Key takeaways

A B2B Google Ads campaign optimized for simple form submissions learns to find form fills, not customers. To make the algorithm go after prospects who actually sign, you have to feed it what happens after the click: qualified lead, opportunity, signed contract. That is the job of offline conversion imports and enhanced conversions for leads, fed from your CRM. The issue is getting urgent: according to the WordStream by LocaliQ 2025 benchmark report, the average cost per lead on Google Ads reached $70.11, up from $53.52 two years earlier. Paying more for every lead without knowing which ones turn into revenue is no longer sustainable.

The setup comes down to four steps: capture the click ID and email on the form, define sales-cycle stages in the CRM, send those stages back to Google Ads with a value, then switch bidding to these new conversions once volume is sufficient.

Why the form fill is the wrong goal in B2B

Google's automated bidding strategies (Maximize Conversions, Target CPA, Target ROAS) optimize for exactly what you tell them is a conversion. If the primary conversion is "form submitted," the algorithm will hunt for the queries, audiences and placements that produce the most forms at the lowest cost. In B2C, the form and the sale are often close together. In B2B, they are separated by weeks of sales cycle, a qualification call and sometimes a buying committee.

Anyone who has managed a B2B account knows the outcome: cost per lead drops in the interface, while the sales team complains about lead quality. Students, job seekers, competitors doing research, off-target micro-businesses: they all fill out forms, and the algorithm has learned to find them efficiently.

Benchmarks confirm the gap between B2B and the average. The WordStream by LocaliQ report, covering more than 16,000 campaigns, measured an average conversion rate of 8.18% across all industries in 2025, while long-cycle B2B categories sit closer to 3 to 5% (Search Engine Land, based on WordStream by LocaliQ, 2025). Fewer conversions, at a higher cost, and only a fraction of them become revenue: that is precisely the context in which the signal you send the algorithm matters most.

Average cost per lead on Google Ads, 2023 to 2025 Bar chart: the average cost per lead rises from $53.52 in 2023 to $66.69 in 2024, then $70.11 in 2025, a 31% increase in two years. $53.52 $66.69 $70.11 2023 2024 2025 Average Google Ads cost per lead (all industries)
Source: WordStream by LocaliQ Google Ads benchmarks, 2024 and 2025. +31% in two years.

What Google means by "offline conversions"

An offline conversion is an event that doesn't happen on your website but is tied back, after the fact, to the original ad click: a lead qualified by a sales rep, a meeting that actually took place, an opportunity created, a signed quote. Google Ads offers two mechanisms.

  • Classic offline conversion import: you store the Google click ID (GCLID) at form submission, keep it in the CRM, then send Google the combination "GCLID + conversion name + date + value." Google attributes the conversion to the click if it is less than 90 days old.
  • Enhanced conversions for leads (ECL): a modernized version that relies on user-provided data, typically the email address, hashed with SHA-256. Google positions it as more durable and more accurate than the classic import, with cross-device conversions taken into account as well (Google Ads Help, enhanced conversions for leads).

A recent change to know about if you run an in-house integration: Google has announced that starting June 15, 2026, offline conversion imports and enhanced conversions for leads go through the Data Manager API and are no longer accepted by the Google Ads API (Google Ads Help, offline import upgrade guide). Native connectors (HubSpot, Salesforce) and Zapier handle the switch, but a script built in-house two years ago may have stopped sending data without any visible alert.

The setup in 4 steps

1. Capture the click and the identity on the form

Turn on auto-tagging in Google Ads so the GCLID is appended to URLs. On the site side, a script picks up this parameter (and its GBRAID/WBRAID variants for iOS), stores it in a first-party cookie, then injects it into a hidden form field. The same form passes along the business email, which will be used for enhanced conversions. Both pieces of data must make it all the way to the contact record in the CRM: this is where most projects fail, because an intermediate tool (Typeform, Calendly, a chat tool) doesn't pass the hidden field through.

If your browser-side tracking is already losing part of your sessions, deal with that first: we detailed the method in our article on B2B server-side tracking.

2. Define 3 sales-cycle stages, not 10

Google needs events that happen often enough to learn from. Sending ten pipeline stages dilutes the signal. In practice, three stages are enough:

  • Qualified lead (MQL or SQL): the prospect fits your target, validated by an objective criterion (company size, job function) or by a sales rep.
  • Opportunity: a need, a budget and a timeline have been identified during a conversation.
  • Closed-won customer: with the actual contract value, or the margin if you know it.

The qualification rule must be written down and stable. If every rep decides in their own way what a qualified lead is, the algorithm learns from noise. This is as much a CRM project as an Ads project, and we cover it in our 3-week CRM setup method for B2B SMBs.

3. Send the stages back to Google Ads with a value

In Google Ads, create one conversion action per stage, of the "import" type, then connect the data flow: native HubSpot or Salesforce connector, Zapier, or an n8n/Make workflow that sends events to the Data Manager API every time a stage changes. Assign a value to each stage, even a rough one. A simple method is to multiply the average contract value by the stage's conversion rate to a signed deal.

Illustrative example: with an average contract of €12,000, if 25% of opportunities close and 30% of qualified leads become opportunities, you assign €12,000 to the signed deal, €3,000 to the opportunity and €900 to the qualified lead. These relative values tell the algorithm that a qualified lead is worth far more than an anonymous form fill, and that a signed deal is worth more still.

4. Switch bidding gradually

Don't change your bidding goal the day the import starts working. First keep the offline conversions as secondary conversions for three to four weeks: you check that the volumes coming through match those in the CRM, without disrupting the campaigns. Next, make the qualified lead your primary conversion, keeping the form fill as secondary. Moving to value-based bidding (Target ROAS) only makes sense if you send a steady monthly volume of qualified conversions per campaign; below that, consolidate campaigns or stick with the most frequent stage.

Mistakes that corrupt the signal

  • Uploading too late. The conversion must be imported within 90 days of the click. With a six-month sales cycle, the signed deal will land outside that window: one more reason to rely on the intermediate stages.
  • Double counting. If both the form fill and the qualified lead are primary conversions, the algorithm adds them up. Only one primary stage at a time.
  • Ignoring disqualified leads. Sending nothing for a bad lead is already a signal, but a weak one. Make sure your CRM rules clearly flag off-target leads so you can measure each campaign's share of them.
  • Breaking the chain at the first form change. A page redesign or a new booking tool is enough to lose the hidden GCLID field. A weekly check (share of Google Ads leads with a populated GCLID) saves you from discovering the problem three months later.
  • Judging too fast. After the switch, the reported cost per form fill almost always goes up. That's expected: the metric that matters becomes cost per qualified lead and pipeline generated.

What to measure next

Once the loop is in place, reporting changes in nature. Instead of comparing campaigns on cost per form fill, you compare them on cost per qualified lead, cost per opportunity and pipeline generated per euro spent. These metrics belong in the leadership dashboard, alongside the other channels: that is the purpose of our 5 Looker Studio views for the executive committee. For cost benchmarks by channel, our B2B Google Ads and Meta Ads benchmark remains a useful baseline.

Our take, without overselling: importing offline conversions won't make a poorly positioned offer profitable, and it won't make up for volume that's too low. What it does fix is a structural flaw in B2B accounts: an algorithm you pay to optimize for the wrong goal.

Frequently asked questions

What is an offline conversion in Google Ads?

It's an event that happens off the website, such as a qualified lead, an opportunity or a signed contract, which you import into Google Ads and which Google ties back to the original ad click using the GCLID or the prospect's hashed email. It lets automated bidding optimize for business outcomes rather than form fills.

What's the difference between offline conversion import and enhanced conversions for leads?

Classic offline import relies on the GCLID stored in the CRM. Enhanced conversions for leads rely on user-provided data, such as the hashed email, which makes them more durable and allows cross-device conversions to be measured. Google now recommends enhanced conversions for leads.

How long do you have to import an offline conversion?

The original click must be less than 90 days old at the time of import. For long sales cycles, it's therefore better to send intermediate stages such as the qualified lead or the opportunity, which happen within that window.

Do you need a well-structured CRM to import offline conversions?

Yes. The CRM must store each lead's GCLID and email, and track qualification stages defined by stable rules. HubSpot and Salesforce have native connectors to Google Ads; other CRMs go through Zapier or an n8n or Make workflow.

Why does my cost per lead go up after switching to offline conversions?

Because the algorithm stops looking for the cheapest form fills and starts looking for prospects who qualify. Cost per form fill often rises, but cost per qualified lead and pipeline generated per euro spent are the metrics to track.