LAB LIVE
SCANNING416 CAMPAIGNS
ANOMALIES15 FLAGGED
AVG LIFT+4.40× ROAS
Back to cases
SL-MT-203REV.ACLASSIFIED · CLIENT-VIEW
HEALTHCARE · BERLIN, GERMANY

Mitte Dental

Private dental practice · WhatsApp + web attribution

Healthcare
[00]DIAGNOSIS
Attribution gap: 58% of leads arrived via WhatsApp, invisible to all reporting. Bidding strategy optimized on form fills — the minority channel — actively penalizing the audience most likely to convert. CPL appeared high because half the conversion volume did not exist in the data.
WhatsApp share of leads
58%
invisible pre-tracking
Attribution gap closed
100%
WhatsApp + web unified
CPL (after)
€42
all channels combined
A/B winner
Specialist
vs. free consultation
Tracking method
Server-side
GA4 · session-accurate
Dashboard cadence
Weekly
plan-fact client review
[01]SITUATION

Mitte Dental is a private dental practice in Berlin offering general dentistry alongside specializations in implants, veneers, and cosmetic treatments. Patient acquisition runs via Google Ads targeting high-intent queries in the Berlin market.

At audit, the practice had two active lead channels: a website contact form and a WhatsApp click-to-chat button. Only the form was tracked. The WhatsApp channel — preferred by the majority of patients — produced no data visible to the campaign. Smart Bidding had been optimizing against an incomplete signal for months.

[02]ATTRIBUTION ARCHITECTURE

The full attribution rebuild covered both channels and added server-side collection to recover the sessions lost to consent mechanisms.

  • WhatsApp click-to-chat trackingPrimary conversion signal

    WhatsApp button tap tracked as a conversion event in GA4 and passed to Google Ads — first time this channel contributed to bidding signal

  • Server-side GA4Session attribution recovery

    Browser-based GA4 was losing 30–40% of sessions due to German cookie consent rejection rates and Safari ITP. Server-side collection with first-party data closed the gap

  • Unified conversion modelForm + WhatsApp combined

    Both lead channels merged into a single conversion type with channel source tagging — bidding now optimizes on the real combined lead volume

  • Plan-Fact dashboardWeekly client reporting

    Automated dashboard showing actual leads vs. monthly target by channel, week, and campaign — replaced manual reporting with a shared live view

[03]A/B TEST — OFFER POSITIONING

With attribution fixed and a reliable lead count established, two positioning strategies were tested against each other across identical campaign structures.

Free consultation offer

Lead volume higher. Appointment-to-show rate lower. Many leads came in expecting a completely free initial service rather than a consultation credit.

Higher volume · lower close rate
★ WINNER
Specialist positioning

Messaging emphasized implant and veneer expertise. Lead volume slightly lower but appointment-to-show rate significantly higher. Higher-intent patients self-qualified.

★ Lower volume · higher close rate · better CPL adjusted for quality
[04]KEY INSIGHTS
01WhatsApp is the primary lead channel in German healthcare — and almost always untracked

German patients, especially in urban private dental practices, strongly prefer WhatsApp for initial contact. They will not fill out a form. They will not call. They will tap the WhatsApp button and send a message. If that tap is not tracked, the campaign that drove it gets zero credit — and Smart Bidding optimizes away from the audience most likely to convert.

02Optimizing on form fills alone inverts the bidding signal

In a practice where 58% of leads come through WhatsApp, a bidding strategy optimizing on form fills is actively selecting against the majority of actual clients. The algorithm learns that the audiences who convert through WhatsApp are 'non-converting' — and reduces bids accordingly. The result is a steadily worsening feedback loop.

03Free consultation offers attract the wrong audience in high-CPL dental markets

The A/B test showed specialist positioning — emphasizing clinical expertise and implant/veneers specialization — outperformed the free consultation offer on lead-to-appointment conversion. Free consultation attracted volume but lower closing rate. Specialist positioning attracted fewer but higher-intent leads at a better CPL after accounting for close rate.

04Server-side GA4 is necessary for accurate attribution in consent-heavy markets

Germany has among the highest cookie consent rejection rates in Europe. Browser-based GA4 was losing 30–40% of session attribution due to consent non-acceptance and Safari ITP. Server-side collection with first-party data recovered the gap — making campaign-level decisions based on what users actually did, not what the browser was able to record.

[05]OUTCOME

Full lead attribution established across WhatsApp and web for the first time. 58% of leads — previously invisible — entered the conversion model and began informing bidding decisions. Combined CPL stabilized at €42 across all lead channels. A/B testing confirmed specialist positioning as the stronger offer for high-intent patient acquisition.

Root cause: WhatsApp channel (58% of leads) invisible to bidding signal
Fix: WhatsApp tracking + server-side GA4 + unified conversion model
A/B winner: specialist positioning over free consultation · CPL: €42 unified

Client name is a pseudonym. Report ID is internal. Metrics reflect specific account conditions — results vary by market, budget, and funnel maturity.