Two years ago something I made reached about forty times my usual audience over roughly ten days. It was not the best thing I had made and I still do not fully understand why that one.
What I learned was mostly about the aftermath, which is less discussed than the event and matters considerably more.
The first two days are not about the video
The volume of comments went from a manageable number to something I could not read, let alone respond to.
A meaningful fraction were hostile, which is what happens when something reaches beyond the audience that chose it. People who found it accidentally have no relationship with you and no reason to be generous.
I spent both days reading them, which was a mistake. It did nothing for the video and a great deal to me.
What I would do now is turn off notifications entirely, read a sample once a day, and respond to the questions rather than to the reactions.
The subscribers who arrive are not your audience
This is the part that surprised me most and it matches what others describe.
Several thousand people subscribed. Engagement on subsequent uploads went up by a fraction of that, and the proportion of subscribers watching new work fell sharply and stayed lower for months.
The people who arrived had subscribed on the strength of one thing that happened to interest them. They were not interested in the channel and mostly never watched anything again.
Which had a real cost. My subsequent videos went out to a subscriber base that was substantially indifferent, the early engagement rate was poor, and performance was worse than before for about four months.
I am reasonably confident the event made my channel worse in the medium term.
The advice about capitalising on it
The standard suggestion is to publish more immediately, to capture the attention.
I did this and it did not work. The follow-ups performed at roughly my normal level, not at the elevated level, because the audience arriving for one specific thing was not there for anything adjacent.
What I have seen work, for other people, is having something already made that is genuinely the closest neighbour of the thing that spread. Not a follow-up produced in a hurry, something that already exists and is good.
Which means the useful preparation is having a solid back catalogue before it happens, not reacting quickly afterwards.
The revenue was less than people assume
Worth saying plainly since it comes up.
The rate paid on that traffic was substantially below my channel's normal rate. Broad, non-targeted audience, a lot of it from places with lower advertising rates, a lot of it short viewing sessions.
Forty times the views produced considerably less than forty times the revenue. It was a good month and it was not a transformative one.
The people I know who make a living from this do it on consistent moderate performance with an audience that comes back, not on occasional spikes.
What was genuinely valuable
Two things, neither of which was the views.
A small number of the people who arrived were exactly the right audience and became long-term viewers. Small in proportion, meaningful in absolute terms. That group is still around.
And two work opportunities came directly from it, from people who would not otherwise have seen anything I made. Both were worth more than the advertising revenue by a wide margin.
Which suggests the value of reach is in the tail rather than in the aggregate, and that the aggregate is what everybody looks at.
What I would do differently
Notifications off, immediately.
Pin a comment pointing at the two or three things on the channel most similar to the one that spread, since a fraction of people will follow it and that fraction is the useful part.
Not change the publishing schedule at all.
And be prepared for the following months to look worse, so that it does not feel like failure when it happens.
The thing about wanting it
I wanted this to happen for years before it did.
Having had it, my honest assessment is that a steady audience of people who chose you is worth considerably more than an event, and that the event is mostly a disturbance in the data plus a period of unpleasant reading.
The channels I admire were not built by one of these. They were built by a long run of things that were reliably good, watched by people who came back, which is slower and is the actual mechanism.
What the analytics looked like
Worth describing since the shape is informative and is not what I expected.
The rise was not gradual. Views were normal for about three days, then roughly doubled each day for a week, then fell away faster than they had risen.
Traffic source shifted almost entirely to suggested and home page placement. Search was a rounding error throughout.
Average view duration was substantially below my channel's normal figure, which makes sense — a broad audience with no particular interest samples and leaves.
And the geography changed completely. My usual audience is concentrated in a few countries and during those ten days it was spread everywhere, which is the main reason the advertising rate collapsed.