Screencast Workflow Best Practices

We recently created a number of screencasts for clients to support software training. As public-facing screencasts for enterprise software, the standards had to be very high. That meant 1 single person couldn’t do everything; we needed a professional voice actor, a software SME, and someone from our team to take care of instructional design and editing. While we’ve certainly created many screencasts, I looked online for workflow best practices for creating high-quality screencasts with multiple resources.

Unfortunately, very little guidance and a lot of complaints were all that I could find. For this type of screencast you can’t wing it; having the SME record the video and then building a script around it just doesn’t work. And getting the voiceover work done prematurely, without proper planning, tends to mean rework. So what’s the best approach?

While it might not work for every situation, we’ve developed an approach to screencasting that generally works well for Uncanny Owl. Here are the steps we follow:

1. Plan everything. We start off by looking at the planned outcomes and objectives. What does the learner need to get out of the screencast? What’s the best way to achieve it using the software? This certainly requires a lot of collaboration with an SME and production of a draft script. We can have the software up on the screen and walk through exactly what the learner needs to see and how best to explain it.
2. Record the video. The SME does the actual interaction with the software while someone from our team observes the recording to do initial script reconciliation and verifying that the pace is appropriate. With 2-3 good takes, there’s typically enough video to perform a more detailed reconciliation and to tailer the script as needed. If anything needs to be recaptured, it can be done almost immediately. It may even help to have someone read the script while the SME is performing the activities.
3. Finalize the script. This needs to get as close as perfect as possible before it goes to the voice actor. Try recording it yourself and listen to it both with and without the video to make sure it flows well and is straightforward to learners.
4. Record the audio track. Since the voice talent may not have the video context to work with, include directional cues in the script as required.
5. Put everything together. Cut the audio up and add it to the video, syncing everything so it seems like the person interacting with the software is the same person speaking. Add or shorten the video as appropriate, and make sure any edits don’t hurt the pacing or cause problems with mouse movements. Once that’s all done, the screencast is ready for publishing!

That workflow has produced generally good results for us, but we’d love to hear your tips for better screencasts in the comments.