If you're trying to decide whether your launch video should be thirty seconds or a minute, start with what has to happen inside it. A duration target is useful as a constraint, but it doesn't tell you which parts of the explanation you can remove.
I'd build a rough sequence around the reason for the announcement, the product change and the next step. Then check how long each part needs to be understandable. That gives you a reason for the length instead of a number you've borrowed from a different kind of video.
Separate the publishing limit from the editorial decision
First check the destination's current limits with whoever will publish the video. A hard upload or placement limit is a production requirement. Your preferred length is an editorial choice within that requirement, and the two shouldn't be confused.
A launch announcement someone encounters while scrolling has a different context from a walkthrough they deliberately open on a product page. Decide which situation you're making the cut for before comparing durations. You may need separate versions rather than one compromise that doesn't explain enough in either place.
Don't treat a round number as evidence of what your audience will watch. Without relevant data from your own videos, it's a planning assumption. Write it down as such so the team can question it when the explanation doesn't fit.
Identify the evidence you cannot cut
Imagine a fictional scheduling product announcing shared availability. The viewer needs to see which people are included and the available time that follows. If you remove the selection to save time, the result may look like an ordinary calendar rather than the new feature.
This is a hypothetical planning example, not a measured video result. Write down the equivalent proof for your own announcement. Ask what the viewer would misunderstand if each part disappeared, and protect the scenes that answer that question.
You can often cut repeated claims before cutting evidence. If narration, a headline and a closing card all say that scheduling is easier, decide which statement actually helps. The available time appearing beside the chosen people may explain more than another sentence praising the feature.
Budget time for finding and reading the important detail
A scene needs time for someone to orient themselves as well as time for the action. You know where the control is because you made the product; a new viewer has to locate it before understanding what changed.
Put the rough scenes together and watch without pausing. If the result disappears just as you identify it, extend that moment or simplify the scene. Faster transitions won't recover the missing reading time.
Consider competing information too. If a viewer has to read a subtitle, compare two values and follow a cursor at once, you may need to sequence those tasks. The answer isn't always a longer overall video; sometimes you can remove a secondary element and make the same moment easier to follow.
The hold-time example gives you a visual reference for the pause after a reveal. Treat its timing as an example, then adjust for the amount your viewer needs to read.
Use a timing sketch as a hypothesis
Make a rough timing sketch and label it as a proposal. The example below allocates 30 seconds to a fictional announcement so the team can discuss tradeoffs. It isn't a recommended benchmark or a claim about conversion.
Hypothetical 30-second planning cut: 0-5 seconds: Establish that the announcement concerns shared availability. 5-12 seconds: Show which people are included. 12-22 seconds: Show the available time and why it corresponds to the selection. 22-27 seconds: Show the verified next action. 27-30 seconds: Give a readable next step. Review question: Can a new viewer identify the panel and result without pausing? If not: Remove competing copy or give the result more time before shortening essential steps.
Make a shorter cut by narrowing its promise
If that sequence doesn't fit the placement, consider asking it to explain less. A short announcement could introduce shared availability and show one clear result, then link to a separate walkthrough for configuration. That is a scope decision, not permission to make a multi-step workflow look automatic.
Keep any qualification needed to understand the result honestly. If availability depends on connected calendars or a particular permission, decide how the video and surrounding explanation will make that clear. Don't cut the condition merely because the narration sounds cleaner without it.
A longer cut may still be the right choice when the action needs context. If the viewer deliberately opened a demonstration, an understandable extra step can be more useful than an abbreviated sequence that leaves them unsure how the product works.
Review comprehension before arguing about a few seconds
Show the rough cut to someone unfamiliar with the feature and ask what changed, what caused the result and what they would do next. Avoid explaining the missing context while they watch; otherwise you're evaluating the video plus your live commentary.
If they misunderstand the workflow, identify the moment responsible. You may need a clearer label, a different order or more time on a result. Simply extending the opening doesn't solve a missing connection later in the sequence.
Use the feedback to revise the timing sketch, then export and review again. Keep notes about the issue each timing change was meant to resolve. That makes the conversation about length concrete rather than a contest between people who prefer fast edits and people who prefer slow ones.
Judge the published cut by the job you assigned it
After release, distinguish people starting a video from people understanding or acting on the explanation. Any available viewing data needs context: placement, audience and how the video starts can affect what you observe.
Use that information to form the next question rather than declaring one duration universally best. For the current cut, the practical goal is a complete, honest explanation with enough time to recognize the important result and a clear next step.