Audience retention is, quite often, a big fat disappointment. Why doesn't your audience love your video as much as you do? You're trying to communicate valuable information here! What don't they like about it? Is it too long, perhaps? Are you doing something wrong? What can you do to get your audience to watch the whole thing?
I assert the following: you can structure your content in such a way that your audience will, indeed, watch your video all the way to the end!
But first, a disclaimer: producing a web series is a wholly new experience for me, so I'm hardly an expert on the subject. But it looks like retention is one thing I'm getting (mostly) right from the beginning.
Here's the audience retention for the Green Screen That Doesn’t Suck! video:
And the retention for the Hot Lights vs. Fire Sprinklers video:
Among those who make it past the first 15 seconds, most of my audience watches the whole thing (although some don't make it all the way to the end, which I discuss later).
If you're a celebrity amongst your audience, people will listen to anything you have to say. But considering that, as of this writing, no one really knows who I am, I'm pretty satisfied with my results so far.
I... kinda did this without trying. My process is just something I've honed over years of video production; it's not really a step-by-step guide that I follow when I sit down to write the script. So I reverse-engineered my own intuitive process, and I now present to you: Nick's Guidelines for Making Instructional Video That Will Be Watched to the Very End!
1. Let the audience know what this is all about.
This is a trend I've noticed in every FYV episode so far: there's an immediate dropoff right after the beginning.
Now, maybe I'm rationalizing when I say this, but I'm gonna say that that's a good thing.
Why the heck would it be a good thing? Because this video isn't necessarily for you.
I do quite a bit of research before I start writing my script. I know the kinds of pains that people face. So by resonating with my intended audience right off the bat, they can say, “Holy crap, I do have that problem!” Simultaneously, everyone else is saying, “I don't use incandescent lights, so I don't have to worry about sprinkler heads!” And they leave, and I don't waste any of their time, and everyone's happy.
But even if they do leave right away, there's still a small chance that they'll recognize the style in a future video and say, “Oh yeah, didn't they make some kind of sprinkler video?” And they watch, and I look like a little bit more of an expert, and they stick around for a video that's more relevant and useful to them, and – once again – everyone's happy.
(I talk more about resonance below.)
2. Never let the air out of the balloon.
In the medium of video, you need to trim the fat relentlessly. Your audience can't just skim over the dull parts as easily as they can with text. Your goal, really, is to do the opposite of that: jam-pack your video with so much information that they wouldn't even think about skipping around, because they might miss something!
The tempo is pretty fast in a Fixing Your Video! video. It progresses linearly. One thought leads directly to another. I try to keep this kind of thing to a minimum: “Okay, we've finished explaining this concept, so now let's go to Topic #2!” If I find myself doing that, I consider the possibility of saving topic #2 for a separate video.
If I'm covering an event (like one of our experiments) where a script wasn't written beforehand, I distill the resulting footage down into its most basic, indispensable elements. A five-minute conversation might end up as two or three sentences. It's one of the few instances where, perhaps, it is better to tell instead of show: by including Robbie's line of “Science is a lot of__ waiting!” in the sprinkler video, I don't have to waste valuable seconds showing that absolutely nothing happened during most of the experiment.
How fast is fast enough? My most important principle is that I explain concepts as clearly and concretely as I can. That doesn't mean I explain every facet of a topic, but it does mean that I'm confident my audience will understand the parts I do explain. I never go so fast that I skip or obfuscate the really juicy parts.
3. Stick to the script.
I'm gonna offer a counter-argument to the just-be-natural strategy: I, for one, can always tell when someone on camera is struggling to think of what to say next, whether they're trying to be extemporaneous or just can't quite remember a line. Those mirror neurons in my brain – you know, the ones that make our audience laugh and smile, the ones we're trying so hard to stimulate through video – go nuts when this happens. My eyes unconsciously narrow. I can feel my muscles getting tense. I'm feeling awkward on their behalf.
Do the naturally-occurring smiles outweigh the naturally-occurring tension? Probably, but that's a topic for another blog post.
If you're highlighting your organization's world-renowned culture, acting natural is the way to go. But my main goal is to communicate tightly-packed, concrete, actionable information (while not coming across as a total dud). I could only ever achieve that density by writing a script and sticking to it.
You know what else? I use a teleprompter.
You know how when you're shooting, and you find yourself saying over and over: “God, I wish I had a [GEAR] right now!”? For me, nine times out of ten, the thing I really wanted was a teleprompter. I'm so glad I finally built one.
I mean, come on. How much time would I waste trying to memorize those lines? (How much of my clients' time have I wasted because they couldn't remember all of theirs?)
And as for being extemporaneous? Yeah, I'm way too awkward to pull that off well. Extemporaneity is a lot easier when I'm working with Rob and Aaron because of the resulting joviality and camaraderie, but I make a lot of these episodes myself.
“But wait!”, I can hear you saying. “Aren't you supposed to just let the words flow when you're on camera? Doesn't a teleprompter make you sound like a robot?” Yes. That is a very real risk. So I do what news reporters do: read your copy over and over again before you roll. But practice a natural-sounding cadence. Try your best to not sound like you're reciting the Pledge of Allegiance. (“And to the rePUBlic, for which it STANDS, one NAtion...”)
I don't know, maybe I do sound like a robot and don't realize it. I think I'm okay with that. 🙂
By the way, if you've never heard of a two-column rundown >before, lemme tell you something: this is the best way I know to flesh out a video quickly. More importantly, it encourages – nay, requires! – you to write your script in recognition of the fact that video is a visual medium. _ It is a constant reminder that you need something to _see at all times, and it needs to line up with the things that are being said.
4. Show cause and effect.
Okay, so, like... am I the only one who gets really annoyed when someone just stands in front of a camera and talks about the right way to do something?
There's one shot in the green screen video where I went to some trouble to pull off a neat special effect: I put myself in an After Effects window, while a mouse pointer changed the effects in real-time. It wasn't as easy as it looked; I had to stack a bunch of layers and mess around with difference mattes and add keyframes one frame at a time.
In a later shot, I demonstrate how auto-exposure and auto-white-balance will wreck your key effect. I could've faked it in post, but I thought it would be a lot more useful if the camera's settings were actually changing. So I wrote a Magic Lantern script that changed the exposure and white balance settings at random (which, sometimes, is how full-auto mode seems to behave).
Why did I go to all that trouble? It wasn't to look cool. (Of course, it always helps when it does look cool.)
My goal is to explain things as clearly as possible. And I couldn't think of a more efficient way than showing the audience exactly what they should be looking for, whether they're diagnosing a problem or trying to reproduce my steps.
After all, if you're not gonna communicate visually, why not just use text?
5. Drop gold coins.
This one's pretty simple: every once in a while, give your audience a reason to keep watching.
It's like training your dog: by providing reinforcers at unpredictable intervals, you're encouraging the desired behavior. In this case, we simply want the audience to keep watching!
(Behaviorism, by the way, always sounds a lot more sinister than it actually is. Positive reinforcement is all about a net benefit for both parties. Mouse, for instance, loves him some food, and I love watching him stand on his hind legs while giving me a high-five.)
In an instructional video, just as long as you provide a lot of good information that your audience can actually use (which I discuss below), you're providing plenty of reinforcement. But if, say, your faithful, smart, loveable, and cuddly Labrador Retriever follows you around everywhere you go, why not toss it in there? People love dogs!
You have no idea how hard it was to write this post with a 60-pound dog on my lap.
One important caveat: if it detracts from the instruction, it's best to leave it out.
6. Don't discount showmanship.
Here's the dirty little secret of the sprinkler test: I didn't need to put water in the pipes. In fact, I didn't need pipes at all. I could've just dangled a sprinkler from a string above each light, and that little glass vial still would've exploded once it hit the right temperature.
So why did I go to all that extra work?
Simple: can you imagine how boring it would've been if water hadn't been involved?
By putting water and electricity so close to one another (without being so negligent as to not include several safety mechanisms), suddenly there's an element of anticipation and danger. You want to see what happens.
In retrospect, I could've put way more emphasis on the precariousness of the experiment. I was really wishing I'd set up another camera just to capture the beads of sweat on our foreheads during that final test. It was... intense.
So, when applicable, find a way to make it interesting. And think long and hard before you decide it's not applicable!
For further reading, I recommend Seth Godin’s excellent book on this topic.
My number-one rule at Fixing Your Video is this: never just make shit up.
Never say, “Oh, I bet my audience cares about this topic, because it's of personal interest to me!” Because they probably don't.
Know your audience. Dive into the community. What do people have trouble with? What do your clients have trouble with? The people you talk to IRL?
Addressing those pains head-on is your best way to make sure your audience will pay attention! And they get something they can actually use!
So, what’s up with that dip at the end of the graph?
I have this really bad habit of ending each episode with a lame, “That's all for now!”
This tells the audience that they can stop watching now, so they close the window before they ever see my call-to-action. Undesirable!
(Oh, and in the case of the sprinkler video, the graph tells me that me dressing up in a blue poncho and beating the shit out of a sprinkler head is freakin' hilarious – and that people stop caring once the funny part is over.)
How can I fix it without just abruptly ending the video? The serial position effect dictates that whatever's at the end will be the most remembered thing, so I'll add a bad joke, or some fun little bonus material. Unfortunately, it's not working as well as I'd like.
Maybe, once again, it's secretly a good thing. Those people wouldn't enter their email address even if they did watch to the very last frame.
I'm trying to ramp up my newsletter subscriptions at the moment, which probably means that incentives are the next thing on my list. Maybe I could end each video by mentioning a specific incentive for that particular video, like a printable cheat-sheet version of the stuff I talked about. (That's more work, but holy crap I love that idea. Except... I kind of hate it too?)
I'll let you know how that works out in a future post.
That's all for now!