Tighter targeting is not a free upgrade. It is a purchase, and most B2B teams never check the price.
Every audience filter you add to a paid campaign does two things at once. It removes people you did not want, and it raises what you pay for the people you kept. The second part almost never makes it into the conversation. This is for B2B marketers running paid social or audience-targeted campaigns who keep narrowing in the name of lead quality and cannot explain why cost per qualified lead went up instead of down.
Narrowing an audience raises your cost per click, and that cost is real
When you add a company-size filter, a seniority filter and a job-function filter on top of each other, you are not refining a list. You are bidding into a smaller, more contested pool of impressions that every other B2B advertiser also wants.
In the accounts we look at, moving a LinkedIn audience from a wide mid-market definition to a very-large-company definition roughly doubles cost per click. Not a few percent. Roughly double.
That is the pattern, not a promise, and your number will differ. The point is that it is a number at all. Teams treat a targeting change as a settings change, the way you would fix a typo. It is closer to accepting a price increase from a supplier.
Ready for paid ads that pay off?
Book your free auditThe one calculation that tells you whether a filter is worth adding
A filter pays for itself only if it multiplies your qualified-lead rate by at least as much as it multiplies your cost per click.
Work it with round illustrative numbers. Say you have $20,000 a month on paid social. At a $14 cost per click, that buys about 1,428 clicks. If 3.5% of those clicks become a lead and 24% of those leads are sales-accepted, you end the month with 12 qualified leads at roughly $1,667 each.
Now add the enterprise headcount filter and let cost per click go to $28. The same budget buys about 714 clicks. To hold that same $1,667, those 714 clicks have to produce 12 qualified leads, which means your qualified rate per click has to go from 0.84% to 1.68%.
So ask the honest question. Does your landing page convert twice as well for someone at a 12,000-person company than for someone at a 3,000-person company?
In my experience the answer is no. It converts somewhat better. Maybe 20% or 30% better. A 30% quality lift against a 100% price increase is a losing trade, and it is the trade most B2B teams are quietly making every time they tighten an audience.
Which means quality is the wrong justification. Deal size is the right one
Here is where the argument turns. If a filter cannot double your lead rate, it has to change the deal, not the lead.
Run the same two scenarios on pipeline value per ad dollar instead of cost per lead.
→ Wide audience. 0.84% qualified per click at a $40,000 average deal is $336 of pipeline created per click. Divide by the $14 click and you get about $24 of pipeline per ad dollar.
→ Narrow audience. 1.09% qualified per click (the modest 30% lift) at a $160,000 average deal is $1,744 of pipeline per click. Divide by the $28 click and you get about $62 of pipeline per ad dollar.
The narrow audience wins by a wide margin, and it wins for a reason that has nothing to do with lead quality. It wins because the deals are four times bigger.
That is the whole decision. If the segment you are narrowing to buys a materially larger deal, pay the premium and stop apologizing for the lead count. If it buys roughly the same deal as the segment you are cutting, you are paying double for a rounding error, and you should widen back out.
Teams get this backwards because cost per lead is the number on the dashboard and deal size lives in the CRM. The fix is not a better dashboard. It is refusing to approve a targeting change until somebody states the average deal size of the segment being kept and the segment being cut.
Price the segment before you pay to narrow to it
You do not have to guess. Google Ads has a setting built for exactly this, and almost nobody in B2B uses it deliberately.
Google's documentation on Targeting and Observation settings is blunt about the difference. "Targeting restricts the reach of your ad group." The alternative does not: "The 'Observation' setting does not restrict the reach of your campaign or ad group." Google tells you to use it "if you don't want to narrow your campaign's or ad group's targeting any further, but you want to monitor how certain criteria perform with your ads", and recommends it for all Search campaigns.
Observation gives you the segment's performance without the price. You add the audience, you leave reach alone, and after a few weeks you can read what that segment actually did inside traffic you were already buying. The same page notes that under Smart Bidding, first-party segments added as observations "will be used as signals for Smart Bidding strategies", so the data is not sitting idle while you decide.
The sequence that follows from this is simple. Observe first, read the segment's real conversion behaviour, then narrow only if the arithmetic above clears. Narrowing on a hunch and reading the damage two months later is the expensive order to do it in.
Some filters do not filter at all, and the platform will tell you if you read
Before you argue about whether to narrow, confirm that the control you are using narrows anything.
Microsoft Advertising sells LinkedIn profile targeting on company, industry and job function, which sounds like the sharpest B2B filter available on search. Its own help page says otherwise: "LinkedIn profile targeting will not narrow your ads' audience. For example, by targeting a specific company, you aren't excluding everyone who does not work for that company." Microsoft tells you to think of it as "bid only", as opposed to "target and bid".
That is a useful feature and a terrible surprise. Plenty of B2B accounts have a Microsoft campaign that the team believes is running only to a named account list and is in fact running to everyone, with a bid nudge for the list.
Meta goes the other direction. Its Advantage+ detailed targeting documentation says the system can "reach a broader group of people than you defined in your detailed targeting selections", and that if it "finds better performance opportunities outside your defined audience", it will "dynamically expand the audience" and "may expand your audience further to include similar opportunities."
So on one platform your filter is a bid hint. On another your filter is a starting suggestion the system is allowed to walk away from. Build a targeting strategy without knowing which behaviour you are dealing with and you will be reporting on an audience you never actually bought.
Inclusions are suggestions. Exclusions are rules
There is a durable asymmetry buried in that same Meta page. When Advantage+ expands past your detailed targeting, "any exclusions and targeting selections outside of detailed targeting (such as age, gender, location and language) continue to apply."
Read that as an operating principle. The inclusion list is the part the platform is allowed to overrule. The exclusion list is the part it honours.
Exclusions are also cheaper. Removing a job function you never wanted does not bid you into a premium pool, it just stops you buying inventory you were already buying at the normal price. An inclusion raises the floor on every click. An exclusion lowers the volume of the clicks you did not want.
When lead quality is the problem, reach for the exclusion first. We went deeper on the quality side of this in our post on fixing bad LinkedIn leads, and on why the underlying company data is shakier than most teams assume in our piece on firmographic targeting.
Three situations where narrowing is correct even when the math is ugly
Not every reason to narrow is an efficiency reason, and pretending otherwise leads to bad calls.
Sales capacity. If two reps can genuinely work 30 opportunities a month, buying 60 cheaper leads is not a win, it is a queue. Narrowing here is a staffing decision wearing a media costume, and it is a legitimate one. Say so out loud when you make it, so nobody grades it on cost per lead afterwards.
Restricted ad categories. Financial products and a handful of other categories lose lookalike and similar-audience options on the major social platforms. When the modelling tools are switched off, first-party lists and engagement audiences carry the whole targeting load, and a tight inclusion is the only instrument left.
A genuine deal-size cliff. If the segment above a size threshold buys three times larger and the segment below it barely closes, the premium is trivially repaid. That is the enterprise case from earlier, and it is real. It is just far rarer than the number of accounts that claim it.
One situation where narrowing is not correct: making a report look better. A tighter audience produces a tidier demographics breakdown and a worse business. If the only thing that improved is the share of impressions matching your persona definition, you bought a slide.
The monthly read that keeps this honest
Once a month, pull spend broken out by the attributes you care about. On paid social that is job function, seniority and company size. On search it is your observation segments.
Then do two things with it, in this order.
1.) Find where the money went outside the profile. If more than about a quarter of spend landed on people you would never sell to, that is an exclusion to write this week.
2.) Compute cost per click inside each segment, not just conversion counts. This is the number that tells you what the next filter will cost before you add it, and it is the one almost nobody pulls.
The second step is what turns this from an opinion into a policy. After two or three months you will have a small table of what each audience attribute costs you per click, and targeting arguments stop being about who sounds more confident.
Where to start
Open your paid social account and find the narrowest audience you are running. Write down its cost per click and the cost per click of your widest audience. Divide one by the other. That ratio is the quality lift the narrow audience owes you, and you now have a specific number to test it against.
If it does not clear, widen the audience, move the qualification job into exclusions, and go find the deal-size evidence before you narrow again.
We do this pricing exercise on every account we take over, and it is one of the faster ways to free up budget that was being spent on precision nobody was measuring. If you want a hand running it on your own account, that is the kind of work our LinkedIn Ads team does, and if you are still deciding whether audience-targeted social belongs in the plan at all, start with how to pick a B2B paid channel.
