15 May 2026
Most B2B companies are optimising their paid campaigns towards the wrong outcome. They track form submissions, trial signups, or demo requests, tell Google and Meta to find more of those, and then measure success by cost per lead. The problem is that in most B2B businesses, a large fraction of those leads never become paying customers. The algorithm gets good at finding the cheap leads. Those cheap leads churn. Meanwhile the expensive leads, the ones from industries that actually use the product, from companies with the right profile, are underrepresented because the algorithm was not told to find them.
Offline conversion tracking fixes this. It is one of the highest-impact changes a B2B company can make to their paid acquisition, and most companies never do it.
When someone clicks a Google Ad, Google assigns a unique identifier to that click, the GCLID, or Google Click Identifier. This identifier is passed to your website and can be captured in your form submission data or CRM. When that lead later becomes a paying customer, you have the GCLID that originated the engagement. You send that GCLID back to Google along with the conversion value, and Google's algorithm learns what a high-value customer looks like from an acquisition perspective, not what a lead form submission looks like.
Meta uses a similar mechanism, the fbclid, with its Conversions API enabling the same offline event matching. The principle is the same: close the loop between your CRM outcomes and your ad platform's optimisation signal.
It requires engineering work. You need to capture the click identifier in your CRM at the moment of form submission, which most CRMs do not do out of the box. You need a process for pushing closed-won events back to the ad platform, either through a direct API integration or via a CSV upload. And someone has to own the process and keep it running accurately.
Most early-stage companies prioritise building product over plumbing attribution. That is understandable. But the cost of not doing it compounds over time: every month you optimise towards form submissions rather than revenue, the algorithm gets better at finding the wrong leads.
Step one is GCLID capture. Your landing page forms need to pass the Google Click Identifier from the URL parameter to a hidden field in the form, and that hidden field value needs to be stored in your CRM against the contact record. This typically requires a small JavaScript snippet on the page and a custom field in your CRM.
Step two is defining the right offline conversion event. For most B2B SaaS companies this is a closed-won deal or a paid subscription conversion. You could also use a qualified opportunity stage if your sales cycle is long and you want to give the algorithm an earlier signal with more volume.
Step three is the upload. Google accepts offline conversion data via the Ads Manager interface (manual CSV upload), via their API, or via integrations through CRMs like Salesforce and HubSpot. The upload includes the GCLID, the conversion name, the conversion time, and optionally the conversion value.
The algorithm reorients over several weeks. It stops finding as many cheap, low-quality leads and starts finding fewer but better-qualified ones. Your cost per lead typically goes up. Your cost per qualified opportunity or cost per customer typically goes down. This can feel uncomfortable initially if the business has been measuring success by CPL, but the unit economics improve materially.
Companies that implement offline conversion tracking properly typically see a meaningful improvement in lead quality within one to two billing cycles, as the algorithm recalibrates towards the conversion signal that actually reflects business value.
If you want help setting this up for your account or designing the right conversion architecture for your model, that is exactly the kind of work I do.