One of the most powerful outputs of Marketing Mix Modeling is the response curve — a chart that shows how revenue changes as you increase spend on a channel.
If you've never read one before, they can look intimidating. But once you understand the shape, response curves become the clearest way to answer the most important question in marketing: where should I spend my next dollar?
What is a response curve?
A response curve plots channel spend (X-axis) against incremental revenue (Y-axis).
The curve is almost always concave — it rises quickly at first, then flattens out. This is called diminishing returns. The first $10k you spend on Google Ads generates more revenue per dollar than the 10th $10k, which generates more than the 50th.
Reading the curve: three zones
Zone 1: Underinvested (steep slope)
The left side of the curve is steep. Every additional dollar of spend drives significant revenue lift. ROAS is high. You're likely underspending here — there's cheap growth left on the table.
Action: Increase budget until the curve starts to flatten.
Zone 2: Efficient frontier (moderate slope)
The middle of the curve is the sweet spot. You're spending enough to capture most of the available demand, but not so much that you're chasing diminishing returns. ROAS is still positive, but declining.
Action: This is where most channels should operate. Monitor closely — if competitors increase spend or demand shifts, you may need to adjust.
Zone 3: Saturated (flat slope)
The right side of the curve is nearly flat. Additional spend barely moves revenue. ROAS is approaching 1:1 or worse. You're past the point of efficient return.
Action: Cut spend and reallocate to underinvested channels.
What to do with this information
Response curves tell you two things:
- Which channels are saturated. If your Meta spend is deep in Zone 3, cut it by 20% and watch revenue barely move. Reallocate that budget to an underinvested channel like podcasts or CTV.
- How much headroom each channel has. If Google Ads is in Zone 1, you can confidently increase spend knowing each dollar will drive meaningful lift.
A real example
One of our D2C clients was spending $400k/month on Meta and $50k/month on Google. Their response curves showed Meta was deep in saturation (ROAS 1.8×) while Google was in Zone 1 (ROAS 5.2×).
We shifted $100k from Meta to Google. Total revenue increased by $180k the next month — because the marginal dollar on Google was 3× more efficient than the marginal dollar on Meta.
That's the power of response curves. They make reallocation decisions obvious.
The fine print
Response curves assume all else is equal. In reality, your creative quality matters, your audience targeting matters, and external factors (seasonality, competitors) shift the curves over time.
Run your MMM monthly. Update the curves quarterly. Test reallocations with small experiments before committing big budget shifts.
But once you have response curves, you'll never look at a channel in isolation again.