Google Ads CPA Reduction Case Study (2026): Kathmandu Service Business
This case study follows the STAR-L framework from our Marketing Case Studies Hub — anonymized to protect client confidentiality while sharing actionable tactics any Google Ads practitioner can apply.
→ Case Studies Hub · Negative Keywords Master List · Google Ads Audit Checklist · CPA (dictionary)
At a glance
| Metric | Before | After | Change |
|---|---|---|---|
| Monthly spend | NPR 85,000 | NPR 72,000 | −15% |
| Conversions/month | 18 | 28 | +56% |
| CPA | NPR 4,722 | NPR 2,571 | −42% |
| Conversion rate | 3.1% | 5.8% | +87% |
| Wasted spend (irrelevant queries) | ~35% | ~8% | −27pp |
| Timeline | — | 90 days | — |
S — Situation
Client profile (anonymized):
- Home services business in Kathmandu (plumbing + electrical)
- Running Google Ads for 14 months with a previous agency
- Single Search campaign, broad match keywords, Maximize Conversions bidding
- Landing page: generic homepage, not service-specific
- Tracking: GA4 + Google Ads conversion import (phone calls + form fills)
Starting metrics (Month 0):
- NPR 85,000/month ad spend
- 18 conversions/month (phone calls + form submissions)
- CPA: NPR 4,722
- Quality Score average: 4.2/10
- 580 clicks/month, 3.1% conversion rate
The owner knew ads were “working” but suspected money was wasted — especially on queries that never converted.
T — Task / Challenge
Three specific problems identified in the initial Google Ads audit:
| Problem | Evidence | Impact |
|---|---|---|
| No negative keywords | Search Terms Report showed 35% spend on irrelevant queries | NPR ~30,000/month wasted |
| Single campaign structure | All services in one campaign — no budget control per service | Plumbing CPA 2× electrical CPA, hidden in aggregate |
| Homepage landing page | 3.1% CVR vs. industry benchmark 5–8% for service pages | Losing 40%+ of potential conversions |
| Broad match without guardrails | Keywords like “plumber” triggered “plumber jobs Kathmandu”, “plumber course” | Irrelevant traffic inflating spend |
Goal: Reduce CPA below NPR 3,500 while maintaining or increasing conversion volume.
A — Actions
Month 1: Stop the bleeding
Week 1–2: Negative keyword blitz
Applied 180+ negatives from our Negative Keywords Master List plus account-specific terms mined from 90 days of Search Terms Report:
| Category | Example negatives | Spend saved |
|---|---|---|
| Job seekers | “jobs”, “vacancy”, “salary”, “hiring” | NPR 8,200/mo |
| DIY learners | “how to”, “tutorial”, “course”, “training” | NPR 4,100/mo |
| Wrong service | “carpenter”, “painter”, “AC repair” (not offered) | NPR 3,800/mo |
| Free seekers | “free”, “cheap”, “DIY fix” | NPR 2,900/mo |
| Competitor brands | 4 competitor business names | NPR 1,400/mo |
Week 3–4: Campaign restructure
Split single campaign into three service-specific campaigns:
| Campaign | Budget share | Keywords | Landing page |
|---|---|---|---|
| Plumbing — Emergency | 45% | “emergency plumber kathmandu”, “pipe leak repair” | /plumbing-emergency/ |
| Plumbing — General | 20% | “plumber near me”, “plumbing service kathmandu” | /plumbing-services/ |
| Electrical | 35% | “electrician kathmandu”, “wiring repair” | /electrical-services/ |
Each campaign got its own ad groups (exact + phrase match), RSA ads with service-specific copy, and dedicated landing pages.
Month 2: Optimize and scale
Landing page improvements:
- Added click-to-call button above fold (mobile — 85% of traffic)
- WhatsApp contact button
- Service-specific pricing ranges in NPR
- 6 genuine Google reviews embedded
- Page load time reduced from 4.2s to 1.8s (mobile)
Bidding transition:
- Switched from Maximize Conversions to Manual CPC for 3 weeks (clean data collection)
- After 45 conversions in Plumbing — Emergency campaign, enabled Target CPA at NPR 3,200
- Electrical campaign stayed Manual CPC (insufficient conversion volume)
Ad copy refresh:
- Added call extensions with tracking number
- Location extensions linked to GBP
- RSAs with service-specific headlines: “24/7 Plumber Kathmandu — Call Now”
Month 3: Refine and document
- Weekly Search Terms Report review — added 15–20 new negatives per week
- Paused 3 ad groups with CPA > NPR 5,000 after 30+ clicks
- Increased Plumbing — Emergency budget by 15% (best CPA campaign)
- Set up offline conversion import for jobs completed (not just leads)
R — Results
| Metric | Month 0 | Month 3 | Change |
|---|---|---|---|
| Spend | NPR 85,000 | NPR 72,000 | −15% |
| Conversions | 18 | 28 | +56% |
| CPA | NPR 4,722 | NPR 2,571 | −42% |
| CVR | 3.1% | 5.8% | +87% |
| Quality Score avg | 4.2 | 6.8 | +2.6 |
| Impression share | 42% | 58% | +16pp |
Campaign-level breakdown (Month 3):
| Campaign | CPA | Conversions | % of total |
|---|---|---|---|
| Plumbing — Emergency | NPR 2,100 | 14 | 50% |
| Plumbing — General | NPR 3,400 | 6 | 21% |
| Electrical | NPR 2,950 | 8 | 29% |
ROI context: Average job value NPR 4,500 (plumbing) / NPR 6,200 (electrical). At NPR 2,571 CPA, the account became profitable on first-visit jobs — previously breakeven at best.
L — Learnings
What worked
- Negative keywords first — biggest single impact. 35% waste eliminated in 2 weeks.
- Service-specific landing pages — CVR nearly doubled. Never send paid traffic to homepage.
- Campaign separation by service — revealed that aggregate CPA hid a 2× difference between services.
- Manual CPC before automated bidding — cleaner data before enabling Target CPA.
What we’d do differently
- Offline conversion import from day 1 — many “conversions” were low-quality leads (price shoppers who never booked). Tracking completed jobs would have optimized sooner.
- Call recording review — would have identified that 20% of phone leads were wrong-service calls despite negatives.
- Ad schedule analysis earlier — emergency plumbing converts 3× better 6 AM–10 PM; night spend was wasted.
Applicable to your account
| If you see this… | Do this |
|---|---|
| CPA above industry benchmark | Run Google Ads Audit — check negatives first |
| Broad match with no negatives | Apply Negative Keywords Master List immediately |
| Single campaign for multiple services | Split by service with dedicated landing pages |
| CVR below 4% on service pages | Build service-specific landing pages with click-to-call |
| Fewer than 30 conversions/month | Stay on Manual CPC — don’t enable Target CPA yet |
Tools and resources used
- Google Ads Audit Checklist
- Negative Keywords Master List
- Google Ads Glossary
- Conversion Tracking Mastery Guide
- Marketing Case Studies Hub
Related case studies
More proof from the Case Studies Hub:
- SEO traffic growth case studies (coming)
- Meta Ads ROAS improvement case studies (coming)
- E-commerce conversion rate optimization (coming)
Have a Google Ads account that needs this level of optimization? See Google Ads Nepal for service details.
