Why Google Ads barely delivers qualified B2B leads

The question comes up in B2B more than any other, and it almost always sounds the same: “Our campaigns get plenty of clicks, but barely any qualified inquiries. Why?”

The answer can’t be read off the ad alone. The ad is the only part you can see from outside, so that’s where people start tweaking. But the cause usually sits one layer deeper, in the question of which search terms you’re actually buying and what your account is told counts as a success. As long as Google doesn’t know what a good lead looks like, it reliably delivers what it can measure: form submissions. Including the bad ones.

This article is the diagnostic path I work through, in this order. No numbers borrowed from other people’s accounts, no benchmarks that wouldn’t say anything about your market anyway. What follows is mechanics.

Define the quality question first

Before you exclude any keywords or shift any budget, sit down with sales and define what “qualified” actually means: the right company and need, a reachable contact, a realistic budget, a plausible timeline. Log the rejection reasons separately: private individual, job application, no real need, spam. “Didn’t close” isn’t a quality verdict yet in a long sales cycle.

Then compare inquiries from the same intake window with enough time to mature. Otherwise a new campaign automatically looks worse, simply because its leads haven’t had time to close yet. The number of qualified leads, their rate, and the cost per qualified lead each answer a different question; none of them should carry the budget decision alone.

Need an independent read on the order for account, measurement and inquiry flow? My Google Ads audit (from €1,900) checks exactly these handoffs and delivers an action plan. What that kind of measurement plan can look like is documented in the Neumann & Müller case study. That case documents the planning, not a proven revenue increase.

Part 1: why B2B clicks turn into junk

These four sources are worth investigating as candidates. Which one applies to you only shows up once you compare against real inquiries.

Consumer intent hiding in the same words

The classic B2B trap: your industry term is also a consumer word. Bid on “industrial cleaning,” “access control system” or “data protection officer” in German, and alongside the procurement department you also get private individuals wanting something entirely different, and they don’t convert worse, they convert better, because they fill out a form more casually.

That gets amplified by two settings that together are the single most common cause: broad match keywords plus a thin negative list. Broad match isn’t bad by itself, but it’s an amplifier. It stretches your keyword into neighboring meanings, and without a negative list, those are exactly the spaces your audience isn’t in. Running broad match without maintaining search terms isn’t a campaign, it’s a subscription.

Job seekers, students and suppliers

Three groups click on B2B ads systematically without ever buying anything:

• Job seekers. Someone searching your industry plus a location is sometimes looking for a job, not a vendor. On an ad that says “get in touch now,” that turns into an unsolicited job application in your inquiry form.
• Students and trainees. They’re researching for coursework or theses, and they want exactly what good B2B pages offer: explanations, data sheets, whitepapers.
• Suppliers and competitors. One group wants to sell you something, the other is checking what you offer and what it costs.

None of these groups is a tracking error. They’re real people with real search intent, just not yours.

Research instead of buying intent

B2B buying processes are long and involve several people. A large share of the searches are therefore research: comparing, understanding, building an internal case. Those clicks aren’t worthless, but they aren’t an inquiry either. Buy them with the same expectation as a tender-stage search, and you end up thinking your account is broken when you’re really just judging it against the wrong target metric.

The special case here is the do-it-yourselfer: anyone searching alongside words like “myself,” “guide,” “template,” “free” or “sample” doesn’t want to hire you, they want to handle it themselves. Those searches are great for content and expensive for a sales campaign.

The form as the wrong door

The fourth source isn’t in Google at all, it’s on your own site. A form asking only for name and email invites everyone in, including the person who just wanted a PDF. Add the bot and spam share that lands in every open form. To Google, both count as a conversion, and that’s exactly what the bidding automation keeps learning from.

Part 2: the diagnostic path, in this order

The order matters. Reverse it and you end up optimizing a measurement that’s still measuring the wrong thing.

Step 1: read search terms, not keywords

Not the keyword report, the search terms report, what people actually typed. Take the last 30 to 90 days and go through the terms pulling the most click share. You’re looking for three patterns: job words (salary, position, apprenticeship, internship), learning words (what is, definition, guide, myself), and consumer words drawn from your own vocabulary.

That’s the one analysis that teaches you something without guesswork. If you’d rather have the report sorted automatically instead of reading a thousand rows by hand, that’s exactly what the search terms analyzer is for: upload a CSV, get wasted spend and negative keyword suggestions back, no account access needed.

What you find here becomes your negative keyword list. And it stays that way permanently: a negative list isn’t a project, it’s a routine.

Step 2: check the hurdle at the form

Now the other direction. A form has a hurdle, and it’s a dial with two extremes:

• Too little hurdle, just name and email, no context, no reason. Result: lots of submissions, few real projects. Sales calls into thin air and loses trust in the channel.
• Too much hurdle, 14 required fields, a budget question in step one, a mandatory phone number. Result: almost nothing comes through. And because almost nothing comes through, the bidding automation gets too few signals to learn from at all.

There’s no universal right value, only the right fit: the hurdle should be high enough to capture the reason, what it’s about, for which company, by when. Anyone unwilling to answer that wasn’t going to buy anyway. Anyone who does answer it hands your sales team the conversation material at the same time.

A side effect that’s easy to miss: a field like “what’s this about?” also happens to be your best spam filter, because automated entries become obviously meaningless there.

Step 3: check the ad’s promise against the page’s promise

Open your ad and your landing page side by side and read both out loud. Very often the ad promises something different from what the page delivers: the ad says “quote within 24 hours,” the page shows a generic contact form. Or the ad names a specific industry, the page speaks to everyone.

That gap creates exactly the pattern everyone’s puzzled by: good click-through rate, plenty of submissions, poor quality. The ad attracts people expecting something specific, and the page collects them without any specificity. Fix this, and you often lose clicks while gaining real inquiries, which looks like a decline in the click statistics.

Step 4: tell the system what a good lead actually is

This is the structural part, and without it the steps above stay cosmetic. As long as every form submission triggers the same conversion with the same value, Smart Bidding optimizes for volume. It can’t do anything else, that’s the only signal it has.

The way out is lead stages. Instead of “form submitted,” you report back what became of the lead, qualified, quote sent, won, and at what value. Technically that runs through offline conversion import: the click ID travels into your CRM, and once sales sets the status, that information goes back to Google Ads. From that point on, the automation stops learning who fills out a form and starts learning who actually signs.

Two things regularly go wrong here: the click ID never makes it into the CRM because the form doesn’t capture it. And nobody has defined what “qualified” even means internally, so every salesperson reports something different back. Both are a matter of craftsmanship, not chance.

If nobody on your team can or wants to own this part, that’s the point where handing it off pays off: you can have conversion tracking set up with lead stages included (in German), instead of letting the bidding automation keep learning from a signal you already know is the wrong one.

Part 3: what doesn’t help

Three reflexes that show up reliably in this situation, and that make things worse.

More budget on a broken structure. Budget is a multiplier, not a fix. If your account is mostly buying the wrong search terms, more money just buys more of them, faster. The effect can even feel right at first, because the raw number of inquiries goes up.

Optimizing for clicks. Click-through rate, cost per click and Quality Score are diagnostic numbers, not goals. Switching a campaign to “maximize clicks” because too little is happening reliably delivers the opposite of what you need: more traffic with broader intent. In B2B, the most expensive click is often the best one, and the cheapest one is often the one that just makes the stats look good.

“B2B targeting” in search. In display and social networks you can target by company attributes. In Google Search, that doesn’t exist the same way: you’re buying a search term, not an audience. Audience signals and customer lists help as an observation layer and a bid adjustment, but they don’t turn a consumer search into a business inquiry. Anyone waiting for a setting to solve the intent problem is waiting for something this channel doesn’t have.

To be fair, sometimes the honest answer is that search is simply the wrong channel for your offer, because nobody’s searching for it. That’s a result, not a failure, and it’s cheaper the earlier you find out.

The checklist

Work through this in order:

1. Read the last 30 to 90 days of search terms and sort them into job, learning and consumer words.
2. Build a negative keyword list from that, and set a recurring slot for it, not a one-off.
3. Check whether broad match is running without a maintained negative list. If so: fix the list first, talk strategy after.
4. Hold the form against the actual reason for contact: does it ask what it’s about, or just for an email?
5. Read the ad and the landing page side by side. Does the ad promise what the page delivers?
6. Check whether the click ID actually reaches the CRM. Without it, no feedback loop is possible.
7. Define what “qualified” and “won” mean internally, in writing, one line per status.
8. Feed lead stages back as conversions and switch the bidding strategy to match.
9. Only talk about budget after that.

Points 1 through 5 cost you an afternoon and no money. Points 6 through 8 are the part that actually changes something, and the part where most accounts get stuck, because it reaches beyond Google Ads into the website, the form and the CRM.

Getting plenty of inquiries and not many usable ones? In a free first call, I go through the search terms and the conversion definitions with you and tell you whether the problem sits in how you’re buying clicks or in what your account is told counts as success. Write to me to set one up. If you’d rather hand off the ongoing management, Google Ads management (in German) covers what that includes.