Your first CTV campaign: the complete checklist
A first Connected TV campaign fails in predictable places: no written success criterion, mobile creative pasted onto a TV, measurement configured after launch. This checklist walks the whole setup in order — goal, audience, inventory, creative, measurement, launch — so your first flight produces learning you can act on, whatever the numbers say.
Step 1: Write down the goal and the kill criterion
Before anything else, one sentence: "This campaign succeeds if X, and we stop it if Y." X might be a target cost per incremental install, a scan rate per completed view, or measurable lift in branded search within the test region. Y is the number that makes you stop and rethink rather than keep spending.
First campaigns are learning purchases. A pilot that hits its learning goal and misses its CPI target is a success; a pilot with no written criterion is an opinion generator at the next budget meeting.
Step 2: Pick one audience and one geography
The beginner mistake is testing the channel, three audiences and five geographies at once — producing data too thin to read. Choose the audience you understand best from your existing channels and the geography where your product already converts. You are isolating one variable: the channel.
CTV targeting is household-level: geography, household composition, content context, and first-party list matching. If you have a customer list, a household-matched audience is often the strongest first test — you are re-reaching people in a new context, which reads cleanly in measurement.
Step 3: Choose inventory you can verify
Decide where your ad may run before you decide anything else. The safe starter set: inventory with declared supply paths (app-ads.txt, sellers.json), named apps and channels rather than anonymous bundles, and placement-level reporting so you can see exactly what ran where.
Whether you buy through a DSP, a network or a partner like us, ask the five seller questions from our buying guide — what inventory, what supply path, what verification, what reporting, what happens when verification flags a placement. Concrete answers are the difference between a test and a donation.
Step 4: Adapt creative for the big screen
A mobile video pasted onto a TV fails in three places: text too small to read from a couch, audio treated as optional, and a call to action built for a tap that cannot happen. The adaptation rules: large high-contrast text; audio that carries the core message because sound is on; and a response mechanic that works from a couch — a QR code with a stated reason to scan, plus a simple verbal call to action for everyone who will not scan.
Keep the spot to 15 or 30 seconds, put the brand in the first five seconds, and test the final cut on an actual television from an actual couch. Ten minutes of that test catches problems no review on a monitor will.
Step 5: Set up measurement before launch
CTV has no clicks, so measurement must be designed, not defaulted. The minimum setup: your MMP receiving postbacks or log-level exposure data; a smart link behind the QR code carrying campaign parameters; and a defined comparison — a holdout geography or unexposed household group — so you can read incrementality rather than coincidence.
Verify the full chain with a test scan on both platforms before spending a cent: QR to store page to install to MMP event. A campaign with unverified measurement is a branding campaign whether you meant it or not.
Step 6: Launch small, watch daily, prune weekly
Launch with a bounded budget and frequency caps — two to three exposures per household per week is a sane starting band. Watch delivery, completion and placement mix daily for the first week: which apps are serving, at what times, with what completion. Prune placements that deliver impressions without downstream signal; concentrate on what works.
Read results against the criterion from step one, on the schedule you set — not daily. CTV effects arrive partly through cross-device and branded-search paths that need weeks, not days, to show.
Step 7: The post-flight review
When the flight ends, answer four questions in writing: did we hit the success criterion; which inventory, audiences and creatives drove the outcome; what did the incrementality read say versus last-touch; and what is the next test. That document — one page is enough — is what you actually bought with the pilot budget.
We run first CTV campaigns for app advertisers with exactly this structure: bounded budget, verified inventory, big-screen creative adaptation, measurement set before launch, and a written read at the end. Register as an advertiser to scope yours.