I’ve decided to view our YouTube analytics favorably. I’ve made this decision in part because I have nothing to compare these statistics to. Churches weren’t exactly doing rigorous engagement-data collection before. In fact, back when we were meeting in person if you were awake during the sermon, I chalked that up as a positive. So now I have some data. The most I can do is to compare it to the next quarter’s data and see what’s what. Mostly I’ve decided to like these numbers because it shows a steady viewership. I’ll take that.
To further my analysis I looked at my top 10 most highly viewed sermons of all weeks of the pandemic. The number 1 video had 343 views. Number 10 had 132 views. Those two sermons were delivered 2 weeks apart. The middle 8 don’t seem to bear any chronological trend unless you can extract a pattern from April, March, May, March, May, June, May, July, the months in which they were viewed. I excluded sermons from the other pastor who preaches with me. Her sermons are more popular and I wanted to compare apples to apples. I examined the viewership rate at 30 seconds into the video, 1 minute into the video, and 1 minute from the end of the video. Then I looked at the viewership numbers for the first 6 days of the video’s release.
At 30 seconds into the video, 69% of viewers remain. This makes sense to me. Most of the videos I see on the interweb are by accident and I spend the first 30 seconds trying to figure out how to make it stop. These are the people who succumbed to my click-bait and now realize that I am not Adam Hamilton. At 1 minute into the video we have 60% of viewers left. By this point I’m probably reading Scripture in the video and the people watching have decided either to give this sermon a try, or that God might zap them if they turn off while the Bible’s being read. When there is only 1 minute left in the video, about 50% of the viewers are still hanging in there. I’ll take a 50% level of engagement on an in-person sermon anytime. That’s probably close to my own personal level of engagement for sermons I’ve listened to in person.
Next I looked at the views of these sermons during the week of their release. One trend is clear. The vast majority of people are viewing sermons on Sundays. Views on the next six days combined raise the total number of views by just over 19%. I think this tells me something about the in-home spiritual practice of most viewers. They are preserving Sunday as their Sabbath.
Now, let’s talk about those 19% of people who view a sermon after Sunday. These 10 sermons in my ersatz study were more likely to be viewed for longer spans of time later in the week. Out of the 10 sermons, 9 of them had their highest average view times later in the week. So while fewer people were watching, those viewers seemed to be doing so with intention. Here are a couple of examples of what I’m talking about. The second most popular sermon had a first day average view time of 6 minutes. That same sermon on day 6 had an average view time of almost 11 minutes. Sermon number 7 had an initial average view time of just over 7 minutes. On day 3 that same sermon had an average view time of just over 12 minutes.
At the very least some of my pre-COVID-19 thinking about posting sermons to YouTube has been confirmed. Going in, I thought that my audience was going to be members of the congregation I serve. I was under no illusion that this would be my chance, or the congregation’s chance, to reach thousands of people. We are getting some views that we wouldn’t normally get, but the congregation is still the core viewership. If our goal was to expand our viewership our videos would need to be very different. Next, I’m impressed and inspired by the dominance of Sunday viewership. As we continue to think about how to cultivate faith formation in the home, I think this constancy will be a factor.
What do you think? Is there something else we should be measuring? If you worship with another congregation what do your analytics tell you? How else would you interpret this data? God bless you, and stay faithful.