Google Ads for B2B lead generation: when it pays and when it burns cash

Paid search is the only channel where a stranger tells you what they want before you spend anything. That is its whole advantage. It is also why so many B2B accounts leak budget: the people typing your keywords are a small slice of the market you actually need to reach, and the rest of the spend goes to everyone else who happened to type something similar.

What Google Ads can and cannot do for B2B

Google Ads is a demand capture channel. Someone has a problem, decides a solution category exists, types a query, and you pay to appear in front of that moment. Everything good about paid search follows from that sequence, and so does everything bad. No query, no ad. No category awareness, no query.

This matters more in B2B than in consumer markets because B2B categories are often narrow, technical or new. A firm selling procurement software to hospital groups is competing for a keyword a handful of people type each month across a whole country. A firm selling industrial cleaning in Lithuania is competing for a term dozens of buyers type every week. Same channel, completely different economics.

Does search intent exist in your category?

Answer this before you open an account. Run three checks.

• Volume. Pull monthly search volume for your five most commercial keywords in each country you sell to. If the combined figure is in the low hundreds, paid search cannot carry a pipeline on its own.
• Language. Volume in English tells you nothing about volume in German, Polish or Lithuanian. European B2B search splits by language, and the local term is often the only one buyers use.
• Vocabulary match. If buyers describe the problem but have no name for the solution, you are selling a new category. Ads will find people searching for the old workaround, not for you.

When those checks come back thin, the honest conclusion is that the demand is not there to capture yet. That is a targeting problem, not a bidding problem, and no landing page fixes it.

What do B2B Google Ads actually cost?

Cost per lead is cost per click divided by landing page conversion rate, and both numbers swing hard by category. Competitive software, legal and financial services keywords are bid up by companies with large deal sizes, and clicks on those terms regularly reach double digit euros. Niche industrial, technical and local service terms often stay in the low single digits because almost nobody else is bidding. Cold paid traffic to a B2B landing page typically converts to a form fill at a low single digit percentage, higher on branded and comparison terms, lower on anything research shaped.

Chain those together and the arithmetic gets clear fast. Suppose you spend EUR 6,000 and clicks average EUR 6. That buys 1,000 clicks. At a 2.5% form fill rate you get 25 enquiries, so roughly EUR 240 per raw lead. If 40% of those are actually in your target profile, you have 10 real opportunities and your qualified lead cost is closer to EUR 600. Nothing in that chain is unusual, and nothing in it is visible from the click report alone.

Run the same maths with your own numbers before you commit a budget. It is the same exercise we walk through in our B2B lead generation pricing guide, and it usually settles the question of whether ads or another channel gets the money.

High intent keywords versus broad terms

Split your keyword list into three tiers and treat them as separate businesses.

• Brand and competitor terms. Cheapest clicks, highest conversion, smallest volume. Defend your own brand name and bid on the competitors you genuinely displace.
• Bottom of funnel commercial terms. Queries with a buying verb or a vendor shape: provider, supplier, agency, software for, quote, pricing, near me. These are where B2B paid search earns its budget.
• Broad and research terms. What is, how to, best practices, template, salary. These pull students, job seekers, consultants and competitors. They belong in your content and SEO plan, not in a lead generation campaign.

Google's automation pushes against this split. Broad match, Performance Max and volume oriented smart bidding all widen the net, which is exactly what you want in consumer retail and exactly what you do not want when a wrong click costs EUR 8 and a right one is rare. Keep match types tight, build a negative keyword list from day one, and review the search terms report weekly rather than monthly.

Why your leads look worse than your click report

The most expensive mistake in B2B paid search is optimising for form fills. Smart bidding learns from whatever conversion you feed it. Tell it a submitted form is the goal, and it will reliably find people who submit forms. Those people include everyone with no budget, no authority and no intention of buying.

The fix is a feedback loop. Import offline conversions from your CRM so Google learns from qualified lead, meeting held and closed deal rather than from the form. Long B2B sales cycles make this awkward, since the signal arrives months after the click, so most teams optimise on the earliest reliable proxy: a lead your sales team accepted. That single change usually does more for lead quality than any amount of ad copy testing.

Two smaller fixes matter too. Match the landing page to the query rather than sending every campaign to the homepage. And ask for the minimum information that lets you qualify, because every extra field cuts conversion while a missing company field makes the lead unusable.

Google Ads and outbound solve different problems

Paid search reaches buyers who are looking right now and selects itself. You do not choose who sees the ad, and you stop existing the moment the budget stops. Outbound inverts both properties: you choose the exact accounts, including the large majority who are not searching this quarter, and the list and the relationships stay yours. Cold email reply rates across B2B typically run between 1% and 5%, which sounds modest until you notice you picked every single recipient.

That difference decides the mix. In a large market with real search volume, run both: ads to capture the in market minority, outbound to create conversations with the rest. In a market of a few hundred qualifying companies, ads cannot find enough queries to matter and outbound covers the entire universe deliberately. Where budgets are tight, most European teams get further starting with outbound and adding paid search once they know which messages convert. Our flat monthly pricing exists partly so that comparison is easy to make.

When Google Ads burns cash

• No search volume. A new category, or a country where nobody uses your terminology.
• A tiny addressable market. A few hundred target companies produce too few queries to build a campaign on.
• Outbid by funded competitors. If rivals can pay more per click because they close bigger deals, you buy their leftovers.
• Homepage as landing page. Query and page mismatch caps conversion no matter how good the traffic.
• No CRM feedback. Without offline conversion import, the algorithm optimises toward the wrong outcome from week one.
• Performance Max as the only campaign. Convenient, opaque, and prone to spending B2B budget on the cheapest available clicks.
• Judging on cost per form fill. The number that looks best is usually the one that hides the quality problem.

A six week test that settles it

Do not commit a year of budget to find out. Run a contained experiment instead. Pick the ten most commercial keywords in one country and one language. Build one search campaign on exact and phrase match only, with a negative list. Send each ad group to a page that answers that specific query. Set a budget you can lose without flinching. Track cost per click, cost per form fill, and cost per sales accepted lead separately, and tag every lead in the CRM.

After six weeks you will know whether the volume exists, what a qualified lead really costs, and whether that number beats your other channels, including LinkedIn ads and outbound. If the answer is yes, scale a proven structure. If it is no, you spent a small budget to avoid a large one, which is the point of the test.