Here are some thoughts on video content - form, length, frequency and more, especially in a learning and sharing context, and specificially in a developer advocate content.
Just now my good friend Ronnie Sletta drew our attention to a question by Chris Roberts on video content: “I’m thinking of making some free courses and putting them on YouTube. Do you prefer a series of short videos, or one long video? If a series, should I release one a week? If one long video, how long?”.
I started to reply on Twitter, then found myself needing to use the “1/n” tweet thread approach, which I’ve never really liked, so I thought I’d take a leaf out of Scott Hanselman’s book and reply once in a form that’s arguably more permanent and easier to read.
First off, I think it’s important for us to remember that folks have different preferences when it comes to consuming content and learning. Some prefer the written word, and some prefer video instruction or demonstration. But it goes deeper than that.
Descend into the video category and within that there are some that like to settle in and watch a “feature length” video. Others prefer short, sharp and to-the-point videos, the moving picture equivalent of good answers on Stack Overflow. Then there’s the question of time available, which also factors into the decision on which videos to watch. Often I find myself thinking:
“OK, I’ve got 15 mins before the next meeting, and I’d like to learn more about X - what videos fit that combination of time and topic filter?”
It seems obvious to me that when it comes to the time filter, shorter time slots are going to be more common, so in one sense shorter videos in general seem like a good idea.
But there are cases to be made for both short and long form videos. They serve different purposes and contexts.
Before I proceed with this thought, let me be clear on what I mean by “long form”. There are recordings of live streams from YouTube or Twitch sessions that are very often far more than an hour or so. A live stream that is just one hour is actually quite unusual.
For me though, an “hour” is long form. Anything longer than that is beyond the question at hand, and for me fits into the “I want to watch my favourite streamer and relax, so I’ll watch this recording as they’re not live right now, or because it covers something I’m interested in” category.
It’s worth pointing out at this point that the main live stream episodes that I put out are deliberately limited to one hour. That’s for many reasons, here’s the main one:
- the folks that join have other things to do in their day, meetings to go to, and so on; I don’t want to make it harder for them to do that, and by streaming from the top of the hour to the top of the next hour makes it easier for them to fit the live stream attendance into their work calendar
This applies to me too - I have other tasks to accomplish and meetings to attend in my working day.
With this in mind, I’d say that the “long form” categorisation of one hour videos only applies if they’re not live. I’d almost go so far as to say that a live stream of less than an hour has more disadvantages than advantages:
- the time to establish the topic and focus is limited
- some folks joining don’t get enough time to “warm up” and feel confident enough to participate in the chat
- given that a live stream has a fixed broadcast start time, if you miss it, you miss it, so a longer live stream gives latecomers (not anything to be ashamed of!) chance to actually catch something, participate and get something from it
While short form videos are great for focused search, confirmation and consumption, there’s a key feature that’s essential for making longer form videos more easily consumable and more useful, and that’s the video chapters feature.
I think they go some considerable way towards bringing the “long form” videos closer to “short form” in consumability and relevance.
I’ve been making use of video chapters for a while, and I can thoroughly recommend it.
I use video chapters in my live stream recordings; after a live stream is finished, I scrub through the recording and then add video chapter information in the description, making it much more accessible and useful for consumers. Here’s an example from an episode of the Getting the most out of the SAP TechEd Developer Keynote repository series on the Hands-on SAP Dev show*.
*See An overview of SAP Developers video content for more details.
My other SAP Developer Advocate colleagues use video chapters too.
Talking of SAP Developers video content, we’ve just launched the first video in a new shorter form show - SAP Tech Bytes. The videos here are deliberately short, to be more consumable in a shorter amount of time, to be focused on one specific topic, and also to provide form variation.
Here’s the first episode: SAP Tech Bytes: Tutorial - Create SAP HANA Database Project. Note that there are video chapters even in this shorter form content.
In fact, we use video chapters in even our shortest form videos - the episodes of the SAP Developer News show, where each video is only around five minutes long - deliberately a coffee break length.
Frequency and schedule
I’ve just got to here and realised I haven’t even talked about frequency. I thought it might be at least helpful to give some examples; I’ve been live streaming in my role as SAP Developer Advocate since January 2019, and have kept more or less the same frequency since then, which is weekly. I’ve done the occasional second live stream in a single week, but I treat (and refer to) those as “off piste” and not really part of the usual cadence.
Perhaps more importantly than the frequency, at least for live streams, is the consistency of day and time, i.e. the schedule. I think of my episodes of Hands-on SAP Dev as episodes of a TV programme, and again, because it’s broadcast live, the best way to help folks not to miss it is to be consistent and predictable with the schedule. That’s why I broadcast my episodes on Fridays, at 0800 GMT.
Incidentally, I chose a relatively early morning slot because that’s when I’m most awake and my brain is buzzing - I’m a morning person and the plate spinning that’s required to stream live is slightly less difficult then.
I think the same scheduling rules apply to YouTube Premieres too. Premieres are videos that you can pre-record but have the first broadcast set for a fixed date and time in the future, with all the build up and excitement of a live stream, but during playback you can attend and interact in the chat with the viewers. Sort of a combination of live stream and recording, which can work really well.
When it comes to recorded videos, i.e. neither live streams nor premieres, then the schedule is not that important, so it just comes down to frequency. And that really depends on two factors:
- primarily: how much content you can produce
- secondarily: how much content your viewers can consume
There’s a balance you need to find between these two factors, and there’s no algorithm I know of that will provide a solid answer here; it really depends on how you work, what you have to share, and so on.
That said, here’s a general piece of advice: If you have a load of pre-recorded videos waiting to upload to YouTube, don’t be tempted to just publish them all at once. Resist the urge to flood your viewers’ brains with all that wonderful goodness, and instead publish them in a spaced-out fashion instead. That has two advantages:
- they don’t feel overwhelmed and tempted to give up because there’s too much to consume
- you don’t burn yourself out, and instead give yourself time to think of other content
Anyway, this post is already far longer than I expected it to be; I’ll bring it to a close now, but as it’s just blog post content, I may come back to it in the future and update it as I see fit. That’s the wonderful nature of blog posts and how they’re still the backbone of many communities.