Behind the scenes of Railway's launch video: from request to release in under two weeks

Launch videos have a strange reputation. Everyone wants one, everyone assumes it takes a month, and most teams quietly discover that the video becomes the thing holding up the launch. So when Railway came to us with a request, the interesting constraint wasn't the creative brief — it was the calendar. From the initial request to the final release, we had less than two weeks.

Here's the behind-the-scenes look at how it came together.

What a launch video actually has to do

If you've never been on the inside of a product launch, it's worth spelling out why this asset matters so much. A launch video is the thing sitting at the top of the announcement — the first frame in the tweet, the hero on the landing page, the clip replayed on stage or dropped into a Slack channel by someone saying "look at this." It's usually 30 to 90 seconds, it's usually autoplaying without sound, and it has to communicate what changed and why anyone should care before a viewer's thumb keeps scrolling.

That's a lot of weight for one asset, and it's why teams tend to over-engineer the process around it. The irony is that over-engineering is exactly what makes the timeline slip. Every extra round of "let's explore another direction" costs days, and days are what you don't have when the launch date is already announced.

Why two weeks is the hard part

A typical video production splits into stages that are traditionally done one after another: agree on the message, write the script, storyboard it, design the visual frames, animate, add sound, review, revise, deliver. Run those in sequence with a review cycle between each one and you're comfortably at four to six weeks — most of which is spent waiting rather than making.

Compressing that into under two weeks isn't about working faster in the animation software. It's about restructuring the order of operations:

• Lock the narrative first, and only once. Before anything visual happens, we agree on what the video says and in what order. That's the single most expensive thing to change later, because every frame after it depends on it. Once the story is signed off, it doesn't reopen.
• Run stages in parallel instead of in series. Sound design, motion, and the final scenes don't all have to wait for each other. While one part is being animated, the next is being designed. It only works if the narrative is genuinely locked — otherwise parallel work becomes parallel rework.
• Review on a schedule, not on availability. Short projects die in inboxes. Fixed review windows with the decision-makers present turn a three-day feedback loop into a two-hour one.
• Design for the format it'll actually live in. Legible without sound, readable on a phone, strong first frame, no critical information in the last five seconds where most viewers have already left.

None of that is glamorous, but it's the difference between a video that ships with the launch and a video that ships two weeks after everyone stopped paying attention.

Speed is a creative constraint, not just a logistical one

The assumption worth pushing back on is that a fast video is a compromised video. In my experience it's the opposite more often than people expect. A tight timeline forces you to make decisions — to pick one idea and commit to executing it well rather than hedging across three half-explored directions. The videos we're least happy with are almost never the fast ones; they're the ones that got sanded down over a month of committee feedback until the original idea disappeared.

What a short schedule genuinely does cost you is optionality. You can't shoot on location, wait for a reshoot, or discover the concept halfway through. So the concept has to be right at the start, which means the first few days matter far more than the last few. That's where the real work goes: getting the story, tone, and visual language agreed before a single production hour is spent building the wrong thing.

If you're planning your own launch video

A few things I'd tell any team with a date on the calendar and no video yet:

1. Come with the message, not the shot list. "Here's what we want people to understand" is a brief we can move fast on. "Here's a reference video we like" is not.
2. Name one decision-maker. Two is manageable. Five means the timeline is fiction.
3. Decide what the video is for — announcement hero, docs explainer, conference opener — because those are different videos, and trying to make one asset do all three is the most common reason projects stall.
4. Start earlier than you think you need to, but don't confuse an early start with a long process. Two focused weeks beat six distracted ones.

Working with Railway was a good reminder that a compressed schedule isn't an excuse for a lesser result — it just moves where the effort has to go. If you have a launch coming up and a timeline that looks impossible, that's usually a conversation worth having rather than a reason not to.