The 11% Your Law Firm Never Sees: First-Party Tracking for Google Ads

By Glenn Gilmour · Published 7 July 2026 · Last updated 29 October 2026

A law firm calls us in a familiar state: the Google Ads account reports 34 conversions last month, the intake team remembers something closer to 50 real enquiries, and nobody can reconcile the two. The marketing company says the discrepancy is “attribution.” The firm’s owner hears that as “we don’t know.”

There is a specific, measurable, fixable reason for most of that gap, and it has nothing to do with attribution models. It is the request that carries the measurement in the first place.

Where the signal actually dies

When a prospective client lands on your firm’s website from an ad, the page asks the browser to fetch Google’s measurement tag. In a standard setup, that request goes to a Google-owned domain. The conversion event, when the person submits a form or taps your phone number, goes to a Google-owned domain too.

Both are third-party requests, and third-party requests are now the most heavily policed traffic on the internet:

• Content and ad blockers refuse them outright, and a meaningful share of every audience runs one.
• Safari’s tracking prevention limits how long script-written cookies survive, which breaks the link between a visit and a conversion that happens days later.
• Corporate and school networks filter tracking domains at DNS level.
• Consent tools block the tag before it ever loads.

None of this stops the prospective client from calling your office. It stops Google from knowing they called. That distinction matters enormously, because Google’s automated bidding decides where to spend your budget based on what it can see.

What first-party serving changes

Google tag gateway for advertisers lets you deploy the Google tag using your own domain, so that measurement events are sent to your domain and then forwarded on to Google (Google Ads Help).

If your site sits behind Cloudflare, the implementation is genuinely straightforward. Cloudflare intercepts the request at its edge, rewrites the path to the original Google endpoint, fetches the script in the background, and serves it from your domain. Cloudflare’s own engineering write-up describes it as converting third-party calls into first-party requests, with no changes required to tags already on the page, and it is free on every Cloudflare plan including the free tier (Cloudflare).

Setup runs from either end: through Google Tag Manager’s admin settings, or through the Engagement tab of the Cloudflare dashboard.

The number, and the honest version of the number

Google’s published figure is specific: “Advertisers who configured Google tag gateway for advertisers saw an 11% uplift in signals” (Google Ads Help). Cloudflare reports the same 11% average from early testers. Agency case studies report a wider band, roughly 9% to 18%, depending on how much signal the site was losing to begin with.

Now the part most people selling this leave out.

An 11% uplift in measured conversions is not 11% more clients. Those enquiries were already happening. Your intake team already answered those calls. What changed is that the measurement system can now see them.

So why does it matter commercially? Because Smart Bidding spends your budget based on what it can see. When 11% of your conversions are invisible, they are invisible unevenly. The campaigns, keywords, devices and counties where signal loss is worst look worse than they are, and budget drifts away from them. Restoring the signal does not create new clients on day one. It corrects where the money goes over the following weeks, and that is where the real gain shows up: in cost per signed matter, four to eight weeks later.

Any partner who shows you the conversion graph jumping the week they switched this on, and calls it growth, is telling you something untrue with accurate numbers.

The three ways this goes wrong

The domain is not actually proxied. A Cloudflare account that only handles DNS, with the proxy switched off, gives you none of this. The gateway will sit in a pending state and nobody will notice for a month.

The firewall eats the measurement path. Aggressive bot protection can challenge the very endpoint your conversions travel through. The symptom is not an error message. The symptom is conversions quietly stopping. Every deployment needs a live conversion test at launch and again 24 hours later, and every subsequent firewall change needs the same test.

Someone treats it as a consent workaround. It is not one. Serving a tag from your own domain does not change what a visitor consented to. Consent signalling still has to be implemented properly, and a firm that gets this wrong has a compliance problem rather than a measurement one.

What we do about it

Law Ops Forge deploys first-party measurement as the first step of every Google Ads and Local Services Ads engagement, before touching a single bid. We verify first-party status on every domain, run a live conversion test, re-test at 24 hours, and check that the firm’s own security rules are not challenging the measurement path.

Then we report the change honestly: what was recovered, what was always there, and what actually moved as a result.

If you would like to know how much of your firm’s enquiry volume is currently invisible to Google, ask us for a measurement audit. It is free, it takes about twenty minutes of your time, and the findings are yours whether or not you ever work with us.

Sources

Every claim above comes from Google’s own documentation. These are the pages it came from, so you can check any of it yourself, or send it to whoever else needs convincing.

• Google Ads Help: Google tag gateway for advertisers, where the 11% figure is published
• Google Ads Help: set up Google tag gateway with Cloudflare
• Google Ads Help: set it up in Google Tag Manager with Cloudflare
• Cloudflare: Google tag gateway for advertisers
• Google Ads Help: about consent mode

Related reading

• The Conversions You Are Allowed to Model