Module 2 · Card 1 of 8

Where marketing attribution gets real now

Marketing analytics has been promising data-driven decisions for fifteen years. AI is the first thing that delivers the 'what'. But the 'why' is still yours — and that gap is where most teams get it wrong this year.

Module 2 · Card 2 of 8

What AI is genuinely good at

Pattern detection at scale. Anomaly flagging across hundreds of campaigns. Multi-touch attribution rollups that a human couldn't realistically build by hand. Channel-mix optimisation suggestions. Weekly performance summaries in plain English. Stakeholder-shaped narrative drafts that you then sharpen.

Module 2 · Card 3 of 8

What the evidence shows

Adobe Digital Trends 2026: 76% of CMOs now use AI in performance reporting. McKinsey Marketing & Sales 2026: AI-aided campaign optimisation produced an average 14% lift in ROAS for B2B marketers — but with high variance (-8% to +41% by team). The variance is the story: whether your team gets the lift depends on what you do with the AI output.

Module 2 · Card 4 of 8

Where the dashboards confidently lie

AI doesn't know your brand-narrative goals. It optimises to whatever metric is loudest in the data — usually clicks or short-term conversions. If your strategy is brand-led, the AI will systematically tell you to defund brand and feed performance, week after week, and the numbers will agree with it. Until they don't.

Module 2 · Card 5 of 8

Own the why, let AI handle the what

What the data is doing: AI's job. Anomaly detection, summarisation, attribution rollups. Why the data is doing it: yours. Did we underspend brand last quarter and is that why CAC went up? Did the new website hurt conversion or are we just attribution-tracking it differently? The AI explains the patterns; you decide the cause.

Module 2 · Card 6 of 8

Stakeholder reports specifically

AI is now extremely good at drafting the weekly performance report. But the smooth narrative buries the campaign that's quietly leaking budget, because the LLM optimises for prose flow not for what your CFO needs to act on. Use AI to assemble the deck; you write the narrative section that names the underperformer and the plan.

Module 2 · Card 7 of 8

Channel mix — the tactical play

AI channel-mix tools (GA4 Insights, HubSpot Performance Hub, Adobe Marketo Engage) routinely produce defensible recommendations. The trap: they recommend short-term ROAS optimisation, which means more of what worked last week. If you only follow the AI's recommendation you end up over-indexed on the best-attributable channel, which is rarely the best long-term channel.

Module 2 · Card 8 of 8

End of module

Next: five questions. Two minutes. Then open your latest AI-generated performance report and ask: which narrative did it write FOR me, and which sentence did I need to write myself?

Where marketing attribution gets real now: AI, channel mix, and the numbers you actually defend

Where marketing attribution gets real now