The Starting Point
The company in this case study is a B2B SaaS offering project management software for teams of 10-200 people. They were spending €15,000/month on Google Ads across three campaigns: Brand (branded search terms), Generic (category keywords like "project management software"), and Competitor (competitor brand names).
The numbers at the start of the engagement:
- Average CPA: €280, nearly double the €160 target
- Monthly conversions: 54 (trial signups, 14-day free trial as conversion action)
- Average Quality Score: 5.2 across all keywords
- Keyword match types: 60% broad match, remainder phrase and exact
- Scheduling: 24/7 delivery, no ad schedule adjustments
- Negative keyword lists: none applied at campaign level
The gap between actual CPA (€280) and target (€160) was costing approximately €6,500/month in excess acquisition costs, on top of the budget already being spent. The mandate was clear: hit €160 without increasing budget.
An initial audit gave the account a health score of 34/100, low Quality Scores, zero negative keyword coverage, and no ad scheduling were the primary flags. Each issue was independently fixable, which made this a strong candidate for optimization before any structural changes.
The Audit: What We Found
A systematic audit of the account surface-level data, search term reports, Quality Score breakdown, and conversion path analysis revealed four critical problem areas:
The search term report was particularly damning. Broad match keywords like "project management" were serving for queries including "project management certification", "project management degree", "agile project management course", and dozens of other informational terms with zero commercial intent. These weren't edge cases, they represented 28% of total impressions.
The overnight hours were the single biggest wasted spend category. The Generic campaign was spending approximately €2,100/month between 10pm and 7am, generating 6 clicks that resulted in zero trials over the 90-day sample period. This was an immediate, zero-risk fix.
Month 1: The Quick Wins
The first month focused exclusively on high-confidence changes with immediate impact and minimal risk. No structural changes, no new campaigns, no bidding strategy changes.
Month 2: Structural Changes
With the quick wins banked, Month 2 focused on deeper structural improvements that required more lead time but offered larger long-term impact.
Keyword Match Type Restructure
The 60% broad match composition was replaced: every broad match keyword was converted to phrase match equivalents, and the top 30 highest-converting terms were also added as exact match with higher bids. This gave precise control over which queries triggered ads while maintaining discovery capability through phrase match.
RLSA Campaign Layer
A remarketing layer was built targeting users who had visited the pricing page but not signed up (a high-intent cohort). These users saw a dedicated set of ads emphasizing the free trial with social proof ("Join 12,000 teams already using [Product]"). This remarketing campaign launched in week 5 and ran on Maximize Conversions bidding.
Bidding Strategy Transition
The Generic campaign transitioned from Manual CPC to Target CPA with an initial target of €220 (the actual current CPA, not the aspirational goal). The Brand campaign remained on Manual CPC due to very high conversion rates making manual bidding optimal. Competitor campaign moved to Maximize Conversions to build more data before setting a target.
Landing Page Optimization
Two landing page changes were implemented in partnership with the client's design team. First, social proof was moved above the fold, a testimonials bar showing 3 company logos (mid-market customers) was placed directly below the hero headline. Second, the trial signup form was simplified from 7 fields to 4 (Name, Work Email, Company Name, Team Size), removing the phone number, job title, and "How did you hear about us" fields.
The form simplification alone produced a 58% increase in form completion rate, from 18% of landing page visitors starting a trial, to 28.4%. Combined with the ad copy improvements, the blended conversion rate across all campaigns rose from 1.2% to 2.9% over 60 days.
The Numbers After 60 Days
Results vs. starting point, same €15,000/month budget throughout:
| Metric | Day 1 (Before) | Day 60 (After) | Change |
|---|---|---|---|
| Average CPC | €3.40 | €2.10 | −38% |
| CTR (avg) | 2.1% | 3.8% | +81% |
| Conversion Rate | 1.2% | 2.9% | +142% |
| CPA | €280 | €160 | −43% |
| Monthly Conversions | 54 | 94 | +74% |
| Avg Quality Score | 5.2 | 7.1 | +37% |
| Wasted Spend/mo | ~€4,200 | ~€620 | −85% |
The most important result: 74% more conversions from the same budget. The team had previously believed they needed a larger budget to hit their acquisition targets. The audit revealed that a significant portion of every euro spent was generating no return, and fixing that efficiency gap was worth more than any budget increase.
Key Lessons
Three takeaways that apply to virtually any Google Ads account running with similar symptoms:
Most accounts have never done a thorough search term audit. The 140 negative keywords in this account were not obscure edge cases, they were obvious, high-volume irrelevant terms that had been burning budget for months. Pull 90 days of data, sort by spend, look for zero-conversion terms with 10+ clicks, and add negatives systematically. This is the highest ROI activity in most accounts.
B2B SaaS buyers are almost never making purchase decisions at 2am. If your audience is business users making considered purchases, turning off or significantly reducing bids during non-business hours is a near-zero-risk change. The conversion rate differential (0.3% vs 4.1% in this account) is extreme but not unusual. Check your own hourly conversion data before assuming 24/7 delivery is optimal.
Optimizing ads while ignoring the landing page is like improving the packaging on a product with a broken lid. In this account, the landing page had a 7-field form that users were abandoning at 82%. No amount of ad optimization could compensate for a form that was too long. Landing page conversion rate is a campaign metric, not a UX metric, it belongs in the same optimization conversation as keywords and bids.