Liam Collins, Founder of We Scale SaaS, joins E-Coffee with Experts, with our host Dawood Bukhari, to talk about how paid marketing works now that tracking is patchy and AI runs most of the bidding. Liam started in large network agencies, moved in-house, and then built his own agency after seeing performance stall for lack of experienced operators and infrastructure.
He explains how his team measures paid campaigns when opt-in rates can drop below 50%: platform attribution used directionally, outcomes checked against the CRM, and media mix modeling and incrementality testing where the data allows. In one LinkedIn test, people came back from the original ad touch around the two-and-a-half-month mark, and revenue showed up around month four. Judged on multi-touch attribution alone, the campaign would have looked like a failure.
Liam also covers why automated bidding is old news for performance marketers, how a cost-per-lead goal handed to AI can fill a pipeline with poor-fit leads, and why AI-generated creative is pulling brands toward the same average ideas. He shares how We Scale SaaS feeds sales call transcripts into the CRM to shape content strategy, and why he sees brand as the real moat as AI starts shortlisting vendors for buyers.
Takeaways:
- Use platform attribution directionally, then check results against your CRM.
- One LinkedIn campaign drove revenue around month four, not week one.
- Judged on multi-touch attribution alone, good campaigns can look like failures.
- Tie every platform metric down to closed-won opportunities.
- Pitch paid budgets in commercial terms your CFO actually understands.
- Keep 80% of budget on proven campaigns and 20% for testing.
- Launching too many campaigns at once hides what is actually working.
- AI chasing cheap leads can fill your pipeline with poor-fit buyers.
- AI-made creative pulls brands toward the same average ideas.
- Sales call insights make better content inputs than generic AI research.