Key takeaways
- Volume is easy, quality is the work. Cheap leads show up fast, the right leads take a system.
- Diagnose first, don't guess: lead profiles, the Demographics tab, and the segment breakdown show you the pattern.
- Most common cause: the audience is too broad. Layering exclusions sharpens targeting more than building new audiences.
- A qualifying question in the form costs you volume and lifts quality. Usually a good trade.
- The feedback loop decides: sales rates leads, targeting learns, and excludes the wrong-fit people.
Marketing and sales argue about lead quality, and both are right: they're looking at different ends of the same campaign. Generating volume is easy: lower the barrier, widen the audience, make the offer low-commitment, and the leads pour in. Quality is the actual work, because it needs a system that catches bad leads, sharpens targeting, and involves sales. Anyone optimizing purely for cost per lead is almost inevitably optimizing for volume and against fit. It's exactly this cheap-volume trap that's easy to fall into when the channel is outsourced and runs far away from sales, one of the reasons many teams would rather build LinkedIn Ads inhouse than outsource it.
Diagnose first, don't guess
Before you start turning dials, find out where the bad leads are actually coming from. Campaign Manager gives you three tools for this, no extra software needed:
- Read the lead profiles. Look at your last batch of poor-fit leads and search for shared traits: same industry, same company size, same job title. That pattern is your clue.
- The Demographics tab. It shows who sees, clicks, and converts on your ads, broken down by job function, seniority, industry, and company size. Once you've got a meaningful amount of data, roughly a few hundred actions, the picture becomes reliable.
- The segment breakdown before launch. Even while you're building the audience, you can see its composition. If half of it consists of a role that never buys, you'll know before launch, not after a hundred leads.
The most common cause: the audience is too broad
In most cases, the problem isn't the channel, it's an audience that's too broad. The reflex to keep building new audiences rarely helps here. Layering exclusions is more effective: deliberately removing the roles, industries, and company sizes that have turned out to produce bad leads from your targeting. One detail worth knowing: job titles on LinkedIn are a free-text field that the platform groups into categories, and according to market data it only cleanly recognizes about half of all titles. That means a too-broad exclusion can accidentally remove titles you actually wanted, too. Exclude, yes, but with precision and a review, not with a broad brush. For how to build precise audiences from the start, see the Targeting guide.
The qualifying question in the form
A Lead Gen Form doesn't have to produce junk by default. The single strongest lever for quality is a real qualifying question in the form that separates the right inquiries from the wrong ones, for example asking about the person's role in the buying process, their specific need, or their timeframe. That costs you completion rate, so you'll get fewer leads. But that's the whole point of the exercise. A slightly more expensive lead that's actually right for sales is almost always cheaper than three cheap ones that cost sales time and patience. One more thing: don't ask for a phone number on low-commitment offers, it scares off exactly the early-stage marketing leads who aren't ready to talk to sales yet.
The strongest filter kicks in before the first bad lead even happens. Run an inexpensive video first, then target the lead form only at people who watched more than 50 percent of it. Anyone who sticks around that long is genuinely interested in the topic. One or two qualifying questions in the form sift out the rest. Fifty qualified leads beat 500 that nobody can do anything with.
The feedback loop: sales rates, targeting learns
The real quality engine is a loop between sales and the campaign. Without it, you're optimizing blind. With it, the campaign gets better every round:
- Sales rates every lead against a simple scorecard, for example: ready to talk, needs more info, follow up later, not a fit. No complicated scoring model, just a scorecard that takes seconds.
- You look for patterns among the poorly rated leads and take those traits into your targeting as exclusions. If a certain job function is consistently a poor fit, for instance, it gets cut.
- You track segments and campaigns by outcome, not lead volume. Segments that deliver plenty of leads and no customers get cut. In the end, what counts is who actually buys, and according to market data that's only around a tenth of leads anyway.
For this loop to work at all, marketing and sales first have to agree on what a good lead actually looks like. That definition belongs before the campaign, not after it. For how a qualified lead turns into a booked meeting, see B2B Meetings via LinkedIn Ads.
Cost per lead is the most seductive wrong number in B2B. A segment with a €40 lead price and a 2 percent close rate is more expensive than one at €90 with an 8 percent close rate. So always break lead cost down by industry, company size, region, and seniority, and weigh it against cost per customer won. Only then do you see where your budget is actually paying off most.
Quality is also a funnel question
Even the best targeting doesn't change the fact that a cold audience delivers weaker leads on average than a warm one. Someone who already knows you fills out a form with real intent, not on a whim. That's why leads from retargeting are almost always the better ones. And quality starts even earlier, with the asset itself: a weak lead magnet that people could just as easily google pulls in lukewarm contacts, real expert knowledge pulls in strong ones. Lead quality is never just a filtering question, it's tied to the funnel stage and the offer. For why the impact of good leads often only shows up with a delay, see the Shadow Funnel.
