ROAS Is the Wrong Polynomial

@Faherty Brand is expanding retail while running a thriving DTC channel. Smart strategy. Painful measurement.

The reflexive graph here has a specific structure: DTC team and retail team both optimize for their own ROAS. The board sees two ROAS numbers and assumes both are real. But the polynomial each team is solving has different variables. DTC ROAS credits online revenue. Retail ROAS credits in-store revenue. Neither captures cross-channel influence.

The invariant: "If my channel ROAS is above 4:1, I am winning." That closure prevents anyone from seeing the real decision equation — which is portfolio ROAS net of cannibalization.

When I took over omnichannel strategy for a $300M retailer, 34% of DTC ad spend reached customers who had already purchased in-store. We were paying premium CPMs for sales that would have happened anyway. Channel ROAS looked strong. Portfolio ROAS was bleeding.

The Portfolio ROAS Framework:

1. Subtract cannibalization from each channel's ROAS before evaluating. If DTC ads reach existing retail customers, that is not acquisition — it is expensive retention.
2. Build a unified attribution model. My team makes decisions on one dashboard, not three competing ones.
3. Measure incrementality by channel, not efficiency. We lifted total ROAS 67% by killing efficient but cannibalizing campaigns.
4. Coordinate channel budgets against the portfolio polynomial, not individual channel equations.
5. Assign channel roles: DTC for discovery, retail for consideration, wholesale for volume. Stop competing with yourself.

The S4 construction here: build a unified commercial engine where channels are nodes in one graph, not independent operators.

I have enabled a £790M IPO. Channel unification was how we got there.

Are your channels allied or competing? What does portfolio ROAS look like when you subtract cannibalization?