Picture a customer in their second week. They want to set up an automation, so they open your help center and press play. The video shows a sidebar with six items. Their sidebar has four, and the one they need is now inside a menu. They pause, look around, go back, try again. Then they open a support chat, or close the tab.
Nobody on your team sees this. There is no error, no alert, no failed build. The video still has a play count, and the play count is still going up. That is what makes outdated videos expensive: they look like they are working.
Where the cost actually lands
Support tickets
Every customer who follows a wrong video and then asks for help is a ticket the video was supposed to prevent. Worse, the ticket is harder than the original question, because the customer is now confused by the mismatch as well. Support answers it, the video stays wrong, and the next customer does the same thing.
Onboarding that stalls
Onboarding screens are exactly the screens product teams redesign most, because they care most about them. So onboarding videos tend to go stale first. A new customer who cannot follow setup does not always ask for help. Often they just do not come back.
Trust
When a video shows an old interface, the viewer does not only doubt the video. They start to doubt the rest of your help, and sometimes the product. "If their own tutorials are wrong, what else is?" It is the same reason a broken link on a pricing page makes people hesitate to pay.
Sales and marketing
The homepage demo shows last year's design. A prospect signs up, sees a different product, and spends the first minutes reorienting. Sales reps learn to say "ignore the video, it's a bit old", which is an awkward way to start a call.
Internal training
New support and success hires learn the product from the same videos. If those are out of date, your own team learns the old way, and passes it on.
A ten-minute estimate
You do not need exact numbers to see the size of the problem. Make a list of your product videos and, for each, note roughly which screens it shows. Then look at the last two quarters of releases and mark which of those screens changed.
Say you have 40 videos, each showing about four screens. Your team redesigned or renamed things on 15 screens in six months. If those changes are spread across the product, it is very likely that well over a third of your videos now show something that is no longer there. Do the count for your own library before deciding it is fine.
| Question | Where to find it |
|---|---|
| How many product videos do we have? | Help center, YouTube, onboarding emails, website, sales folder |
| Which screens does each show? | Watch at 2x speed and note them, once |
| Which screens changed recently? | Release notes, changelog, design tickets |
| Which videos get watched most? | Your video host or help center analytics |
| Which tickets mention a video or article? | Search support tool for "video", "tutorial", "doesn't match" |
Where to start
- Sort videos by views. The top ten carry most of the risk.
- Watch each of those ten next to the live product. Mark it fine, needs a re-record, or retire.
- Retire generously. A video nobody will maintain is a future wrong answer. Fewer, current videos beat a big, stale library.
- Re-record the top ten that need it, keeping each to one task.
- Give every remaining video an owner and a "last checked" date.
- Add one line to your release checklist: which videos show the screens this release changes?
Make the fix cheap, or it won't happen
Teams do not leave videos outdated because they do not care. They leave them because re-recording, re-narrating, and re-editing a video for one moved button loses every prioritization fight against new work. The lasting fix is making updates cheap enough that they happen in the same week as the release. We cover the practices in how to keep product videos up to date after every UI change, and the same problem in written docs in why help docs rot after every release.
That is a big part of why we built Steperly the way we did. A guide's video, clickable demo, and written steps come from one recording, so they cannot disagree. Narration is editable text, so renaming a button means changing a sentence. And when a flow changes, you refresh one guide instead of three separate assets.
If your team has already stopped making videos because of this, read why SaaS teams stop making product videos.