LinkedIn's targeting is the best in the business. Nowhere else can a small company put its ad specifically in front of, say, operations directors at 50-to-200-person manufacturers in the Midwest. That precision is the whole reason LinkedIn is worth its famously high prices. It's also the thing that quietly wastes more small-business budget than anything else on the platform, because your audience can be wrong in two completely opposite directions, and LinkedIn's own default settings nudge you straight toward the one that costs the most.
On one end, your audience is too broad: a giant, loosely-defined pool full of people who will never buy, and you're paying LinkedIn's premium rates to reach every one of them. On the other end, it's too narrow: you've stacked so many filters that only a few hundred people qualify, and the campaign can barely run at all. The money leaks either way. The skill, and it is the single most important skill on LinkedIn, is landing in the middle: an audience tight enough that nearly everyone in it is a real prospect, but big enough that the campaign can actually spend and learn. Here's how both failures happen, and how to find that middle.
Too Broad: The Direction LinkedIn Pushes You
Start with the broad end, because this is the one LinkedIn actively steers you into. The platform makes most of its money when you spend your whole budget, and the fastest way to spend a budget is on a big audience. So several of its defaults are quietly set to inflate yours.
The worst offender is a checkbox called Audience Expansion, and it's turned on by default in every campaign you create. What it does is let LinkedIn add people to your audience who don't actually match the targeting you carefully set, on the theory that they're "similar." One of the most respected LinkedIn ads specialists calls this feature the closest thing the platform has to a virus, because you are paying a premium to reach exactly the people you specified, and this quietly overrides you to reach people you didn't. Uncheck it, every time.
Next to it lurks a second one, the LinkedIn Audience Network, also usually on. This takes your ads off LinkedIn entirely and scatters them across a network of random third-party apps and websites. The impressions are cheaper, which sounds good until you realize you're now paying for your B2B ad to appear who-knows-where, to who-knows-whom. For a small business trying to reach a specific kind of buyer, turn it off and keep your spend on LinkedIn itself, where the targeting actually applies.
Then there's the human nudge. If you talk to a LinkedIn sales rep, they will very likely tell you your audience should be at least 300,000 people, maybe more. Understand what that advice is really for: it makes sure you can spend a large budget, which is their goal, not necessarily yours. The advertisers who genuinely need audiences that large are Fortune 500 brands with money to burn. A small business does not need, and usually should not have, a 300,000-person audience. Aim it at the people who would actually make good customers, and ignore the pressure to bloat it.
The reason all of this matters is that a too-broad audience doesn't just waste money in the abstract. You feel it twice: once when your reports show a fine-looking cost per click that never turns into real conversations, and again a few weeks later when your sales team starts complaining that the leads are junk. Broad targeting on a premium platform is the most expensive way to discover who your customer isn't.
Too Narrow: The Over-Correction
Now the opposite failure, which tends to happen right after someone reads a warning like the one above and over-corrects. LinkedIn's targeting options are so rich and so tempting that it's easy to keep stacking them. You start with a job title. Then you add a seniority level. Then a company size. Then an industry. Then a specific skill. Then you narrow it to one metro area for good measure. Each filter feels smart in isolation, and by the end you've built an audience of four hundred people.
An audience that small simply can't work, for mechanical reasons. LinkedIn won't even run a campaign below a certain audience floor, and hovering just above it isn't much better. There aren't enough people for the system to find the responsive ones and learn who to prioritize, so it never optimizes. The same handful of people see your ad over and over, get tired of it fast, and your costs climb as the platform strains to keep serving a tiny pool. Your budget trickles out slowly and you conclude, wrongly, that "LinkedIn doesn't work for us." LinkedIn worked fine. The audience was strangled before it had a chance.
There's a subtler version of this trap too: over-segmenting by geography on a small budget. Slicing your audience into separate campaigns for each city or region feels precise, but every slice divides your already-modest budget into a piece too small to do anything. Unless you're spending heavily, keep the geography wide, one country or a broad region, and let the budget stay concentrated enough to actually produce data.
Finding the Right Size
So the target is the middle: an audience where nearly everyone is a genuine prospect, yet large enough for the campaign to breathe. A few principles get you there.
First, resist the urge to stack every filter. Pick the two or three that truly define your buyer and stop. If operations directors at mid-sized manufacturers are your people, then "job title" plus "company size" plus "industry" probably describes them completely. Adding a seniority filter and a skills filter and a specific-city filter on top doesn't make the audience more accurate, it just makes it smaller. Every filter you add is a bet that everyone who doesn't match it is worthless, and that bet is usually wrong.
Second, do your narrowing with exclusions rather than more inclusions. This is the move most small businesses miss. Instead of piling on inclusion filters to tighten the audience, keep your inclusion targeting relatively simple and then subtract the people you don't want: exclude your competitors, exclude your current and past customers so you're not paying to advertise to people who already know you, and exclude the obvious out-of-scope job titles that LinkedIn's imperfect data will otherwise sweep in. Exclusions sharpen quality without shrinking your reach into the danger zone.
Third, size the audience to your budget, not the other way around. A bigger budget can support and reward a tighter, more segmented audience, because there's enough money for each segment to gather real data. A smaller budget needs a broader, more consolidated audience so the spend isn't spread too thin to matter. Let how much you're spending decide how finely you're allowed to cut.
If you want a rough target to aim for rather than a rule to obey, most small-business LinkedIn campaigns are healthiest with an audience somewhere in the tens of thousands, not the hundreds and not the hundreds of thousands. That's specific enough that the people in it genuinely resemble your buyer, and large enough that the campaign can deliver consistently and actually learn who responds. If you find yourself down in the low thousands or below, that's your signal to loosen a filter; if you're up past a few hundred thousand, that's your signal to tighten one, or more likely to turn off the expansion settings inflating it. The exact number matters less than the habit of checking it and asking, honestly, whether everyone in there is really a prospect and whether there are enough of them to run.
Here's what that looks like in practice. Say you sell CNC machine tooling and your buyers are operations and maintenance leaders at mid-sized manufacturers. The over-stacked version goes: maintenance-and-operations job titles, AND manager-level-or-above seniority, AND 51-to-500 employees, AND the industrial-manufacturing industry, AND the specific skill "CNC machining," AND located in Ohio. Each filter felt reasonable, and together they've produced an audience of maybe three hundred people that LinkedIn can barely serve. Now rebuild it: keep the job titles, the company size, and the industry, because those three genuinely describe your buyer. Drop the seniority filter (the titles already imply it), drop the skill filter (it's redundant with the titles and it's the kind of profile data LinkedIn gets wrong constantly), and widen Ohio out to the broader Midwest. Then add exclusions: cut your two main competitors and your existing customers. You've gone from three hundred strangled people to a healthy audience in the tens of thousands, every one of them a plausible buyer, and the campaign can finally run and learn. Same targeting instinct, right-sized.
A couple of smaller settings are worth getting right while you're in there. Set your location targeting to "permanent," so you're reaching people who actually live and work in your area rather than someone passing through on a trip. And when you do want to reach a full buying committee, the pain-feelers and the decision-makers who sign off, don't cram them all into one audience. Split them into separate campaigns so you can see which group actually responds, an idea we've written about in why one LinkedIn campaign can't do three jobs at once.
Precision Is a Tool, Not a Goal
The thing to hold onto is that LinkedIn's incredible targeting is a tool, and like any tool it can be used too little or too much. We've written before about how remarkable it is that you can put your ad in front of exactly the companies you want, and that power is real. This is simply the other half of that lesson: the goal was never maximum precision, and it was never maximum reach. It's the deliberate middle, and getting there means overriding the defaults that push you broad and resisting the temptation to slice yourself narrow.
Turn off audience expansion and the audience network. Ignore the pressure to hit some giant audience number. Use two or three real filters, sharpen with exclusions, and size the whole thing to what you're actually spending. Do that, and LinkedIn's targeting finally becomes what it's supposed to be: the reason your ads reach the right people, instead of the reason your budget quietly disappears. If you'd like a second set of eyes on whether your audience is sized right, that's a core part of what we do with LinkedIn ads at BrandRocket. Either way, the fix is in your hands, and it starts with a single audience that's neither 300,000 nor 300.
