End Cards and Calls to Action: What Actually Works

End cards and calls to action are the closing frames of a video, the moment where the viewer is told what to do next. They are also the part of most productions that gets the least attention. The script is argued over for weeks, the hero animation is polished frame by frame, and then someone drops a logo, a URL and three buttons onto the last two seconds of the timeline the night before delivery. That is backwards. The end card carries the action the whole video was made for, whether that action is booking a demo, subscribing to a channel, downloading a guide or visiting a store.

This guide matters to marketers commissioning video, in-house designers building a series, and editors who inherit a finished cut and are asked to "add a CTA." Rather than list rules in the abstract, it follows one project from brief to delivery and explains each decision about the end card along the way: how long it stays on screen, how it avoids the overlays that platforms place on top of it, how the action is spoken as well as shown, and how the design holds up across a series of videos.

The project below is an illustrative example, a composite built to show the reasoning. The company, the numbers and the schedule are realistic but invented for teaching purposes; they are not a client story or a measured result.

The Illustrative Project: A Product Explainer That Had to Earn Demo Requests

Picture a mid-sized software company that sells scheduling tools to clinics. Its marketing team commissions a 90-second animated explainer for its YouTube channel and website, plus two shorter cutdowns for social feeds: a 30-second vertical version for short-form platforms and a 15-second square version for paid placements. The goal written in the brief is blunt: the videos exist to generate demo requests. Brand awareness is welcome, but it is not what the budget is being judged on.

That single sentence in the brief changes how the end card should be treated. When a video is expected to drive a measurable action, the end card stops being decoration and becomes the conversion point. Every earlier scene is, in effect, a setup for the last few seconds. The production team therefore schedules end card design as its own workstream, with its own review, instead of leaving it to the final edit.

Those figures are the project's planning targets, not benchmarks from any study. They are useful because they force a conversation early: if the master is 90 seconds and the end card holds for 15, then roughly one-sixth of the running time is spent on the close. Stakeholders who have never thought about the end card tend to react to that ratio, and that reaction is exactly the discussion you want to have at the brief stage rather than during final approval.

The rest of this walkthrough follows the project in the order the decisions were made. If you want to see how the full service is scoped at the studio level, the animation and motion graphics production service page describes the typical phases; here we zoom in on the last few seconds.

Discovery: Choosing One Action Before a Single Frame Is Designed

The first working session is not about visuals at all. It is about deciding what the viewer should do when the video ends. The marketing team arrives with a list: request a demo, start a free trial, subscribe to the channel, follow on social, read the case studies and call sales. All six are legitimate business goals. Putting all six on the end card would guarantee that almost nobody does any of them.

The working principle in the reference material is simple: one action performs better than several. The reasoning is practical rather than mysterious. At the end of a video the viewer has a few seconds of attention left and is already reaching for the next thing. A single, clearly stated action asks for one decision. Three competing actions ask the viewer to compare them first, and a viewer who has to compare options often chooses none.

How the team ranked the candidate actions

To settle the question without a week of debate, the team scored each candidate action against three criteria: how directly it served the brief's goal, how much effort it asked of a viewer who had just finished the video, and whether it could be measured cleanly. Request a demo won on the first and third criteria. Subscribe to the channel was the lowest-effort action and made sense as a secondary element on YouTube only, where the platform provides a native subscribe element. Everything else was moved to the video description, the landing page, or a later email.

Notice what the decision does not say. It does not say "Learn more," which is the most common and least useful call to action in video. "Learn more" describes a vague activity rather than a concrete next step, and it gives the viewer no idea what will happen when they act. "Book a 20-minute demo" tells them the action, the commitment and roughly what they will get. Specific verbs and specific outcomes are the cheapest improvement you can make to any end card.

Writing the action as a short, repeatable phrase

The team also agreed on the exact wording before scripting began, because the phrase would appear in several places: spoken in the voiceover, set as on-screen type, printed on the landing page button and used in the video description. Keeping those identical matters. When the voiceover says "book a demo," the card says "Get started" and the landing page says "Contact us," the viewer has to translate between three phrases at the exact moment they should be acting. One phrase, used everywhere, removes that friction.

Scripting the Close: Writing the Call to Action Into the Voiceover

With the action chosen, the scriptwriter works backward from it. A common mistake is to write the whole story and then tack "Visit our website" onto the last line. The stronger approach is to set up the action in the body of the script so that, by the time the end card arrives, the viewer already understands why they would want to take it.

In the illustrative script, the middle section shows a clinic manager rebuilding a week's schedule in minutes. The closing lines then connect that outcome to the action: the voiceover says the viewer can see the same workflow with their own schedule in a 20-minute demo, and then says the action plainly, including where to go. The end card appears on the line where the action is spoken, not after the voiceover has finished.

Why the action is spoken, not only shown

The reference material is explicit that on-screen actions need spoken or captioned equivalents. There are two reasons. The first is accessibility: a viewer who is blind or has low vision cannot read a URL or a button that exists only as pixels, and a viewer who is listening while looking elsewhere, which is common for videos playing in a second tab, will miss a silent card entirely. The second is plain effectiveness: hearing an instruction while reading it reinforces it, and the spoken version often carries the reason to act, which the card has no room for.

The W3C Web Accessibility Initiative guidance on making audio and video media accessible covers the underlying principle: information conveyed visually needs to be available in another form, whether through narration, captions or audio description. For an end card, the simplest compliance route is also the best creative route. If the voiceover says the action, the captions will carry it, and the viewer who cannot see the card still gets the instruction.

Timing the voiceover against the card

The script marks three beats for the close. The first is the outcome line, which is the last line of the story. The second is the action line, spoken as the end card resolves. The third is silence, or music only, while the card holds. That third beat is deliberate. A card that holds under continuing narration forces the viewer to choose between listening and reading. Letting the voice finish and the card sit gives them room to act. If you are unsure how long the story beats before it should run, the principles in pacing for social video animation apply directly to how quickly the close should arrive.

How Long End Cards and Calls to Action Should Hold

End cards need time to be read and acted on. That sounds obvious, yet the single most common failure is an end card flashed for a second or two because the edit ran long and the card was the easiest thing to trim. The illustrative team treated hold time as a calculated number rather than whatever was left over.

The reading-time math

A practical way to set a minimum hold is to count what the viewer has to read and do. The team uses a working rule of thumb common in subtitle and motion work: budget roughly a third of a second per word for a viewer to read short on-screen text, allow for the time it takes the eye to find the text after a cut, and then add time to act, such as moving a cursor or tapping. This is a planning heuristic, not a standard, and you should test it by watching the cut cold, but it produces sensible numbers.

For the vertical 30-second cutdown, the card reads "Book a 20-minute demo" plus a short URL, about seven words in total. At a third of a second per word that is a little over two seconds of pure reading. Add about a second for the eye to land on the card after the transition and two to three seconds to act, and the minimum lands around five to six seconds. The team rounded up to six seconds for the vertical cut and five for the 15-second square cut, where the card text was cut down to four words and the action was a platform button rather than a URL.

Platform rules that set the floor

On YouTube the hold time is partly set by the platform. The end screen feature places interactive elements, such as a video, playlist, subscribe button or approved website link, over the final stretch of the video. According to YouTube Help's guidance on adding end screens, end screen elements can be shown in the last 5 to 20 seconds of a video, and the video itself has to meet a minimum length for the feature to be available. Always check the current help pages before locking an edit, because platform rules change and the numbers in any article, including this one, can drift out of date.

For the master, the team chose a 15-second end screen segment. That is long enough for the native elements to animate in, for the voiceover to finish the action line, and for a viewer to decide. It is short enough that the video does not feel padded. A 20-second hold was considered and rejected because the story had nothing left to say and the music bed would have had to loop visibly.

Designing Around Platform Overlays and End Screens

Some platforms overlay their own end screens and interface elements on top of your video. If the end card was designed as a full-frame graphic with the button in the middle, the platform's elements can sit directly on top of it, hiding text or creating two sets of buttons that fight each other. Text hidden behind platform elements is one of the mistakes the reference material names, and it is almost always the result of designing the card in isolation from where it will play.

Designing the YouTube master as a frame for native elements

For the 16:9 master, the team designed the end card as a background layout with deliberate empty zones rather than as a finished graphic. The layout reserved space on one side of the frame for the platform's video or playlist element and a smaller zone for the subscribe element, following YouTube's end screen templates. The brand's own content, the action phrase, the short URL and the logo, occupied the remaining area, well clear of those zones.

Because the demo booking happens on the company's website, the team also confirmed early that the channel was eligible to add a website link element; where it is not, the action has to be carried by the voiceover, the on-screen URL and the description link instead. That check belongs in discovery, not the week of delivery.

Vertical and square cuts: the invisible interface

Short-form vertical platforms place their own interface over the video: captions and account names along the bottom, action buttons down the right edge, and sometimes a progress bar or header at the top. None of that appears in your editing software, so the card has to be designed with an overlay guide in place. The team built a simple guide layer for each target platform by capturing a screenshot of the live app interface at the target resolution and placing it as a non-rendering reference layer in the composition.

The rule that came out of this was straightforward: the action phrase and URL sit in the central band of the vertical frame, a comfortable distance from the bottom caption zone and the right-edge buttons. The logo sits above them, not in a corner, because corners are where interfaces cluster. The same thinking underpins the classic broadcast conventions described in getting title safe and action safe areas right; platform overlays are simply a modern, less symmetrical version of the same problem.

Layout and Hierarchy on the Final Frame

Once the zones are defined, the layout inside them is a question of hierarchy. The viewer's eye should land on the action first, then the destination, then the brand. The illustrative team's hierarchy looked like this, from most to least prominent: the action phrase, the URL or button label, a one-line supporting reason, and the logo.

Type size and contrast

End cards are viewed on phones far more often than editors assume, so type that looks generous on a 27-inch monitor can be unreadable on a small screen in a bright room. The team set the action phrase as the largest type in the entire video and checked every card on an actual phone at arm's length. They also checked contrast between the text and the background with a contrast checker, aiming to meet the WCAG contrast guidance for text, because a pale brand color on a white card is a common and entirely avoidable readability failure.

The URL problem

URLs are awkward on end cards. They are long, they are often hard to spell after hearing them once, and on most platforms they are not clickable in the video frame itself. The team registered a short vanity path on the company domain, the equivalent of example.com/demo, and redirected it to a tracked landing page. The short path is easier to read, easier to say in the voiceover and easier to remember. The tracking parameters live in the redirect, not on screen.

What was deliberately left off

The first layout draft included a phone number, social handles and a tagline. All three were removed. The phone number competed with the primary action, the social handles were an action the brief did not prioritize, and the tagline duplicated the voiceover. What remained was four elements in a clear order, with generous space around them. Empty space on an end card is not wasted; it is what makes the one thing on it readable in the few seconds available.

Motion on the End Card: How It Arrives and Then Stays Still

An end card is still motion graphics, and how it arrives shapes whether it reads as the natural end of the story or as a bolted-on advertisement. The team made three motion decisions for the close.

First, the transition into the card continues the visual language of the rest of the video. The explainer used a calendar grid as its recurring motif, so the final scene's grid collapses into the card's layout rather than cutting to a new graphic. That continuity tells the viewer they are still in the same piece. The broader principles are covered in transitions and scene changes.

Second, the elements build in order of hierarchy. The action phrase resolves first, the URL a beat later, the logo last. Staggering the arrival guides the eye in the intended reading order. The whole build completes in roughly a second, so that the full reading time is available on a stable card.

Third, once built, the card stops moving. It is tempting to keep a subtle loop running, a drifting gradient or a pulsing button, to prevent the frame from looking frozen. Small ambient movement is fine, but anything that pulls the eye away from the text, or that makes the text itself move, works against the whole purpose of the hold. The team allowed a slow background drift and nothing else. If you are delivering to web players or animated formats, it also helps to check that the close renders smoothly at the target rate; frame rates for web animation explains the trade-offs.

Accessibility: Captions, Spoken Actions and Audio Description

Accessibility was treated as part of the end card specification, not a separate compliance pass at the end. That decision saved rework, because most accessibility fixes for end cards are much easier to build in than to retrofit.

Captions that carry the action

The captions for every version were written from the final script and timed so that the action line appears as a caption at the same moment the card resolves. The team checked that captions did not collide with the card's own text. On the vertical cut, where platform captions and the video's burned-in captions can both appear, the burned-in captions were positioned above the card's text block during the close, and the caption file was reviewed separately for the platforms that display their own.

When the visual carries information the audio does not

In the 15-second square cut there is no voiceover, only music, because it runs in feeds where most viewers watch with the sound off. That reverses the usual problem: now the card is the only place the action exists. The team handled this in two ways. The card text was written to stand on its own, and the paid placement's own text field repeated the action in words. For any version that plays with sound, the rule remained that nothing important is shown without also being said or captioned.

Contrast, flashing and motion sensitivity

Beyond captions, the card was checked for adequate text contrast and for any flashing effects during the transition, which were avoided entirely. The motion into the card was kept smooth and short. These are low-effort checks that remove real barriers for viewers with low vision, photosensitivity or vestibular sensitivity, and they are consistent with the W3C guidance cited earlier.

Versioning the Close for Every Platform and Cut

With three aspect ratios and a series planned, the team needed a way to produce end cards consistently without redesigning each one by hand. The versioning plan was built into the project schedule from the start.

The schedule is illustrative, but the order matters more than the durations. The action and hold times are decided before animation, so they cannot be squeezed later. The test uploads come before delivery, because overlays and native end screen elements can only be fully verified in the real player. Many teams skip that week and discover the collision after the video is live.

What changes between versions and what does not

What stays fixed is the action itself, the brand treatment and the visual motif. What changes is everything the platform dictates: space, timing and the mechanism for acting. Keeping the fixed elements fixed is what lets a viewer who sees the vertical cut on their phone and the master on their laptop recognize the same ask.

Keeping End Cards Consistent Across a Series

The explainer was the first of a planned series of short feature videos. Keeping end card design consistent across a series is one of the good practices in the reference material, and the reason is recognition. When every video in a series closes the same way, returning viewers learn where to look and what the close will ask of them. That familiarity lowers the effort of acting, which is the whole point.

Building the end card as a template

The team built the end card as an editable template with locked layout and animation and exposed text fields for the action phrase, the URL and an optional supporting line. For the in-house editors who would produce later episodes, the template was exported for their editing software with only those fields exposed, so the timing, hierarchy and overlay-safe zones could not be accidentally broken. The approach is described in more detail in motion graphics templates for editors.

What the template locks and what it allows

The locked elements were the layout grid, the safe zones for each aspect ratio, the build order and timing, the type styles and the minimum hold. The editable elements were the words, within a character limit that the layout could accommodate at the required type size. The character limit is important: without it, someone will eventually paste a 60-character action phrase into a field designed for 25, and the type will shrink until it is unreadable.

The team also archived the full project, including guide layers, fonts, caption source files and a short written specification of the end card rules, so that a future team could rebuild or update the template without reverse engineering it. The practical details are covered in archiving animation projects.

Reviewing the Close: A Scorecard Before Sign-Off

Before final approval, the team reviewed every version's close against a short scorecard. The point was to make end card approval a specific check rather than a general impression of the whole video, because stakeholders reviewing a 90-second film tend to judge the story and wave through the last frames.

Every item on the scorecard is binary and can be checked in a few minutes. In the illustrative project, the first review round failed two checks: the vertical cut's URL sat under the right-edge buttons in a test upload, and the square cut's card held for three seconds because the edit had been tightened. Both were fixed before sign-off. Neither would have been caught by a reviewer watching the video once on a desktop monitor.

Planning how the close will be judged

Because the brief tied the videos to demo requests, the team also agreed in advance how they would judge the end card after launch, without promising any particular result. The plan was to track visits to the short URL's landing page, clicks on the YouTube end screen elements in the platform's analytics, and completed demo requests from the tracked landing page. They also agreed on what they could change if the numbers disappointed: the wording of the action, the hold time, or the position of the card relative to the voiceover. Changing one variable at a time across later episodes, instead of redesigning everything, is what lets a team learn what actually works for its audience.

When to Bring In Specialist Help for End Cards

Plenty of end cards can be handled in-house. A team with a clear template, a single action and a reviewer who checks test uploads on a phone will avoid most of the common mistakes. The case for outside help is strongest when video is expected to drive a measurable action and the stakes of getting the close wrong are real: paid campaigns where every impression costs money, product launches, or a series where the end card template will be reused dozens of times.

In those situations, a specialist team brings three things that are hard to improvise. The first is process: treating the end card as its own workstream with locked timing and overlay-aware layouts. The second is a template system that in-house editors can use without breaking it. The third is the discipline of test uploads and binary review checks before anything goes live. If that is the situation you are in, the motion graphics and animation services overview explains how that kind of engagement is usually structured.

Whether you build end cards yourself or commission them, the lessons from the illustrative project hold. Decide the one action first. Write it into the script and say it out loud. Calculate the hold time and protect it. Design around the platform, not on top of it. Test on real devices. Keep the close consistent across a series so viewers learn what it asks. None of these steps is expensive, and together they turn the most neglected seconds of a video into the part that does the work the video was made for.

Where this comes from

• YouTube Help — Add end screens
• W3C Web Accessibility Initiative — Making Audio and Video Media Accessible

The figures and practices above come from the sources listed.

Working on something like this?

We take on Motion Graphics & Animation work for teams who want it done once, properly. Tell us what you are building and we will tell you honestly whether we are the right studio for it. Start a project.

Where to go next

• Seasonal Campaign Variants: The Decisions That Matter
• Motion Graphics in Presentations: The Decisions That Matter