Your YouTube line item sits in the brand budget, but its week-on-week conversion pattern tracks your paid search performance with a two-day lag. This matters because brand and DR channels carry different saturation curves. Applying brand decay rates to a DR-behaving channel will cause your budget optimiser to undervalue YouTube by roughly 20 to 35 percent.

Your podcast sponsorship line shows identical spend in every single week of the panel: $8,750 AUD, no variation. This is almost certainly a fixed-fee placement recorded as a weekly figure rather than a variable buy. That is fine operationally, but a zero-variance channel produces no signal for the marketing mix model. Including it as-is will inflate your model's R-squared without improving predictive accuracy.

Between 28 October and 11 November, your Meta and TikTok platforms both claimed the same conversion events via last-touch attribution. The overlap accounts for an estimated 890 duplicate conversions across those three weeks, representing roughly $67,000 in misattributed revenue. Your actual CPA for both channels during this period is higher than your dashboards show.

The Alchemist flagged 14 percent of programmatic rows for missing impression data across weeks 2, 6, and 9. This is above the 10 percent threshold where spend-per-impression ratios become unreliable. The affected rows are still usable for spend analysis, but impression-based CPM calculations for those weeks should be treated as estimates.

The Auditor attempted to cross-reference your spend panel against category-level search trend data and competitor spend indices, but no third-party data source is connected to your workspace. Connecting one source would allow the MMM to separate your brand's organic contribution from paid lift, which typically improves model accuracy by 15 to 25 percent.

Your paid search panel shows 80 percent of spend on branded terms (brand name + product name variants) and 20 percent on non-branded. Branded keywords capture demand you largely already own. At this ratio, the measured ROAS of 6.8x is significantly inflated by organic demand that would have converted without paid assistance. Your true incremental ROAS on paid search is likely 2.5 to 3.5x.

The Meta and TikTok panels show average frequency climbing from 2.8 in weeks 1 to 7 to 4.7 in weeks 8 to 12. CTR fell 31 percent across the same period while spend held flat. This pattern is consistent with creative fatigue. You are paying for impressions that are actively degrading brand perception rather than driving consideration.

There are no geo holdout, conversion lift, or matched market test results in this workspace. Without incrementality data, the marketing mix model must attribute all sales correlation to paid spend, which overstates the contribution of channels with high organic co-occurrence, particularly branded search and podcast sponsorship. This is not unusual at your stage of measurement maturity. It is worth planning one test before the next budget cycle.