CS 02 B2B SaaS
Case Study 02
B2B SaaS / Performance Marketing

3.8× ROAS.
Six months running.

Not a spike. Not a single platform win. Sustained return across three channels with attribution that held up under audit.

Sector B2B SaaS
Services Performance Marketing, AI Agents
Timeline 6 months
Result 3.8× ROAS · ₹18 avg CPL
The 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×.

01 / The challenge
I

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.

II

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.

III

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.

02 / What we built
1

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.

2

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.

3

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.

4

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.

Results
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.
"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)
Services used in this engagement
Performance Marketing AI Agents & Automation Attribution Consulting
← Back to all work

Find out what your
attribution is hiding.

We'll run a free performance audit — check where your budget is going versus where it should be going, and what the real numbers look like.

Get your free audit
48-hour reply No obligation India HQ · Remote