How to audit your own channel
The step almost everybody skips is explaining the videos that already worked.
An audit sounds like something done to you by somebody else, with a spreadsheet and a serious expression. Most of it is one afternoon, your own analytics page and a notebook, and the hard part is not getting at the numbers. It is deciding what you are looking for before you open anything.
You are not looking for faults. A fault list is easy to produce and nearly useless: it tells you what to stop and nothing about what to do next. What you want by the end of the afternoon is the opposite, the few things on your channel that already work and the reason each one works.
What is an audit actually for?
To find the pattern you cannot see from inside it.
You made every video on that channel, so you remember each as a separate afternoon. The day the tripod fell over. The one you filmed twice because of the bin lorry. Nobody watching experienced any of that. They experienced a body of work with shapes in it, and noticing those shapes means treating your videos as a set rather than as a run of Saturdays. That is all an audit is, and it is why you can run one on yourself.
How do you score a channel without scoring the videos?
By topic first, then sub-topic, and not by individual video.
Individual videos are noisy. One went out the week everybody was away, one got shared by somebody with a large following, one was uploaded at a strange hour because you fell asleep. Topics average that out. Write down the four or five subjects your channel covers, in the words you would use to a stranger rather than the words in your titles, then split each into the two or three sub-topics underneath.
Now compare each group against your own channel average and against nothing else. That is the whole trick. A figure lifted from another channel belongs to a different audience, a different subject and a different length of history. Your own average is the only benchmark that is about you.
Mark each topic and sub-topic out of ten. The number is not the output. Being made to commit to one forces you to decide whether something is genuinely working or whether you enjoyed making it, and those two feelings are harder to tell apart than anybody admits.
Which of your videos already worked, and why?
This is the step people skip, and it earns the afternoon.
For every topic that scored well, write a sentence explaining it. Not that it did well. Why. Was it the most basic question in your subject, asked plainly? Was it the one where you showed the thing going wrong before you showed it going right? Was it the only video that answered a question somebody had actually put to you?
A good score you cannot explain is not a finding. It is a coincidence you are about to build six months of content on. Once you can explain it you have what a fault list never gives you: a hypothesis you can test on the next one, in your own handwriting.
Where do people leave?
Open the retention graph on your three strongest videos and your three weakest, and read the shape rather than the number at the end. You want to know whether the line drops in the same place across several videos. A fall in the first seconds usually points at what you promised on the way in; a slow slide through the middle usually points at pacing. Treat those as the first place to look in the footage rather than the verdict, because no one shape maps cleanly onto a single cause.
Three videos with the same shape is a finding. One video with an odd shape is a Tuesday.
Which thumbnails were actually clicked?
Sort your videos by click-through rate, which is the platform's way of telling you how often somebody shown the thumbnail decided to open it, and read your own list from the top, comparing like with like: click-through moves with where a video was shown, so a search hit and a video pushed onto the home feed sit on different scales, and the honest read is within one traffic source rather than across the whole list.
Do not hold the rate against any published figure. Read the list against itself. Put the top five images side by side, say what they have in common, then do the same with the bottom five. The answer is usually dull: a face doing something, one legible object, an unusual colour. It came from your audience rather than from a category, which is why it beats any rule anybody could hand you.
Which videos brought somebody back?
Your analytics separate people who had watched you before from people who had not, and they tell you which videos gained subscribers. Those two lists rarely match.
A video that pulls in strangers is doing the finding. A video your existing audience turns up for is doing the keeping. Different jobs, both legitimate, and a channel that only does one stalls in a way the view count never explains. Mark each topic as one or the other before you close the laptop.
Why do the good scores need explaining too?
Because you cannot act on a fault.
Knowing that one subject underperforms tells you to stop making it, which is one decision taken once. Knowing that another subject works because those videos begin with the mistake everybody makes is an instruction you can follow twenty more times, on subjects you have not filmed yet.
The explanations on the high scores take longer to write than the ones on the low scores, and they are the part that changes what gets filmed next. It is also the honest limit of auditing yourself: you can score the pattern and explain your own wins, but you cannot read your channel from outside it, against its competitors, or against the analytics a viewer never sees, and that outside read is the part a second pair of eyes is actually for.
What comes out of the afternoon
One page. Four or five topics, their sub-topics, a mark each, a sentence each, and a note saying whether each topic finds people or keeps them.
Then the real decision, which is not technical at all. Of the topics that scored well, which one could you make ten more videos about without repeating yourself or getting bored? That is your next quarter, and everything else on the page is context for that single choice.
Keep reading
This is a service, and the method is written down.
Everything above came out of doing the work rather than writing about it. If you want the method instead of the story, it runs in order on one page.
Start with a free audit
Tell us where your content is now. We will come back with what we would change and what result to expect.
A person reads the channel and writes the audit by hand: a considered read typically takes three working days. That is the usual shape, not a promised turnaround. We use these details only to reply to you: no lists, no lurking.
What you will get
A fit snapshot: where your channel stands, and whether we are a match.
Two to three opportunities: specific, prioritised, yours to keep.
A recommended next step, even if that step is not us.
The audit is free and commits you to nothing: nobody follows up with a call you did not ask for.