Case Study · 10 min read
B2B SaaS — 3.8× ROAS Case Study
Project Overview & Situation
A Series A SaaS company had been running ads in-house for 14 months. ROAS was averaging 1.6× on Google, with Meta at breakeven and LinkedIn effectively off. The founder suspected the attribution was wrong. It was: they were crediting the last click, which made brand terms look like winners and made prospecting look useless. True incremental ROAS was closer to 1.1×.
Engagement Context
- Sector: B2B SaaS
- Services: Performance Marketing, AI Agents
- Timeline: 6 months
- Result: 3.8× ROAS · ₹18 avg CPL
The Strategic Challenges
Broken attribution
Last-click was hiding where leads actually came from. The model was crediting brand keywords that users searched after seeing prospecting ads — making prospecting look useless.
Under-indexed on LinkedIn
The highest-intent B2B channel was turned off because the attribution model made it look expensive. LinkedIn was actually generating 38% of final-stage intent signals.
Creative fatigue
Same three ad formats running for 11 months with no hook testing. Frequency was high, novelty was zero. Effective CPL had quietly doubled.
Our Implementation & Playbook
Attribution rebuild (Weeks 1–3)
Implemented data-driven attribution across GA4 and the CRM. Mapped touchpoints to closed revenue, not MQLs. Found that LinkedIn was generating 38% of final-stage intent signals that were being attributed to branded Google search. The founder's instinct was right — the data was lying.
Budget reallocation (Week 3)
Shifted 22% of Google budget to LinkedIn. Launched proper prospecting campaigns with intent-matched sequences based on job title, company size, and technographic signals. Paused 14 underperforming ad sets immediately.
Creative overhaul (Weeks 4–8)
Built 14 new creative variants based on actual customer language from sales call transcripts. Ran structured hook tests with 48-hour cycles. Winning format: 15-second problem-statement video — no branding in first 3 seconds, pain point stated in first 5 words.
AI bid management (Ongoing)
Deployed automated bid strategy agents that adjusted bids based on CRM conversion lag, not platform-reported conversions. Cut wasted spend 31%. The key insight: SaaS sales cycles mean platform signals lag reality by 3–4 weeks — the agents accounted for this.
Measurable Results & Outcomes
Sustained ROAS
3.8× — Across all three platforms at month 6. Not a launch spike — verified by independent attribution at the 3-month and 6-month marks.
Average CPL
₹18 — Cost per qualified lead at month 6. Baseline at engagement start: ₹74. A 76% reduction in cost per lead.
Wasted spend eliminated
31% — Ad spend eliminated through AI bid management — redirected to highest-performing channel-format combinations.
Client Impact & Perspective
"The attribution rebuild alone was worth the engagement. We found out we were spending ₹40L/year on keywords that looked like winners but weren't." — CFO, B2B SaaS (name withheld)
Core Competencies Applied
Performance Marketing, AI Agents & Automation, Attribution Consulting
Want to apply this strategy to your business?
Understanding the strategy is step one. Implementing it flawlessly is the real challenge. Tell us about your goals and we will suggest the next move in 1 working day.