Most business owners research their competitors the way everyone does. They open the competitor's website, read the homepage, maybe scroll the company's LinkedIn page, and come away with a general feeling about how that business talks about itself.
That is the brochure. It is the version they built for anyone who wanders by.
The ads are different. An ad is what a company is willing to pay money to say, to a specific person, right now. It is the sharpest statement of what they think their buyer cares about, because every word in it costs them something. And on LinkedIn, every one of those ads is sitting in a public database that anyone can open, for free, without logging into anything.
Most owners have no idea it exists.
There Is a Public Record of Every Ad They Run
It is called the LinkedIn Ad Library, and it is not a marketing tool LinkedIn built to be generous. It exists because regulators made it exist. The European Union's Digital Services Act requires large platforms to keep a searchable public record of the advertising they carry, and LinkedIn built the Ad Library to comply. The side effect is that a small business in Boise or Charleston now has a permanent window into what every competitor is running.
Here is what is in it. Ads that ran on LinkedIn after June 1, 2023, and each ad stays in the library for one year after its last impression. You can search by company name, by keyword, by country, and by date range. For each ad you get the creative itself, the advertiser name, the name of the entity that paid for it, the format, and the dates of its first and last impression.
You do not need an account, a subscription, or a tool that costs $99 a month. LinkedIn's own description is that the library is available to members "and the public." You open linkedin.com/ad-library and type a competitor's name into the search box.
Start with a list of five. Not twenty. Five companies you genuinely lose deals to, the ones whose names come up on sales calls. Search each one and see what comes back. Some will have nothing, which is information all by itself. Others will have forty ads, which tells you something different about how seriously they take this channel.
One caution before you get excited about a company that shows nothing: an empty result does not always mean they are not advertising. It can also mean they run under a different legal entity, or that everything they ran has aged out of the one-year window. Check the parent company and any brand names they operate under before you conclude the channel is uncontested.
The Date Field Is Worth More Than the Creative
Everyone who finds the Ad Library does the same thing first. They look at the pictures. They study the layout, the color, the stock photo, and they start thinking about whether their own ads look as polished.
That is the least valuable thing on the screen.
Look at the dates instead. Every ad carries its first impression date and its last impression date, and those two numbers together are the closest thing you will ever get to a public performance metric for a competitor's campaign.
Think about what a long run actually means. Nobody keeps paying to serve an ad that is not working. Ad budgets get reviewed, campaigns get cut, and the ads that survive are the ones somebody looked at and decided to keep funding. So an ad that has been running continuously for nine months is not a design you should admire. It is a message that has survived nine months of someone asking whether it is worth the money.
Read the whole set of dates together and a shape appears:
The long-runner. One or two ads with a first impression date many months back and a last impression date of this week. These are the ones to study hardest. Whatever that ad says is the claim their buyers respond to.
The burst. A cluster of ads that all start and stop within a few weeks, usually around a conference, a product launch, or a quarter end. This tells you they treat LinkedIn as an event channel, not an always-on one, which is often an opening.
The graveyard. A long list of ads that each ran for two or three weeks and stopped. That is a company testing constantly, or a company with no idea what it is doing, and the creative usually tells you which. Either way it means no single message has earned a permanent slot.
One honest limit, and it matters: the library shows you what survived, not why. An ad can run for eight months because it generates pipeline, or because nobody at that company has looked at the account since March. We have written before about how much of the competitor creative you see is really just what somebody was guessing about six months ago. Longevity is a strong signal. It is not proof.
Read the Offer, Not the Layout
Once you stop looking at the design, there are four things worth pulling out of every ad, and they all fit in a spreadsheet row.
The promise. What specific outcome does the ad claim? Not the category they are in, the outcome they are selling. "Field service software" is a category. "Get your crews to six more jobs a week" is a promise. Most B2B ads never make it past the category, which is exactly why the ones that do get remembered.
The offer. What are they actually asking for? A demo, a webinar registration, a guide, a free assessment, a pricing page visit. The offer tells you where they think their buyer is in the process. A company running nothing but demo requests believes their market is already shopping. A company running guides and webinars believes it has to create the demand first.
The named buyer. Does the ad say out loud who it is for? Ads that name a role or an industry in the copy are doing work that targeting alone cannot do, because plenty of people who see the ad were never in the target audience to begin with. If your competitor's ads never name anyone, that is a gap on a plate.
The proof. What do they lean on when they need you to believe them? A customer logo, a number, a named case study, an analyst report, or nothing at all. If four of your five competitors are leaning on enterprise logos and you serve businesses with twenty employees, you have just learned what your ad should say instead.
This is the part that actually changes your own account. On LinkedIn, the targeting options are the same for everyone, which means the message is where the real difference gets made. We have argued that fixing the targeting does not help if the ad still says nothing, and a competitor sweep is the fastest way to find out what "something" sounds like in your category.
If They Advertise in Europe, You Can See Their Targeting
Here is where the library goes from useful to genuinely uncomfortable for the company being looked at.
Because the disclosure requirements come from EU law, ads that are targeted to the European Union carry far more detail than everything else. For those ads, the library shows the total impressions the ad has received, a breakdown of those impressions by country rounded to the nearest percent, and the targeting parameters the advertiser used.
Not a vague description of the targeting. The categories themselves. LinkedIn discloses twelve of them: language, job, education, company, location, audience, age, gender, demographic, member interests and traits, predictive audience, and audience expansion. For each one it also states whether the advertiser used it for inclusion, for exclusion, or for both.
Read that last part again. You can see which audiences a competitor deliberately excluded.
That is a strategy document. Exclusions are where a company admits who it does not want, which is usually a more honest statement of their ideal customer than anything on their website. You can also see whether they left audience expansion switched on, which tells you whether anyone is minding the account.
Now the limit, stated plainly, because this is where an article like this usually oversells. This detail only appears on ads targeted to the European Union. If your competitor is a US company selling to US buyers and never runs anything into the EU, you will see their creative and their dates and nothing else. No impressions, no targeting, no country split.
For a lot of small businesses that is the end of the story, and it is worth knowing before you go looking for something that will not be there. But check anyway. Plenty of mid-sized B2B companies run at least one EU-targeted campaign, and the moment they do, that campaign's targeting is on public display.
What the Library Will Not Tell You
A tool like this is easy to over-read, so it is worth being precise about the edges.
It does not show spend. There is no budget figure anywhere. An impression count on an EU-targeted ad gets you closer, but you still do not know what they paid for it.
It does not show results. No clicks, no cost per lead, no conversions, no pipeline. The library tells you what a company chose to say and how long they kept saying it. Whether any of it produced a customer is invisible.
It hides restricted ads. For ads that fall under legal restrictions, LinkedIn does not display the preview, the advertiser name, or the payer name at all.
It is not a competitor's whole strategy. LinkedIn is one channel. The same company may be spending five times as much on search, or on outbound, or on a trade show circuit you never see. We made this point at more length in the piece on competitors with ten times your budget: most of a big competitor's spend is not aimed at your customer at all, and reading one channel as the whole picture is how small businesses talk themselves out of competing.
There is also an Ad Library API in LinkedIn's developer portal if you ever want this programmatically. For a business with five competitors and a spreadsheet, you will never need it.
This Is Reconnaissance. It Is Not a Swipe File.
Now the part where we contradict what you are probably about to do with this.
The instinct, once you find forty competitor ads in one place, is to build a swipe file. Save the good ones, note the headline structures, and use them as a starting point for your own. It feels efficient. It is the single fastest way to make your advertising invisible.
We have said this before and we still mean it: your best ideas are not in anyone's ad library. If you build your ads out of the ads already running in your industry, you arrive at the same claims, in the same order, in the same tone as everyone your buyer is already ignoring. You end up as the fifth company saying "purpose-built for modern finance teams" to a person who has stopped reading that sentence.
So use the library for the opposite purpose. Do not read it for what to say. Read it for what has already been said.
The most valuable output of a competitor sweep is a list of the claims everybody is making. When you can see that all five competitors lead with integrations, or all five lead with enterprise security, you have found the consensus of your category. And the consensus is exactly where you should not stand. The open ground is whatever none of them is willing to say: the buyer they all ignore, the objection they all dodge, the price they all hide, the use case they all treat as too small to bother with.
That is also the honest test of whether this exercise worked. If you finish your sweep and your next ad looks more like your competitors' ads, you used it wrong. If you finish it and your next ad says something none of them would put their name to, you used it right.
Thirty Minutes, Five Competitors, One Decision
Here is the whole routine, and it genuinely takes about half an hour.
Open the Ad Library. Search your five competitors one at a time. For each company, build one row in a spreadsheet with five columns: the company, how many ads they have running, the earliest first impression date you can find, the offer they ask for most often, and the single claim their longest-running ad makes.
Then look down the columns instead of across the rows.
The claim column tells you the consensus of your category and therefore where not to stand. The offer column tells you whether your market is being sold to as ready-to-buy or still-being-convinced, and whether you are matching that or fighting it. The date column tells you who is serious about this channel and who shows up twice a year. The count column tells you how much noise your buyer is already hearing before your ad arrives.
Then make one decision, not ten. The most common one it should drive: change what your own ad claims. The second most common: change who you name in it. If your sweep shows five competitors all talking to enterprise buyers and your business is built for the twenty-person company, the ad practically writes itself, and the targeting work you do afterward gets easier, because pointing ads at a specific list of companies only pays off when the message is aimed at the same people.
Do it once a quarter. Anything more often and you will be watching competitors instead of serving customers, which is its own kind of waste.
We do this sweep for every LinkedIn account we take on, usually before we write a single ad, because it is the cheapest way to avoid spending a client's money saying what four other companies already said. If you want a hand reading what you find, or you would rather hand the whole channel to someone who does this every day, that is what we do. Either way, go type a competitor's name into that search box. It costs you nothing and you will learn something in the first five minutes.
