CTV and BVOD trafficking gets treated like "video, but on a bigger screen" more often than it should. It isn't. A tag that validates fine in a browser-based video player can still fail silently once it's inside a Roku channel or a broadcaster's app, frequency capping behaves differently when a household shares one screen across five people, and a tracking pixel that would fire normally on desktop can get dropped entirely depending on whether the stream uses server-side or client-side ad insertion. None of this shows up until the campaign is live, unless you check it beforehand.
The short answer: a real pre-launch CTV/BVOD QA pass validates the VAST tag (not VPAID — most CTV apps can't execute it), confirms whether frequency capping is being applied at the household or device level and whether that matches the buy's intent, checks creative duration, resolution, and caption requirements against the specific platform's spec sheet, verifies ad pod position and competitive separation, and confirms whether the publisher uses server-side (SSAI) or client-side (CSAI) ad insertion — because that single detail determines which tracking pixels will actually fire.
1Validate the VAST Tag — and Confirm VPAID Isn't in the Mix
Start every CTV trafficking QA pass by pulling the actual VAST response, not just trusting that the tag "looks right" in the ad server UI. Check that the tag resolves against VAST 4.x if that's what the publisher requires, and count wrapper redirect hops — most CTV publishers cap acceptable wrapper depth well below what's tolerated on open web video, because underpowered set-top boxes and smart TV apps time out on long redirect chains far more often than a desktop browser does. Confirm every tracking event you expect is actually present in the VAST XML: impression, start, firstQuartile, midpoint, thirdQuartile, complete, and any skip or error events the platform requires.
The bigger issue is VPAID. Most CTV apps — Roku, Fire TV, Samsung Tizen, LG webOS, and the native apps for services like Hulu and Peacock — don't run a JavaScript execution environment the way a browser does, so VPAID creatives either get rejected outright at trafficking or fail at render time on-device. If a line item is still pointing at a VPAID-wrapped tag because it was cloned from a desktop video campaign, that's the single most common reason a CTV campaign shows healthy request volume but near-zero fill. Confirm every creative and tag in the chain is VAST-only before it goes near a CTV line item.
2Frequency Capping: Household vs. Device-Level
CTV frequency capping has a wrinkle that doesn't exist in mobile or desktop: a single screen is often watched by an entire household, not one identified user. Many CTV identity graphs cap at the household level — grouped by shared IP, ISP account, or a household-level graph the SSAI vendor or publisher maintains — specifically to avoid hammering one TV with the same ad regardless of who's on the couch. Device-level capping, by contrast, relies on an identifier like a device's advertising ID (where the platform exposes one, such as Roku's RIDA or Amazon's Fire TV advertising ID), which lets you distinguish between two different devices in the same home but says nothing about how many distinct people are watching either one.
Before you traffic the line item, confirm which level the publisher or DSP is actually capping at, and whether that matches the buy's intent — a reach-building campaign generally wants household-level capping to avoid burning frequency on the same living room, while a campaign trying to manage exposure per identified viewer needs device-level capping and needs to know that identifier is actually available on that platform. Also flag a structural limitation up front with anyone setting frequency goals: most streaming apps are walled gardens that don't share frequency data with each other, so a cap set inside one DSP or one publisher's direct deal has no visibility into how many times the same household saw the ad through a completely separate app or demand path. True cross-app, cross-device frequency management generally requires an identity or clean-room solution layered on top — it isn't something the ad server does for you by default.
3Creative Duration, Format, and Caption Specs Per Platform
CTV creative specs vary more by publisher than most traffickers expect coming from display or standard pre-roll. Duration is usually locked to :15 or :30, occasionally with a shorter :06 bumper format supported on select platforms, and non-skippable is the default assumption for CTV inventory rather than the exception — build and QA creative accordingly rather than assuming a skip button will be available. Resolution requirements commonly start at 1080p, with a growing number of premium CTV publishers requiring or preferring 4K source files even if the ad ultimately transcodes down, so check the current spec sheet rather than reusing last year's requirements.
Closed captioning is the item most frequently missed on the first pass. A lot of BVOD and broadcaster-owned CTV inventory — think Hulu, Peacock, and broadcaster catch-up services — either requires embedded or available captions as a platform policy or as a regulatory obligation tied to broadcast content standards in their market, and creative without them can get flagged or rejected during the publisher's own pre-flight review, not just at your QA stage. Audio loudness is worth a spot-check too; some networks apply broadcast-style loudness normalization and will reject creative that's mixed noticeably hotter or quieter than the surrounding content. None of this is exotic, but it's easy to skip when a trafficker is moving fast and treating the creative asset as "just another video file" rather than routing it through proper creative operations review against the specific platform's spec sheet.
4Ad Pod Positioning and Competitive Separation
Pod position matters more in CTV than most planning decks account for. Pure pre-roll is still the norm on many apps, but long-form BVOD content increasingly runs mid-roll pods with multiple positions, and position 1 in a pod behaves differently from position 2 or 3 — completion rates, pricing, and even the platform's own reporting sometimes treat them as distinct inventory. When you're trafficking against a pod, confirm which position the line item is eligible for rather than assuming "in the pod" is specific enough.
Competitive separation is the other piece worth confirming before launch, not after a client notices two competing auto brands back-to-back in the same pod. This is typically handled through IAB category exclusions or the ad server's own competitive exclusion rules, and it's worth double-checking those exclusions actually apply to the pod-fill logic on the specific platform, since some publishers manage pod composition through a separate ad-pod optimization layer that doesn't automatically inherit line-item-level exclusions the way a standard display ad server would. It's also worth pacing-checking pod fill directly — if two of your own line items are eligible for the same pod slot, they can end up competing with each other and quietly depressing delivery on both. This is exactly the kind of pre-launch detail that's easy to miss under a tight go-live deadline, and it's where dedicated trafficking support earns its keep — if your team is finding these issues after launch instead of before, a free ad-stack audit is a faster way to find the gap than adding another line to a checklist nobody has time to run.
5SSAI vs. Client-Side Ad Insertion: Tracking Implications
This is the single most consequential technical question to answer before you traffic a CTV campaign, and it's the one most frequently skipped. With server-side ad insertion (SSAI), the ad is stitched directly into the video stream on the server before it ever reaches the device — which means there's no client-side browser or app JS environment for some tracking pixels to execute in. Verification and viewability tags built around client-side JavaScript execution may simply never fire under SSAI; instead, tracking typically relies on server-to-server pings fired by the stitching platform against the VAST tracking URLs, and it's essential to confirm with the publisher exactly which tracking events their SSAI setup actually supports before you assume standard VAST quartile tracking will report cleanly.
Client-side ad insertion (CSAI), where the app's own player SDK requests and inserts the ad directly, behaves closer to standard web video tracking — the player fires VAST events itself, and most tags work as expected. The practical trafficking implication: don't assume a tag validated against a CSAI test environment will report identically once it's running on an SSAI-based platform, and don't be surprised by lower reported completion or viewability numbers under SSAI if the publisher's stitcher doesn't support every pixel your verification vendor wants to fire. It's also worth watching for stitching failures that fall back to a black slate rather than the intended creative — a manifest-level issue that looks like normal delivery in server logs but shows a blank screen on the actual device. This is one of the places technical QA support pays for itself, because it requires someone actually watching the stream on a real device, not just reading the ad server's delivery report.
6Common CTV Delivery Issues to Catch Before Launch
A short list of issues accounts for most CTV launch-day fire drills. First, non-skippable enforcement that the platform's spec sheet promises but doesn't actually apply consistently — verify it with a real test play rather than trusting the documentation. Second, tag timeouts: CTV devices and networks are often slower than the desktop or mobile environment your default ad server timeout settings were tuned for, and an unadjusted timeout threshold can produce artificially high unfilled rates that look like a demand problem but are actually a latency problem. Third, creative rejected for aspect ratio, bitrate, or codec mismatches that only surface once it's rendering on an actual device — always request a device-rendered preview from the publisher rather than relying on a spec sheet confirmation alone.
Fourth, frequency caps applied at the wrong level for the campaign's goal, as covered above — this is worth a final gut-check right before launch, not just during initial setup. And fifth, missing closed captions triggering a rejection at the publisher's own pre-flight stage after your QA already signed off, which usually means caption verification needs to move earlier in the ad quality and compliance workflow rather than being treated as a formality. None of these are hard problems once you know to look for them — they're just easy to miss when CTV trafficking gets slotted into the same checklist built for standard digital video. For more on how the broader trafficking discipline should adapt across formats, our piece on ad trafficking services covers the fundamentals this checklist builds on.
