
Kalma: 7x the spend, 7x the qualified leads
Weak alignment between the business and acquisition, plus under-tended Paid Media campaigns, could not scale the volume of qualified leads Sales needed to work.
- Spain
- Retainer
- 2 years (and counting)
CPL
-21%
Monthly spend
715%
Qualified leads
+730%
An account struggling to scale: acquisition was not connected to the business. The Paid campaign structure was simple and did not fit Kalma's needs. There was no creative testing and no data-driven iteration.
Pau brings a much broader skill set than just campaign management: creative, tracking, reporting and performance structure, everything that actually makes paid media work over the long term.
Strategy · 01
The approach
We started with the infrastructure. We rebuilt the pixel conversions, sent CRM data back to the platforms via Zapier, and restructured campaigns on both Google and Facebook with a scaling structure.
What changed
With clean data flowing into the ad platforms, we migrated the campaigns: first to Maximize Conversions and then to Target CPL. We tested different landing pages until we found the ones bringing the best qualified leads.
Use case · Paid Media · 02
Scaling spend without breaking CPL
With tracking rebuilt on the signals that matter, we shifted the optimisation of the new campaigns toward CRM conversions. The landing-page structure in Performance Max and Meta helped us scale spend toward the parts of the site with the most volume and the best quality.
Monthly spend
+715%
Qualified leads
730%
Use case · Reporting · 03
From click to sale, in a single panel
We built a dashboard that connects the whole funnel from the ad click to the final sale, recorded inside the CRM and visible thanks to the campaign UTMs.
Time spent on reporting
-100%
4 h/week 0 h
Use case · Anomalies · 04
Alerts that warn before it hurts
A daily monitor watches spend, CPL and tracking signals, and pings Slack when something drifts out of range. That is how we caught volatility on TikTok and the onset of ad fatigue on Meta.

Anomaly detection
-99%
hours seconds
Use case · Creative analysis · 05
AI-powered creative analysis
Every creative is analysed against its CTR and CPL: which message works, which one has burned out, and what to iterate in the next batch. Testing decisions are made on data, in minutes.
Anonymised dataImpact · 06
The results
A year and a half later we have multiplied the spend while holding CPL and cost per qualified lead steady over time.
The key part: it held. Optimising on CPL signals, the algorithms stopped chasing duplicate or irrelevant conversions and performance stayed steady quarter over quarter.
Tools
- GA4
- Google Tag Manager
- Google Ads Editor
- Looker Studio
- Consent Mode v2
- Claude / MCP
Got a similar challenge?
Tell me your situation (platforms, accounts, what you want to move) and I’ll reply within 24h.
Let's talk