How to Measure Amazon Sponsored TV (Beyond ROAS)
The number the console shows you for Sponsored TV is the wrong one. Read it and you’ll kill a campaign that’s quietly growing your brand. Here’s what to watch instead.
Don’t judge Sponsored TV on console ROAS — it’s last-touch and upper-funnel, so it will always look weak. Measure lift instead: brand-search lift (+15–40%), new-to-brand rate (60–80%), detail-page-view lift (+20–50%), and the Amazon Marketing Cloud conversion-lift report. Watch these signals arrive in order over about 30 days.
01Why console ROAS misleads
The Ads Console reports Sponsored TV on last-touch, on-Amazon ROAS — it credits a sale only if a TV impression was the final touch before purchase. But almost nobody buys straight off a living-room TV ad; they see it, and days later they search, compare, and buy through another touchpoint. So the console systematically under-credits TV, and a genuinely effective campaign shows a weak ROAS. Trusting that number is the single most common reason sellers cut Sponsored TV campaigns that were actually working.
02Two layers: delivery vs impact
Separate your metrics into two questions. Delivery metrics — impressions, reach, frequency, video completion rate, CPM — answer “did the ad get seen properly?” They’re necessary hygiene, but a high completion rate doesn’t prove the campaign moved anything. Impact metrics — the lift measures below — answer “did it grow the business?” Don’t confuse the two: plenty of Sponsored TV campaigns deliver beautifully and change nothing, and you can only tell them apart by looking at lift.
03The four lift metrics that matter
These are the numbers that tell you Sponsored TV worked:
| Metric | What good looks like | When it shows | What it tells you |
|---|---|---|---|
| Brand search lift | +15–40% vs prior 4 weeks | ~Week 2 | People are looking you up |
| Detail page view lift | +20–50% in window | Weeks 2–3 | It’s driving consideration |
| New-to-brand rate | 60–80% on promoted ASINs | Weeks 3–4 | You’re reaching new customers |
| AMC conversion lift | Positive incremental lift | Full flight | The defensible truth |
One diagnostic to note: if your new-to-brand rate on promoted ASINs falls below ~50%, your targeting is too narrow — you’re re-reaching existing buyers with expensive TV instead of finding new ones.
04The order signals arrive
Lift shows up in sequence, not all at once, so reading it too early misleads you. Brand-search lift comes first (around week 2) — curiosity is the fastest response. Detail-page-view lift follows (weeks 2–3) as that curiosity turns into research. New-to-brand orders catch up last (weeks 3–4) once shoppers move through consideration to purchase. This is why a clean read needs at least a 30-day flight, and why a week-one verdict — on any metric — will point you the wrong way.
05Frequency & VCR as diagnostics
When lift is missing, two delivery metrics tell you why. Low frequency — fewer than about two impressions per viewer per month — means your budget is spread too thin for anyone to remember the ad; that’s a budget or audience-size problem, not a creative one. Low video completion rate — viewers dropping off before the end — points to a creative problem: a weak hook or a message that doesn’t hold. Use these to diagnose a flat campaign before you conclude TV “doesn’t work” for you.
06AMC: the most defensible read
The gold standard is the Amazon Marketing Cloud conversion-lift report, which compares an exposed group against a holdout to isolate the incremental conversions TV actually caused — the closest you’ll get to true causality on any CTV product. AMC also enables custom cross-media attribution and lets you follow the customer journey over roughly 12.5 months, so you can see how a TV impression contributed to a purchase that closed weeks later through search. If you run Sponsored TV at any real scale, AMC is where the honest answer lives.
07Measure the journey, not the island
Sponsored TV rarely closes the sale by itself — it opens the door that your other ads walk the shopper through. Amazon reports that customer journeys touched by streaming TV plus sponsored ads convert new-to-brand meaningfully better than journeys without TV. So measure the whole path: is TV lifting the branded searches your Sponsored Products then capture? Are more viewers entering your retargeting pool? Judging TV in isolation misses most of its value.
08A measurement setup that works
- Baseline branded search for the 4 weeks before launch
- Track branded keyword volume in Brand Analytics
- Set up branded SP campaigns to catch the lift
- Monitor DPV and NTB by promoted ASIN
- Run 30+ days before drawing conclusions
- Use the AMC conversion-lift report for the final read
- Console ROAS is last-touch and understates upper-funnel TV — don’t judge Sponsored TV by it.
- Separate delivery metrics (did it get seen?) from impact metrics (did it grow the business?).
- Track four lift metrics: brand-search, detail-page-view, new-to-brand, and AMC conversion lift.
- Signals arrive in order over ~30 days — search first, then DPV, then NTB orders.
- Measure the whole journey; TV opens the door your search and retargeting ads walk through.
Frequently asked questions
How do I measure Amazon Sponsored TV?
Measure lift, not last-touch ROAS. Track brand-search lift, detail-page-view lift, new-to-brand rate on promoted ASINs, and the Amazon Marketing Cloud conversion-lift report, over a flight of at least 30 days. These reveal the upper-funnel impact the console understates.
Why does Sponsored TV show a low ROAS?
Because the console measures last-touch, on-Amazon conversions, and TV is upper-funnel — viewers rarely buy immediately off a TV ad. The purchase usually closes later through another touchpoint, so TV’s contribution is under-credited and its ROAS looks weak even when it’s working.
What is brand search lift?
It’s the increase in how often shoppers search your brand name during a campaign versus a baseline period — typically the prior four weeks. A properly-running Sponsored TV campaign often lifts branded search 15–40%, and it’s the earliest signal that the ad is working.
What is a good new-to-brand rate for Sponsored TV?
Roughly 60–80% of orders on promoted ASINs coming from new customers is a strong result, since acquisition is the point of TV. If the rate drops below about 50%, your targeting is too narrow and you’re paying to re-reach existing buyers.
What is AMC conversion lift?
It’s an Amazon Marketing Cloud report that compares an exposed audience against a holdout to isolate the incremental conversions your ads actually caused. For Sponsored TV, it’s the most defensible attribution available, cutting through the last-touch bias of console reporting.
Or return to Module 6: Sponsored TV or the course hub.