Does vertical video lift YouTube Shorts ads? Google's 2024 figures

Yes, according to Google's own help page, but the page quotes three different gains for three different campaign types, and it labels them as Google's 2024 internal data. Adding vertical video to Demand Gen campaigns on Shorts is tied to a gain of more than 35% in conversions. Adding vertical to Video View campaigns is tied to more than 40% higher view-through rate (VTR) on Shorts. Adding vertical to Video Reach campaigns is tied to more than 45% higher viewability on Shorts and 20% longer average watch time.

Those are Google's numbers, not ours, and nobody outside Google can reproduce them from the page. Treat them as a reason to ship a 9:16 cut, not as a forecast for your account. This post lays out which figure belongs to which campaign, what else the same page says about aspect ratios, and how to render the vertical version with the Sume Timeline API.

The three figures, side by side

The page is Google Ads Help, "Use square and vertical video to engage mobile customers". It states that vertical videos with a 9:16 aspect ratio "are best suited for the Shorts format and deliver better performance compared to landscape assets". The three figures sit underneath that sentence, each tied to one campaign type.

Read the wording of each metric before you compare them. Conversions, VTR and viewability measure different things, so a 45% viewability gain does not mean a 45% conversion gain. And "increases" is relative to the same campaign without a vertical asset, per Google's statement, not relative to a competitor.

What the same page says you should not do

The page also contains a warning that gets skipped in summaries: campaigns relying only on vertical content may not serve, because vertical assets no longer serve in YouTube in-feed environments. In other words, the vertical cut is an addition, not a replacement.

It also says square video is still recommended for App, Demand Gen and Performance Max campaigns, and that aspect ratios between 16:9, 9:16 and 1:1 are allowed. So the practical set for most advertisers is a 9:16 master plus a 16:9 or 1:1 companion, not one file.

• Ship 9:16 for Shorts delivery, per the page.
• Keep a 16:9 or 1:1 asset so in-feed placements still have something to serve.
• Do not read the three percentages as one combined lift; they apply to three campaign types.

Rendering the vertical cut with Sume Timeline

Sume Timeline 1.0 returns one MP4 from one audio spine plus ordered video slots, and its default output is 1080x1920, which is exactly the 9:16 recommendation. The public rate in the Timeline docs is $0.10 per ceil output minute, so a 15-second vertical ad bills one minute: $0.10. A free plan call (POST /v1/timeline-1.0/plan) returns duration_seconds, segment_count, billable_minutes and estimated_cost_usd_micros without creating a job.

Every source URL must already be a media.sume.com artifact or asset of your workspace, so import the clip first. The render needs an Idempotency-Key, and the default mode is async: you poll GET /v1/jobs/:id/status and then /result, which returns video_url.

curl -X POST https://api.sume.com/v1/timeline-1.0/render \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: vertical-ad-001" \
  -d '{
    "audio": {
      "url": "https://media.sume.com/artifacts/artf_demo/voice.wav",
      "duration_seconds": 15
    },
    "video": [
      {
        "source_url": "https://media.sume.com/artifacts/artf_demo/hero.mp4",
        "start": 0,
        "duration": 15,
        "fit": "cover"
      }
    ]
  }'

A sensible test plan

If you want your own evidence instead of Google's, run the vertical and horizontal cuts in the same campaign for the same period and compare the metric that your campaign type optimises. Render both from one spine so the audio is identical and only the framing differs: a second Timeline call with output.width 1920 and output.height 1080 costs another $0.10 for a clip of one minute or less.

Do not generalise from a single account. Google's page gives no sample sizes, so the figures tell you the direction Google sees across its own data, and nothing more precise than that.