You are in Campaign Manager, on the targeting step. You set a company size range. You add a couple of job titles. And somewhere in the middle you open the industry dropdown, scroll to the one that looks like your customers, and check it.
That box does not mean what you think it means.
It is not a description of the person you are about to pay to reach. It is not even a description of what that person does for a living. It is a label their employer picked for itself, once, from a menu, and every single human being who works there inherited it. The receptionist, the warehouse lead, the CFO, all of them.
On a platform where a single click can cost many times what it would on Meta, the gap between what that box appears to say and what it actually does is worth real money. So let's look at what you are buying when you check it.
Where That Label Actually Comes From
LinkedIn is unusually direct about this in its own documentation, which is probably why nobody reads it.
Company Industry is defined there as "the primary industry of the company where the member account is employed, as stated by the company." On the member side, the wording is just as plain: "by default, member account industries are derived from the member account's current company."
Read those twice, because two things are hiding in them.
First, "as stated by the company." The industry on a company's page was chosen by whoever built that page. LinkedIn's developer documentation lays out the three ways a company ends up with an industry: someone at the organization creates the page and assigns it, LinkedIn manually annotates it for very large organizations, or LinkedIn's own models infer it from the wider economic graph. For almost every company a small business actually wants to reach, it is the first one. An admin, on one afternoon, picked the closest-looking option from a long list and moved on.
Second, "derived from the member account's current company." The industry attached to a person is not something that person decided about themselves. They inherit it from their employer. Only when the company has no industry at all does LinkedIn fall back to the industry on the individual's own profile. Otherwise the employer's label wins, for everybody on the payroll.
The Receptionist at a Software Company Is "Software Development"
Once you can see the mechanism, the consequence is obvious and slightly alarming.
When you target Software Development, you are not reaching software people. You are reaching everyone employed by companies whose page says Software Development. That includes the engineers. It also includes the office manager, the recruiter, the accounts payable clerk, the facilities coordinator and the two salespeople.
If you have layered a real job function and a seniority level on top, most of those people are filtered back out and no harm is done. If industry is doing heavy lifting in your audience, they are not filtered out at all, and you are paying LinkedIn's premium rate to put your ad in front of a receptionist because of a dropdown her employer's marketing coordinator touched in 2019.
This is the same shape of problem as the job title trap, which we covered in You Targeted the Job Title. LinkedIn Only Knows Half of Them. Both come from one root: a targeting option that sounds like it describes a human being, and in fact describes a data field. The difference is that most owners have at least heard job title targeting is imperfect. Almost nobody questions industry, because it looks like the safest box on the screen.
The List Is Not a List. It Is a Tree
Here is the second thing the interface hides from you.
LinkedIn's industry taxonomy is not a flat menu of equal options. It is a hierarchy, hundreds of entries deep, where broad parent groups sit above increasingly specific children. Construction is a parent. Underneath it sit Building Construction, Civil Engineering and Specialty Trade Contractors. Underneath those sit narrower entries still, down to things as specific as Highway, Street, and Bridge Construction.
In the targeting panel, a parent and a child look identical. Both are a line of text with a checkbox next to it. But selecting the parent can pull in an order of magnitude more companies than selecting the child, and nothing on the screen tells you which kind of thing you just clicked. Owners routinely check the broad group, watch the audience estimate jump, assume the number is healthy because it is big, and go live against a category that contains half the economy.
The reverse mistake is quieter and more expensive. Your ideal customer may not exist in the taxonomy at all. LinkedIn's categories were built to classify the whole labor market, not to match how you segment your buyers. If you sell to independent physiotherapy practices, or to regional food distributors, or to firms that do commercial landscaping, there may be no entry that means that. You end up checking something adjacent and hoping, which is a fine thing to do knowingly and a terrible thing to do by accident.
Three Ways It Quietly Misses
Put the mechanism and the taxonomy together and you get three specific failures, all of which cost money without ever showing up as a problem in your reporting.
One company, several businesses, one label. An organization that runs a manufacturing arm, a logistics arm and a services arm still gets a single primary industry. Every employee across all three inherits it. If you are trying to reach the services side of a company that classified itself as manufacturing, industry targeting will not find them, and if you are trying to avoid the manufacturing side of a company that classified itself as services, it will not exclude them.
Chosen once, rarely revisited. Company pages are set up early, usually by someone whose job that week was to get the page live. Businesses then pivot, add product lines, move upmarket and change what they sell. The industry label is not something anyone gets a reminder to review. You are frequently targeting a description of what a company used to be.
Your category is not their category. The label answers the question "what does this organization sell," which is not always the question you need answered. You often care about how they buy, how big their team is, whether they are regulated, or whether they have the specific operational problem you solve. Industry is a proxy for those, and sometimes a poor one.
None of this makes industry targeting broken. It makes it coarse. Coarse is fine, as long as you know it is coarse and build accordingly.
Use It as a Filter, Never as the Spine
Here is the rule that follows from all of it. Industry should narrow an audience you have already defined some other way. It should almost never be the thing that defines the audience.
What carries the audience instead depends on one question: can you name the companies you want?
If you can, name them. A list of target accounts uploaded to LinkedIn is more precise than any industry selection, because it contains exactly the organizations you chose and nothing else. It does not go stale in the same way, and it does not depend on how each of those companies once described itself. We went through how to build and use that in LinkedIn Can Put Your Ad in Front of the Exact Companies You Want. For a business with a defined market of a few hundred realistic buyers, this is nearly always the better spine.
If you cannot name them, build the audience from the person and the company shape. Job function plus seniority describes what someone actually does and how senior they are. Company size describes whether the organization can plausibly afford you. Those two together do most of the qualifying work. Then, and only then, add industry to trim what is left.
The practical test is simple. Remove industry from your audience and look at what is left. If the remaining targeting still describes a sensible buyer, industry was doing its proper job as a filter. If removing it leaves you with something meaningless, industry was holding the whole audience up, and it is not strong enough to do that.
If You Really Do Sell Into Several Industries, Split Them
There is one place where industry earns a much bigger role, and most owners get it exactly backwards.
If your business genuinely serves several distinct industries, the instinct is to check all of them in one campaign, add every relevant job title, and let a comfortably large audience build up. Anthony Blatner, who has audited a great many LinkedIn accounts, describes this as the single most common problem he sees: every possible industry and every possible title crammed into one campaign, producing an audience of several million people and a campaign nobody can learn anything from.
The problem is not only the size. It is that the same job title means genuinely different things in different industries. An operations manager at a hospital and an operations manager at a freight company do not have the same problems, do not use the same vocabulary and will not respond to the same ad. Blended into one audience, they receive one message, which is written for neither of them.
They also cost different amounts. LinkedIn is an auction, so some industries and seniorities are simply more contested than others. Blended together, an expensive segment quietly consumes budget that a cheaper and better-converting segment never gets to use, and your reporting shows you one averaged number that conceals both.
So when you truly sell into several industries, that is an argument for several campaigns, not one wide one. Same offer if you like, but a separate campaign per industry, with the message adjusted to each. You will see which industries actually respond, what each one costs, and where your next dollar should go. That is real information about your market, and it is only available if you kept them apart.
When Industry Is Exactly the Right Tool
It is worth being straight about the other side, because industry targeting is not a trap to be avoided.
When your product is genuinely bound to a sector, it is the correct filter and there is no better substitute. Compliance software for medical practices, equipment for commercial kitchens, a service that only makes sense to firms holding a particular license: in those cases the industry is not a proxy for anything, it is the qualifying fact, and you should use it.
It is also the right tool when you are deliberately working broad. Early awareness work, where the point is reach within a defined sector rather than precision on a named buyer, is well served by it.
And it is worth noticing how the specialists treat it. AJ Wilcox, who has managed well over 200 million dollars in LinkedIn ad spend, walks through company-side facets like size and industry briskly as the straightforward part of a build, then slows right down on how to find the individual, because that is where he sees people get tripped up. He is right that the company side is the easier half. The catch is that "easier" gets heard as "safe," and a filter nobody examines is exactly the kind of thing that runs unquestioned for a year.
Then Go Look at What You Actually Bought
All of this is a build-time decision, and build-time decisions are worth checking against reality afterwards.
Your account will tell you which companies actually received your impressions, by name. It is the fastest way to find out whether the industry box bought you the market you thought it did, and we walked through how to read it in LinkedIn Tells You Exactly Who Clicked. Almost Nobody Looks. Look once and you will know within ten minutes whether this article applies to your account.
If it does, the fix is not dramatic. Move the weight of your targeting onto a company list, or onto job function, seniority and company size. Let industry narrow what remains. Split genuinely different industries into their own campaigns instead of blending them. None of that requires a bigger budget, only a different arrangement of the one you have.
And if you would rather have someone build that properly than work through it yourself, that is the kind of thing our LinkedIn ads team does all day. Either way, the box is not what it looks like, and you now know what is behind it.
