifty cents to a dollar a head, clearing 300 members costs somewhere between a hundred and fifty and three hundred dollars. Not per month, and not forever. Once, to own a warm audience that did not exist before you started.
Set that against waiting two quarters for website traffic you are not certain is coming.
Cheap Is Not the Same as Warm
Now the correction, because a table of prices will lead you badly astray if you read only the right-hand column.
Two or three cents a head looks like the obvious winner until you ask what that person actually did to earn the label. They advanced one slide of a document. That is a flick of a thumb. They may not have taken in your name, your offer, or what you sell.
Wilcox says as much about his own cheapest tactic, and it is the most useful admission in the whole subject. He questions the quality of that interaction, and when he builds a funnel out of document ads he adds more stages to it, because it takes several shallow touches to add up to one real one.
The same caution applies to static engagement. An engagement on LinkedIn can be a like, a comment, a click on your profile, or a click on "see more" to finish reading the post. Most of them are that last one. Somebody who tapped to finish a paragraph is not the same prospect as somebody who chose to watch you talk for twenty seconds.
So read both columns at once. Price tells you how quickly you can fill the audience. Depth tells you what the audience is worth once it is full. A video viewer costs perhaps twenty times what a document flipper costs, and is worth a good deal more than twenty times as much.
How to Cross 300 Sooner Than You Think
Three practical moves, all of them from running this at volume rather than from theory.
Build the audiences before you need them. An audience only starts collecting once it exists, exactly like the tag does, so an audience created today captures engagement from today onward. Create every audience you could plausibly want on the day you launch, including the ones for stages you have not built yet. They cost nothing to sit there filling.
Create the campaigns early as well. A campaign aimed at an audience that has not yet reached 300 does not throw an error and does not need watching. It sits still, and it starts serving the moment the audience crosses the line. You are not staring at a dashboard waiting to flip a switch by hand.
Then combine audiences to get over that line. This is the move that makes the whole thing workable on a small budget. If you have 100 website visitors and 200 company page visitors, neither audience can run on its own. Put both into the same campaign and you have 300, and it serves. The threshold is checked against the combined audience, not against each ingredient separately.
Guard Rails, and Letting People Move On
Two things go wrong once this starts working.
The first is who you catch. Not everybody engaging with your ads is a buyer. Job seekers research companies they would like to work for. Salespeople research companies they would like to sell to. Both will happily watch your video, and both will land in your retargeting audience and spend your budget sitting there. The fix is to layer ordinary targeting on top of the engagement audience. Job function, seniority, company size. You are telling LinkedIn to retarget the engaged people who also look like customers, rather than everyone who was curious. It also keeps the audience from being spent on people who will never buy, which matters more here than usual, because a LinkedIn audience this size wears out quickly.
The second is people getting stuck. When you build an audience out of the people who engaged with your first campaign, you have to exclude that audience from that same campaign. Otherwise the person who watched your video keeps being shown the video they already watched, at the same time as the follow-up you built specifically for them. You pay twice to reach one person, and your own two campaigns bid against each other in the same auction. Every audience you create should be excluded from the campaign that feeds it, on the day you create it. This is the same discipline as keeping one campaign to one job: the moment two campaigns are chasing the same person, you are bidding against yourself.
The Part That Turns This Around
There is an assumption sitting underneath all of the above, and it deserves to be challenged, because it is the reason website retargeting has the reputation it does.
The assumption is that a website visit is the strongest signal a stranger can hand you. It is not, particularly. A visit tells you somebody loaded a page. It does not tell you why, whether they read a word of it, or whether they meant to be there at all.
The agency PipeRocket Digital published a short video this month explaining why they stopped retargeting website visitors on LinkedIn altogether. Their argument was that undifferentiated traffic carries no real buying intent, so the spend chasing it never traced back to pipeline and ended up written off as influence. They moved to targeting accounts that were showing genuine signals instead.
You can disagree with where they landed and still take the point. Somebody who chose to watch two thirds of your video made a decision. Somebody who opened your lead form and stopped made a decision, and told you roughly where they stopped. Somebody who bounced off your homepage in four seconds made no decision whatsoever.
Which puts the small business in an unexpected position. Without website traffic you are not settling for a lesser version of retargeting. A good deal of the time you are skipping the weakest signal on the menu and going straight to the ones that carry meaning.
Where to Start
If you are starting from nothing, the sequence is short.
Create every retargeting audience LinkedIn will let you create, today, so that they begin collecting. Run one campaign built around a format that produces depth rather than cheap volume, which for most small firms means video from a real person's profile rather than from the company page. Watch what it costs you per retargetable member for two weeks, using your own numbers rather than anyone else's. Then, once the audience clears 300, put your actual offer in front of it.
That is the entire play, and it runs on a few hundred dollars rather than a few thousand.
It is also, in our experience, the part of a LinkedIn account most likely to have never been set up at all. Building it and keeping it running is a good deal of what we do when we run LinkedIn Ads for a small business. If you would rather build it yourself, everything above is what you need.
