Which Buying Lane, Modeling a $15K/Month CTV Budget Across Self-Serve, DSP, and FAST
Objective: Given a fixed monthly CTV budget and a reach/frequency goal, forecast realistic impression volume and blended CPM across the three buying lanes (self-serve, programmatic DSP, FAST channels) to recommend a split.
You're the growth marketer at a ThredUp-style online resale marketplace with a new $15,000/month test budget to move off saturated Meta prospecting and into CTV for the first time.
Model expected impressions and blended CPM for three allocation scenarios, then recommend a split that balances reach against premium-streamer brand lift.
Before you start
What you'll need
Free path (everything below is enough to finish)
No CTV spend required to test the allocation math before committing budget
Paid upgrades (optional, faster/deeper)
The lesson names YouTube via Google Ads as a real CTV buying lane, and it's the one cataloged tool that actually reaches living-room TV screens
The process
2 steps
Step 01 of 02
The lesson splits CTV buying into three lanes by budget size: self-serve managed platforms ($1K-$250K/month), programmatic DSPs (cross-platform unified buys), and FAST channels ($12-20 CPM vs. $25-65 on premium streamers).
At $15,000/month, is this budget better spent entirely on a premium self-serve platform like Roku Ads Manager, split across FAST channels, or blended?
Procedure
- Row 1: $15,000 at a $35 blended premium CPM (Netflix/Hulu range) = ~428,500 impressions
- Row 2: $15,000 at a $16 blended FAST CPM (Tubi/Pluto/Roku Channel range) = ~937,500 impressions
- Row 3: 60% ($9,000) to FAST at $16 CPM + 40% ($6,000) to premium at $35 CPM = 562,500 + 171,400 = ~733,900 impressions
- Compare completion rates: 88-94% on FAST vs. ~95% on premium, both well above social video
Scenario Impressions Blended CPM Completion Rate Premium-only 428,500 $35.00 ~95% FAST-only 937,500 $16.00 88-94% 60/40 Blend 733,900 $20.44 90-94%
Healthy
The blended scenario delivers 71% more reach than premium-only at a completion rate within a few points of it, using FAST volume to fund a smaller premium presence for brand-safe placements.
Unhealthy
Spending the entire $15,000 on premium streamers because they "sound better," delivering under half the reach for a completion-rate difference of only a few points.
What this means
At this budget size, reach efficiency from FAST channels usually outweighs the marginal completion-rate edge of premium streamers, the lesson's own guidance for advertisers under $20,000/month.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| Budget under $20K/month is being allocated 100% to premium streamers | Shift the majority to FAST channels and use premium only for high-value creative moments | 30 min |
Step 02 of 02
The lesson warns that a 3-per-week cap set separately on 3 different DSPs becomes a 9-per-week cap for any household reachable across all three, unless the cap is unified.
The blended plan above runs across 2 self-serve platforms (Roku Ads Manager for FAST, MNTN for premium) with no shared frequency cap set yet. What's the risk and the fix?
Procedure
- Set a 3-per-week cap inside Roku Ads Manager
- Set a separate 3-per-week cap inside MNTN
- Recognize any household reachable on both platforms can see up to 6 ads/week instead of the intended 3
- Either lower each platform's individual cap to 1-2/week, or move to a single DSP that can enforce one cap across both lanes
Household overlap check: 34% of the FAST-reachable audience is also reachable on MNTN's premium inventory -> uncapped combined frequency could hit 6/week for over a third of the target audience.
Healthy
Each platform's cap is set to roughly half the target (1-2/week) to keep the combined ceiling near the intended 3/week for overlapping households.
Unhealthy
Both platforms run their own full 3/week cap independently, quietly doubling frequency for the highest-overlap third of the audience.
What this means
A frequency cap set at the platform level, not the household level, is not really a cap once you're buying across more than one platform.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| Two or more CTV platforms are live with the same nominal frequency cap | Halve each platform's individual cap or consolidate into a single DSP that unifies it | 30 min |
Final deliverable
A 3-scenario impression/CPM forecast model with a recommended budget split and a frequency-cap plan across platforms.
See a reference example
Grab CTV test budget model (excerpt) Premium-only ($20K): 571,400 impressions at $35 CPM FAST-only ($20K): 1,250,000 impressions at $16 CPM Recommended 65/35 blend: 1,043,600 impressions, blended CPM $19.16 Frequency cap: Roku Channel capped at 2/week, premium DSP capped at 1/week, combined ceiling ~3/week for overlapping households.
Success criteria
You're done when you can:
- Correctly calculates impression volume from budget divided by CPM for all three scenarios
- Identifies the frequency-cap overlap risk and proposes a specific fix