Blendpipe homeSend a brief
Menu

Cutting down a 30-second spot: 15s, 6-second bumpers and vertical versions

6 second bumper ads and 15s cutdowns rarely work as plain trims of a 30. Google's length bands, what to plan on set, vertical crops and when to shoot native.

Reader question
Can a 30-second hero spot be cut down into 15-second, 6-second and vertical versions, or should the cutdowns be planned and shot separately?
A small bonsai juniper on a wooden potting bench beside a pair of pruning shears, with one branch drawn in chroma green with a bold black outline growing into the gap left by a cut.

Yes, a 30-second hero spot can yield 15-second, 6-second and vertical versions, but only reliably when those versions are planned before the shoot rather than discovered in the edit. Each length needs its own structure, its own coverage and its own end card, and a centered 9:16 crop keeps only about 32% of a 16:9 frame, so the framing has to be decided on set. Google’s own bumper guidance puts it plainly: fitting into six seconds “requires more than just cutting down”. Plan the cutdowns in the script and shot list, cover them on the day, and shoot native vertical only when vertical placements carry the campaign or the hero frame cannot survive the crop.

This guide starts with the lengths the platforms enforce, as published on their help pages in November 2026. It then covers how each length is structured, what to plan on set, how the vertical and 4:5 crops behave, and how to decide between cutting down and shooting native. It ends with a deliverable matrix built from the fictional example campaign used across this site.

What lengths do the platforms actually run?

The platforms sort video ads by duration, and the length bands decide which format a file runs as. On Google Ads, a video of up to 6 seconds runs as a bumper ad, 7 to 15 seconds as a standard non-skippable ad, and 16 to 30 seconds as a 30-second non-skippable ad shown on connected TV, which must be horizontal (Google Ads Help, “Create a Video reach campaign”). Google Ads “automatically determines the video ad format based on the length of your video”, so the length you cut is the format you buy.

Platform and format Length rule (as published) Orientation and other notes
Google Ads, bumper Up to 6 seconds (5 seconds minimum on the formats page) Non-skippable
Google Ads, standard non-skippable 7 to 15 seconds in Video reach campaigns The formats page lists non-skippable in-stream at 15 to 60 seconds depending on campaign subtype
Google Ads, 30-second non-skippable (CTV) 16 to 30 seconds “Video assets must be horizontal. Square or vertical assets will not be able to run.”
Google Ads, skippable in-stream No maximum (under 3 minutes recommended) Viewers can skip after 5 seconds
YouTube Shorts ads Under 60 seconds recommended Viewers can swipe away at once; a TrueView view counts at 10 seconds or the end
LinkedIn video ads 3 seconds to 30 minutes; 15 to 30 seconds recommended Videos under 30 seconds loop until there are at least 30 seconds of playback
TikTok In-Feed ads (non-Spark) Up to 10 minutes 9:16 recommended
Meta Facebook Feed 1 second to 241 minutes 4:5 at 1440 x 1800
CTV publishers Exact lengths, often 15 or 30 seconds See our guide to CTV ad specs

Sources for the rows: Google Ads Help “About video ad formats” and “Create a Video reach campaign”; LinkedIn’s video ads specifications; TikTok’s In-Feed ads specifications, updated June 2026; and Meta’s Facebook Feed video specifications. All were checked in November 2026.

Two things stand out. First, the only hard caps that bite on a short campaign are Google’s: 6 seconds for a bumper and 15 for a standard non-skippable. Social feeds take almost any length, so their limits are attention and placement fit, not rules. Second, Google’s non-skippable mix lets you combine lengths: it says that “when creating a non-skippable mix, you can choose to add videos with different durations of 6 seconds, 15 seconds, and 30 seconds”. A 30, a 15 and a 6 from one shoot are exactly what that mix expects.

Frame-exact bumpers

A bumper has to be no longer than six seconds of running time, and that is not the same as 144 frames at every frame rate. At 24 fps, 144 frames run exactly 6.000 seconds. At 23.976, which many productions use, 144 frames run 144 / 23.976 = 6.006 seconds. Since Google Ads assigns the format by duration, a conservative edit at 23.976 ends on frame 143, which runs 5.964 seconds. The same arithmetic applies to the 15: at 23.976, 360 frames run 15.015 seconds. The frame and timecode calculator converts between frames and seconds at each rate, and the safest check is to read the duration the platform reports after upload.

Why is a cutdown a new edit, not a trim?

A cutdown is a new edit because a shorter length changes what the spot can do, not only how long it runs. A 30 has time for a setup, a turn, a product moment and an end card; a 15 has time for one idea and the product; a 6 has time for one thought and the brand. Trimming the 30 tends to keep its setup and lose its payoff, or keep the payoff without the context that made it land.

Google’s bumper guidance, written by its creative director Ben Jones in April 2017 (Think with Google, “Success in Six”), is old but still the clearest statement of the craft. Its four points:

  1. Focus on a single, simple purpose. “When you’re working in :06 though, there isn’t time for everything. So rather than trying to do a little of everything, do one thing well.”
  2. Budget time to establish the ad, and stick the landing. Viewers need a moment to orient themselves, and the closing thought or call to action needs time to sink in: “Two seconds is a good length to aim for.”
  3. Your 30-second spots don’t hold all the answers. “Fitting into :06 requires more than just cutting down. Shot length, story arc, and other elements may not translate well.” It suggests retaining “only the aspects that will serve a powerful, memorable :06”, and notes that print or out-of-home visuals can be a better starting point than the 30.
  4. Build in series, rather than singles. It suggests “thinking of bumpers as a set of three to five ads”.

Read against the current length bands, the second point sets the budget for a bumper. If the landing takes about two seconds, the rest of the idea gets about four: at 24 fps, 2 seconds is 48 frames and the remaining 4 seconds are 96 frames. Four seconds hold one or two shots that read instantly, not the four or five that open a 30.

Words follow the same logic. At 150 words per minute, the pace used in the script timing calculator, a voice reads 2.5 words per second, so the rough budgets are 12 words for 6 seconds, 30 for 15 and 60 for 30, before any music-only moments. A 6 with a full end card and a beat of music usually carries far fewer than 12.

How should each length be structured?

Each length should be structured around what it alone can do, with the product and brand placed where viewers of that length will actually see them. The table below is a planning guide drawn from the principles above, not a platform rule; each campaign sets its own numbers in the plan.

Element 30 seconds 15 seconds 6 seconds
Job Tell the whole idea: setup, turn, proof, product, brand Make one claim and show the product Leave one thought and the brand
Shots As the story needs The proof shots and the product, few setups One or two shots that read at a glance
Voiceover budget at 150 wpm Up to 60 words, less any music-only time Up to 30 words Up to 12 words, often 6 to 8
Product on screen Can arrive late if the story earns it Early, then again at the end From the first second or the end card
End card Built for the 30 Built for the 15, not cropped from the 30 About 2 seconds, per Google’s guidance
Music The full cue An edit with its own ending A sting or the cue’s final phrase

Three structural habits separate cutdowns that work from ones that feel like fragments:

  • Front-load the brand in short versions. In a 6, a brand that appears only in the last second has little time to register. Put the product or logo in the opening frames or give the end card its full two seconds.
  • Give every length its own ending. A 15 that ends on the 30’s end card often runs out of time before the card reads. Design end cards per length, with type sized to the time on screen.
  • Keep the claim intact. A short version must not drop the qualifier that makes the long version’s claim accurate. If a claim needs a super, the super goes into every version that makes the claim, sized to be legible in that version’s time and frame.

Skippable, non-skippable and feeds: what the first seconds must do

The format also decides where attention is won. On a skippable in-stream ad, Google says viewers “can skip the ad after 5 seconds”, and with cost-per-view bidding the advertiser pays “when a viewer watches 30 seconds of your video, or the entire duration of the video if it’s shorter than 30 seconds”, or interacts with it. A skippable 30 therefore needs its brand and its reason to keep watching inside the first five seconds, which is the same discipline a 6 demands.

Non-skippable formats change the question. A bumper or a 15-second non-skippable ad will be seen to the end, so the job is clarity rather than retention: one thought, legible type, an end card that has time to read. Feeds behave differently again. YouTube Shorts ads can be swiped away immediately, and a TrueView view on Shorts counts when a viewer “watches 10 seconds or until the end of the video, whichever comes first”. LinkedIn advises showing “the most impactful content in the first 10 seconds” and loops videos under 30 seconds until there are at least 30 seconds of playback, so a 10-second edit plays three times and a 20-second edit twice.

The practical consequence: the opening of every version is the part that has to work hardest. When the boards mark coverage for the short versions, they should mark the openings first.

What should be planned on set so the cutdowns work?

The cutdowns work when the shoot covers them: shots marked on the boards for each length, framing that survives the tallest crop, and audio and music that can be cut to each ending. Most failed cutdowns are not edit problems. The footage simply does not contain a two-second shot that tells the story, or a vertical frame with the product in it.

Coverage for each length

  1. Mark the boards by length. For each shot, note whether it appears in the 30, the 15, the 6 or all three, so the director knows which takes carry the short versions.
  2. Shoot short, self-contained moments. A shot that needs three seconds to develop cannot open a 6. Ask for versions of key moments that read in under two seconds: the pour, the reveal, the product in hand.
  3. Hold heads and tails. Extra frames before and after each action give the editor room to retime for each length without jump cuts.
  4. Get the product early and clean. A clean product shot that works as the first frame of a 6 or a 15 saves a re-shoot or a 3D insert later.
  5. Cover the bumper series. If the plan calls for three to five bumpers, shoot the variations on the day: the same frame with different facets, as Google’s guidance suggests.

Framing for 9:16 and 4:5

  1. Frame for the tallest crop. Keep the subject, the product and any action that carries the idea inside the 9:16 area of the 16:9 frame, and mark that area on the monitor.
  2. Shoot at a higher resolution than delivery. A 3840 x 2160 source cropped to 9:16 keeps 1,215 x 2,160 pixels, enough for a 1,080 x 1,920 delivery; a 1920 x 1080 source cropped the same way keeps only 607 x 1,080 and has to be scaled up.
  3. Leave room for platform interfaces. Captions, buttons and profile overlays cover parts of vertical frames; TikTok says its safe zone depends on the dimension, the ad caption length and any extra formats, and it publishes downloadable safe zone files.

Sound, music and type

  1. Record voiceover per length. A 15-second read is a different performance from the first 15 seconds of a 30-second read. Record the short scripts in the same session.
  2. Choose music with edit points. A cue with clear phrase endings, or a library track with alternate endings and stems, can be cut to 30, 15 and 6. Confirm that the license allows edits and covers every length and version (the music licensing checklist lists the terms to check).
  3. Build supers and end cards per ratio. Type and graphics made in animation can be re-laid out for each ratio and length rather than cropped. Text burned into the live-action frame cannot.
  4. Measure every cut. Each length has its own loudness when measured over its duration, so a 15 trimmed from a compliant 30 still needs its own measurement; our article on loudness for TV and streaming ads explains the targets.

How do you cut the 15 and the 6 from the 30?

Cut the short versions from the hero’s material, but build each one from its own script outward: script first, then the shots that carry it, then the end card, then the music. Working from the 30’s timeline inward is how cutdowns end up as trims.

  1. Lock the hero first. Picture, mix and supers for the 30 are approved before the short versions start, so claims and type are settled.
  2. Time each short script. Read the 15 and the 6 scripts at the planned pace and add any music-only beats; the result sets the room left for picture.
  3. Pick the opening shot. Choose the frame that makes sense with no context, because in a 6 or a skippable placement nothing comes before it.
  4. Choose the proof shots. For the 15, keep the one or two shots that prove the claim. For the 6, keep one image that states the thought.
  5. Place the end card. Budget its time first (about two seconds for a bumper, per Google’s guidance) and cut the rest of the edit into what remains.
  6. Cut the music to each ending. Use an alternate ending or a final phrase rather than a fade, so each version lands on a musical button.
  7. Rebuild type, do not trim it. Re-time supers and kinetic type to the new length so every line stays on screen long enough to read.
  8. Check claims and supers. Every version that makes a claim carries the super that qualifies it, legible in that version’s frame and time.
  9. Count frames. Confirm each version’s frame count at the project rate against the platform’s length rule.
  10. Review each length on its own. Watch the 6 and the 15 as a viewer who has never seen the 30.

Example (fictional). Mapping the eight Orlenna shots to the shorter versions might look like this. The hero runs shot 1 (0 to 4 s, ridge at dawn), shot 2 (4 to 7 s, first sip), shot 3 (7 to 12 s, the drawn chart), shot 4 (12 to 17 s, the 3D wall), shot 5 (17 to 20 s, the blended shot), shot 6 (20 to 24 s, summit pour), shot 7 (24 to 27 s, kinetic type) and shot 8 (27 to 30 s, packshot).

  • 15 seconds: shot 1 for 2 s, shot 2 for 2 s, shot 4 for 3 s (under “two steel walls and a vacuum”), shot 6 for 3 s, shot 7 for 2 s and shot 8 for 3 s: 2 + 2 + 3 + 3 + 2 + 3 = 15 seconds, or 360 frames at 24 fps.
  • 6 seconds: shot 3 retimed to 2 s, shot 7 retimed to 2 s and shot 8 for 2 s: 2 + 2 + 2 = 6 seconds, or 144 frames at 24 fps, with the packshot holding 48 frames as the landing.

The blended shot 5 drops out of both short versions: it is the most expensive shot in the spot, but it needs the context of the 30 to read. That is a typical cutdown decision, and it is why the short versions are planned on the boards rather than found in the edit. The mapping is an illustration, not a production record.

Vertical and 4:5: what survives the crop?

A centered crop from a 16:9 frame keeps 56% of the width for 1:1, 45% for 4:5 and 32% for 9:16, so anything outside the middle third of the frame is lost in a vertical version. The aspect ratio crop calculator shows the numbers for a 3840 x 2160 source:

Ratio Crop from the source Share of the frame Delivered at
16:9 3,840 x 2,160 px 100% 1,920 x 1,080
1:1 2,160 x 2,160 px 56% 1,080 x 1,080
4:5 1,728 x 2,160 px 45% 1,080 x 1,350
9:16 1,215 x 2,160 px 32% 1,080 x 1,920
What each centered crop keeps of a 16:9 frameA 16:9 frame with three centered crop outlines drawn to scale: 1:1 keeps 56 percent of the width, 4:5 keeps 45 percent and 9:16 keeps 32 percent. A subject in the center sits inside all three crops. A product placed on the right third of the 16:9 frame falls outside every crop. product subject 16:9 frame, 100% 1:1, 56% 4:5, 45% 9:16, 32% A product on the right third is outside every centered crop.
Crops drawn to scale on a 16:9 frame; the subject and product are illustrative. / Source: Blendpipe aspect ratio crop calculator default example (3840 x 2160 source), checked November 2026.

The platforms point in different directions. TikTok recommends 9:16 for In-Feed ads. LinkedIn recommends 4:5 at 720 x 900 and also accepts 16:9, 1:1 and 9:16. Meta’s Facebook Feed specification lists 4:5 at 1440 x 1800. Google’s 30-second non-skippable CTV format accepts only horizontal assets. A campaign on all four needs at least three ratios, and each one is a separate composition, not a resize.

Composition rules that survive every ratio

  • Center the idea. If the action that carries the spot sits in the middle third, it survives every centered crop. Compositions that use the full width (two people talking across the frame, a product beside a landscape) need a separate setup or a reframe that moves the crop window shot by shot.
  • Move the crop, not only the frame. A reframe can pan the 9:16 window within each shot to follow the action. That costs edit time per shot, and it works only if the action stays inside the frame height.
  • Keep type off the plate. Supers composited over the picture can be re-laid out per ratio; type photographed on set or burned into the live-action plate cannot.
  • Check the safe zones per platform. Interface overlays differ by platform and placement, so test the vertical end card against each platform’s safe zone template before it is approved.

Automatic vertical versions

Google Ads offers to make the vertical version for you. Its video enhancements feature “uses Google AI to automatically create additional versions of your videos for you in vertical aspect ratio (9:16)”, and the extra versions “play automatically if they pass a quality review”. In Video reach campaigns it is “turned on by default for your new campaigns”, and you opt out by unticking “Get vertical versions of your videos”.

That is a decision for the plan, not a default to discover after launch. An automatic version decides its own crop, so check it against the same tests as a hand-made vertical: product visible, supers and legal lines intact and legible, end card readable. If the brand’s legal lines or claim supers must appear exactly as approved, a supplied 9:16 version gives more control than a generated one.

Should you cut down or shoot native?

Cut down when the hero frame was designed for the crops and the short versions support the hero; shoot native when the vertical or short placements carry most of the media plan, or when the hero composition cannot survive the crop. The table sets out the signals.

SignalFavors cutting downFavors shooting native
Where the media money goesMostly 16:9 placements: TV, CTV, in-streamMostly vertical feeds and Shorts
Hero compositionAction in the middle third, framed for 9:16 on setWide compositions that depend on the full width
What the shots are made ofAnimation, 3D and composited type that can be re-laid out per ratioLive action with type, products or people at the frame edges
Story shapeOne idea that compresses to a single thoughtA narrative that needs its full length to make sense
Platform behaviorFormats that play inside other videos (bumpers, non-skippable)Feeds where content made for the format is the norm (TikTok, Reels, Shorts)
Schedule and budgetOne shoot day; versions built in postRoom for extra setups or a second unit on the same day
Talent and rightsUsage already covers every length and mediumNew lengths or media would need new or extended usage anyway
A planning aid, not a platform rule; most campaigns land in between, with a cut-down set plus a few native shots. / Basis: Google Ads Help length and orientation rules and Think with Google bumper guidance (April 2017), checked November 2026.

In practice the answer is often both: one shoot day that covers the hero in 16:9 with 9:16-safe framing, plus a handful of setups composed natively for vertical, shot back to back with the same lighting, talent and product. That costs setup time on the day rather than a second production, and it gives the editor native vertical moments to open the short versions.

Two cautions apply whichever way you go. First, rights travel with the footage: a new length or medium can fall outside the usage the performers or the music license agreed to, so check the rights before the matrix grows. Second, a native vertical shoot is still part of one campaign, so its claims, supers and end cards follow the same approvals as the hero.

Building the deliverable matrix: lengths by ratios by calls to action

The deliverable matrix multiplies every dimension of the campaign, so each added length, ratio, language or call to action multiplies the files to edit, check and name. Count it before the shoot, because the count decides post time.

Example (fictional). Orlenna, an invented Colorado brand, launches the Ridge Bottle with a 30-second hero spot at 24 fps (8 shots). Its matrix is 3 lengths (30, 15 and 6 seconds) x 4 ratios (16:9, 9:16, 1:1 and 4:5) x 2 languages (English and Spanish) x 1 call to action = 24 deliverables. At 45 minutes of edit and 10 minutes of quality check per version, that is 24 x 45 = 1,080 minutes (18.0 hours) of edit and 24 x 10 = 240 minutes (4.0 hours) of QC, 22.0 hours in all.

The 6-second version keeps the drawn chart and the kinetic type from the hero, retimed to fit, and ends on the 3D packshot, which is identical in every version. Its fictional script, 8 words, reads in 3.2 seconds at 150 words per minute: “Still hot at the top. Orlenna Ridge Bottle.” The 15-second script, 23 words, reads in 9.2 seconds: “Coffee poured at the trailhead is still hot at the top. Two steel walls and a vacuum keep it there. Orlenna Ridge Bottle.” Both are drawn from the 42-word hero script, which reads in 16.8 seconds.

Because the hero was framed for the tallest crop, the live-action shots keep the runner and the bottle inside the 9:16 area (about 32% of the frame), and the drawn and 3D shots are re-laid out per ratio rather than cropped.

Suppose Orlenna then adds a second call to action. The matrix doubles to 3 x 4 x 2 x 2 = 48 deliverables, and post time to 48 x 45 = 2,160 minutes (36.0 hours) of edit plus 48 x 10 = 480 minutes (8.0 hours) of QC, 44.0 hours. Pruning cells no placement will buy (for example, a 30-second 9:16 if no vertical placement runs a 30) brings the count back down. All names, scripts and numbers in this box are illustrations.

The versioning matrix calculator runs this count with your own numbers, and its note is the rule to remember: cut the matrix to the placements you will actually buy before you add another dimension. The full fictional plan, with its blend map and shot list, is on the example campaign page.

Prune before you multiply

Not every cell in a matrix earns its file. Prune with three questions:

  1. Does any placement run this length in this ratio? Google’s 30-second CTV format is horizontal only, and many feed placements favor shorter edits, so a 30-second vertical may have no home.
  2. Does any audience see this language in this placement? A Spanish 4:5 may only matter where the plan buys Spanish-language feeds.
  3. Does this call to action apply here? An offer that runs online only should not generate broadcast versions.

Bumpers in series

If the plan follows Google’s suggestion of three to five bumpers, the series is a dimension of its own: three bumpers in four ratios and two languages are 3 x 4 x 2 = 24 more files, as many as the whole Orlenna matrix. Bumpers often run in fewer ratios than the hero, so decide the series’ ratios separately. Plan the series on the boards with the hero: if all three bumpers share one frame (the same bench, the same light, the same bottle), the variations can be shot in a single setup at the end of the day, and the series reads as a set when it runs. For a product brand, a natural series gives each bumper one fact; in the fictional Orlenna case that might be twelve hours hot, twenty-four cold and the copper layer, each claim a placeholder that a real brand would have to substantiate before it runs.

Reviewing 24 versions without reviewing 24 cuts

A large matrix needs a review plan, or the approvers end up watching the same edit two dozen times and missing the one file with a broken end card. Split the review into creative approval and technical checking:

  1. Approve one reference per length. The brand approves the 30, the 15 and the 6 in the lead ratio and language. These are the creative decisions.
  2. Approve one reference per ratio. For each additional ratio, approve one length (usually the 15) to confirm the framing, type layout and end card.
  3. Approve one reference per language. For each additional language, approve one length to confirm the voiceover, supers and legal lines.
  4. Check the rest against the references. The remaining files are built to the approved references and checked by QC for framing, type, captions, loudness, length and file name, rather than re-approved creatively.

In the fictional Orlenna matrix, that means 3 length references, 3 more ratio references and 1 more language reference: 7 creative reviews instead of 24, with the other 17 files covered by the 10-minute QC per version already in the 22.0 hours. Name in the plan who approves each reference, so a change requested on one version is applied to every file built from it.

Common cutdown mistakes and how to avoid them

Most cutdown problems are visible in review and expensive to fix after the shoot. The ones that recur:

Mistake What it looks like How to avoid it
Trimming the 30 The 15 is the first half of the 30 and stops before the product Script and board each length separately
Brand too late The logo appears in the last half second of the 6 Give the end card about two seconds or open on the product
Cropped end card The 9:16 version cuts the web address or the legal line Design end cards per ratio and check safe zones
Supers too fast A line from the 30 flashes past in the 15 Re-time type to the new length
Lost qualifier A claim survives the cut but its super does not Map claims and supers to every version
Music fade The 6 fades out mid-phrase Use alternate endings, stems or a sting
Over-length bumper A 23.976 edit of 144 frames runs 6.006 seconds End on frame 143, or work at 24 fps
Unplanned ratios A late request for 4:5 meets a frame shot wide Agree the ratios before the shoot
Matrix growth A new call to action doubles post hours Count and prune the matrix before adding a dimension

A checklist for planning cutdowns

Use this list at the brief, on set and in the edit.

Before the script is approved

  1. The media plan lists every placement, with its length rule and ratio.
  2. The matrix is counted and pruned, with edit and QC hours in the schedule.
  3. Each length has its own script, timed at the planned pace, with an end card budget.
  4. Claims and their supers are mapped to every version that makes them.

On set

  1. The boards mark which shots carry the 30, the 15 and the 6.
  2. Short, self-contained versions of key moments are shot.
  3. The monitor shows the 9:16 and 4:5 areas, and the product stays inside them.
  4. Native vertical setups, if planned, are shot back to back with the hero.
  5. Voiceover for each length is recorded in the same session.

In the edit and at delivery

  1. Each length is cut to its own structure and ending, not trimmed.
  2. Bumpers end on frame counts that stay within six seconds of running time.
  3. Type and end cards are re-laid out per ratio and checked against platform safe zones.
  4. Automatic vertical versions are allowed or turned off by decision.
  5. Every file is measured, captioned and named to the matrix.

Captions matter most in the short vertical versions, which often play with the sound off; whether captions are required and how to deliver them is covered in captions and audio description for video ads.

What to do next

Start from the placements, not the hero. List where the campaign will run, read each platform’s current length and ratio rules, and count the matrix with the versioning matrix calculator. Then write the cutdowns into the script and the shot list, so the shoot covers them. Campaign versioning describes how a matrix becomes named, checked files, and if you have a campaign that needs a hero and its cutdowns planned together, you can send a brief with the placement list attached.