Nordic SaaS Reduces CAC 43% with AI Optimization

Challenge

A 150-person Nordic SaaS company was spending €50,000 per month on Google Ads and Meta Ads.

Their customer acquisition cost (CAC) had crept up to €420 per customer. Way above their target of €250.

The marketing team knew something was wrong. But they couldn't find it.

The problem:

• 47 active campaigns across Google and Meta
• 1,200+ keywords in Google Ads
• 23 Meta ad sets with overlapping audiences
• Too much data to analyze manually
• No clear visibility into what was working

They'd tried:

• Hiring freelance PPC experts (inconsistent results)
• Using dashboards like Supermetrics (showed data, no answers)
• Manual analysis (couldn't keep up with volume)

Nothing worked.

CAC kept climbing.

Solution

They connected Cogny to their Google Ads, Meta Ads, and BigQuery (with GA4 data).

Setup took 15 minutes.

Within 24 hours, Cogny's AI generated 23 growth tickets.

The top 5 tickets:

Ticket #1: "Pause 47 Zero-Conversion Keywords"

• Keywords that had spent €4,200 with zero conversions
• AI identified them across all campaigns
• One-click pause

Ticket #2: "Consolidate Overlapping Meta Audiences"

• 3 ad sets targeting nearly identical audiences
• Creating auction competition with themselves
• Recommended consolidation into single ad set

Ticket #3: "Reallocate Budget to High-ROAS Campaigns"

• Campaign "Product Demo Viewers" converting at 8.2 ROAS
• Only getting 8% of total budget
• AI recommended increasing to 25%

Ticket #4: "Add 83 Negative Keywords"

• Search terms triggering ads but never converting
• Examples: "free marketing software", "open source alternative"
• Bleeding €1,800/month

Ticket #5: "Pause Underperforming Geographic Regions"

• Certain regions converting at <1.0 ROAS
• No clear path to profitability
• Recommended pausing, reallocating budget to better regions

Implementation

Week 1:

• Reviewed all 23 tickets
• Implemented top 10 tickets (2 hours total)
• Paused zero-conversion elements
• Consolidated audiences
• Adjusted budget allocation

Week 2:

• AI generated 17 new tickets based on updated campaigns
• Implemented another 8 tickets
• Started seeing CAC improvement

Week 3:

• Continued iterative optimization
• Added more negative keywords
• Refined audience targeting
• CAC dropped below €300

Week 4-8:

• Steady optimization
• New tickets focused on scaling what worked
• AI identified expansion opportunities
• CAC stabilized at €240

Results

Primary Metrics (After 8 Weeks)

Customer Acquisition Cost:

• Before: €420
• After: €240
• Improvement: 43% reduction

Monthly Ad Spend:

• Stayed flat at €50,000/month
• But acquired 75% more customers for same spend

Time Spent on Optimization:

• Before: 20 hours/week (2 team members)
• After: 4 hours/week (1 team member reviewing tickets)
• Saved: 16 hours/week = €6,400/month in labor costs

Secondary Metrics

Campaign Efficiency:

• Google Ads ROAS: 2.8 → 4.9 (75% increase)
• Meta Ads ROAS: 3.1 → 5.2 (68% increase)

Wasted Spend Eliminated:

• Paused elements that spent €11,200/month with poor ROI
• Reallocated to high-performers

Customer LTV:

• Discovered that customers from "Product Demo Viewers" campaign had 2.1x higher LTV
• Shifted more budget there
• Overall customer LTV increased 18%

Team Morale:

• Marketing team shifted from "fighting fires" to strategic planning
• Could finally focus on creative and positioning
• Less burnout from manual optimization

Key Insights from AI

1. Hidden Audience Overlap

The team didn't realize their Meta ad sets were competing against each other.

67% audience overlap across 3 major ad sets.

This drove up CPMs and wasted budget.

Consolidation fixed it immediately.

2. Keywords That "Look Good" But Don't Convert

Many keywords had decent CTR and low CPC. Management thought they were working.

But AI tracked full journey: zero actual conversions.

Just wasting money on clicks that never closed.

3. Budget Allocation Based on Vanity Metrics

They'd been allocating budget based on CTR and impression share.

AI showed that campaigns with LOWER CTR sometimes had HIGHER conversion rates.

Better to optimize for end result, not clicks.

4. Geographic Performance Variance

Different regions had wildly different CAC:

• Sweden: €180 CAC (home market, great fit)
• UK: €310 CAC (decent)
• Germany: €620 CAC (terrible ROI)

Same bid strategy everywhere = inefficient.

AI recommended geo-specific bidding.

5. Creative Fatigue Timing

Meta ads would perform great for 2-3 weeks. Then cliff-dive in performance.

Manual monitoring missed the timing.

AI detected the pattern and recommended refresh schedule:

• Rotate creative every 18 days
• Have 3 variations ready
• Systematic testing approach

What The Team Said

"We knew we had optimization opportunities. But we couldn't find them in the noise. Cogny's AI found €11,000/month in wasted spend in 24 hours. It would've taken us months to discover manually—if we ever found it at all."

— Head of Growth

"The time savings are huge. We went from spending 20 hours a week in spreadsheets to 4 hours reviewing AI tickets. The team is happier, and results are way better."

— Marketing Manager

"Most tools show you data. Cogny tells you exactly what to do. That's the difference. Every ticket is specific: 'Pause this', 'Increase budget here', 'Add these negative keywords'. We just execute."

— Senior Performance Marketer

Lessons Learned

1. Manual Optimization Doesn't Scale

With 47 campaigns and 1,200 keywords, humans can't keep up. You miss opportunities. You miss waste.

AI analyzes everything, continuously.

2. Budget Allocation Matters More Than You Think

Small reallocation shifts (moving 5-10% of budget to better campaigns) compound into huge results.

But you can't reallocate well without seeing the full picture.

3. Zero-Conversion Elements Hide Everywhere

Every account has them. Keywords, audiences, placements, regions that consume budget but never convert.

The longer they run, the more they cost.

AI finds them fast.

4. Audience Overlap is Invisible in Native Platforms

Meta doesn't clearly show when your ad sets compete with each other. Google doesn't highlight keyword cannibalization.

AI spots these patterns immediately.

5. Optimization is Continuous

What works this month stops working next month. Creative fatigues. Audiences saturate. Competitors change tactics.

You need continuous monitoring and optimization. Humans can't do this at scale. AI can.

Why It Worked

Data Completeness: They connected everything: Google Ads, Meta Ads, GA4 via BigQuery.

AI could see the full customer journey. Not just ad platform data.

Fast Implementation: They acted on tickets quickly.

Many teams analyze but don't execute. Speed matters.

Trust the AI: They started by manually verifying recommendations.

After seeing the AI was right 90%+ of the time, they moved faster.

Iterative Optimization: They didn't expect perfection in week 1.

Each week, new tickets. Each week, incremental improvement.

Compounding gains over 8 weeks = 43% CAC reduction.

Replicability

This result is replicable if you have:

✅ At least €20K+/month ad spend (enough data for AI to analyze) ✅ Multiple campaigns and keywords (more data = more optimization opportunities) ✅ BigQuery or GA4 data (for full-funnel visibility) ✅ Willingness to implement recommendations quickly

Typical timeline:

• Week 1: Find quick wins (pause wasted spend)
• Week 2-4: Optimize allocation (shift budget to winners)
• Week 5-8: Scale and refine (continuous improvement)

Expected results:

• 20-40% CAC reduction (depends on starting efficiency)
• 10-20 hours/week time savings
• 30-50% ROAS improvement

What's Next for Them

Now that core efficiency is fixed, they're focusing on:

1. Scaling Profitably

• Increase ad spend while maintaining €240 CAC
• AI identifies new expansion opportunities
• Testing new channels with confidence

2. Creative Optimization

• Using AI insights to inform creative strategy
• "AI says testimonial + demo format works 2.1x better"
• Systematically testing AI recommendations

3. LTV Optimization

• Now optimizing for customer LTV, not just CAC
• AI identifies which sources drive highest-value customers
• Shift budget to quality, not just quantity

4. Predictive Analysis

• AI forecasting: "If you increase spend 30%, here's expected CAC"
• Scenario planning for quarterly budgets
• Data-driven planning instead of guesswork

Want Similar Results?

If you're spending €20K+/month on ads and feel like you're missing opportunities, you probably are.

On average, companies waste 20-30% of ad spend on:

• Keywords/audiences that don't convert
• Budget misallocation
• Invisible audience overlap
• Geographic inefficiency

The waste is there. You just can't see it without AI analyzing everything.

See what AI finds in your campaigns:

Schedule a demo and we'll do a live analysis of your accounts.

We'll show you exactly what tickets Cogny would generate. Most teams find $5K-$20K/month in quick wins in the first session.

---

About This Case Study

Written by the Cogny team—built by the founders who created AI optimization systems for Netflix, Zalando, and Momondo at Campanja, and scaled growth for Kry, Epidemic Sound, and Yubico through GrowthHackers.se.

Company details anonymized to protect client confidentiality. Results verified and representative of typical outcomes.

Last Updated: December 5, 2024