Timeframes in Data
Looking at data with the perspective of timeframes applied to it.
2-3 years ago while I was working on some onboarding plan/documentation to help the company I was working for at the time; to onboard new employees more efficiently and with clearer objectives, I came across the idea of timeframes.
The timeframes were: first day, first week, first month, first quarter, and first year.
In the plan, it highlighted what the new employee should have achieved by each milestone. And if they missed a checkbox in the plan, at the next check-in with their manager, they would discuss why they missed it and we would adjust the plan accordingly.
Since then, I've applied this concept of timeframes to most things I do.
In this post, I want to share how timeframes have changed how I view data - specifically my YouTube analytics data.
Background
I used to just look at the view count, cos, let's face it. We look at the single metric that matters most to us - and in this case, it was the number of views.
An increased number of views generally led to an increase in the number of subscribers. And the easiest lever I could pull to increase the number of views was to push out more videos.
This was exactly what I did, I increased the frequency that I pushed out videos. I got more views and the subscriber count increased.
However, because I was constantly looking at the data.
Whenever a video didn’t perform as well as I had hoped. Or there was a subscriber who unsubscribed. I felt a bit of a panic and a need to immediately rectify it by pushing out another video.
Obviously this was not sustainable.
Data Analysis
After pushing out all the time-sensitive videos. I took some time to do some data analysis.
This was really easy with YouTube. As YouTube provides a very good analytics interface, where all the key and important information is presented to you without any need to manipulate or transform the data.
I looked into:
when the majority of my viewers watched my videos.
which videos were viewed the most and least.
and how viewers found my videos.
With this information, I can apply timeframes to the data to turn the simple data into more meaningful insights.
Applying Timeframes
Due to the nature of social videos. I wasn’t uploading videos on set schedules. Therefore, if I uploaded a video during peak viewing hours, I would get a lot more views within the first hour or so compared to a video I uploaded during off-peak viewing hours.
Knowing this, I no longer panic when I see a video performing badly during off-peak hours. In fact, I no longer panic when a video no longer performs as well as I had hoped.
The reason for this is because I now apply contextual timeframes to my videos.
I look at my videos under the following timeframes: first hour, first 24 hours, first week and first month.
And just like that onboarding plan, I have different expectations and criteria for each timeframe.
The first hour is about discovery and appeal to new viewers.
The first 24 hours is about how well the video appeals to my subscribers.
The first week is about how well the video relates to the topic in general.
The first month is about whether the video is an evergreen video.
Conclusion
Rather than go into panic mode and churn out more videos that would turn away subscribers to increase view count.
Applying timeframes to data, enables you to make more informed decisions.
It is much easier to increase the view count when you have a higher subscriber count. And it is much easier to increase the subscriber count if you don’t need to compensate for the lost subscribers.