A few people have now asked me, both in the comment section of this blog and elsewhere, to explain how we produce our YouTube videos. I might well do a version of this on YouTube in due course, where such videos are very popular, but I thought I might also offer an explanation here. I'm aware that not everybody will be interested, but that is true almost any post on this blog.

When I restarted my YouTube channel earlier this year after a hiatus of a couple of years I realised that it was quite important to work out what I was trying to achieve by doing so. Channels that have a theme do, unsurprisingly, tend to do better than those which post random content.

The theme of the YouTube channel is a little more focused than that of this blog. Both address political economy but overall I tend to think that the YouTube channel is slightly more basic in its approach. To put this another way, I generally assume that the readers of this blog have some understanding of the continuing narrative that develops here over time, and I do not therefore feel the need to repeat explanations whenever I touch on some subjects in a blog post. I am much more likely to reiterate a point on the YouTube channel.

The secondary theme of the YouTube channel, perhaps more than is the case again on this blog, is to be educational. There is, in my experience, an enormous lack of understanding in society around economics, politics, finance, tax and accounting. I have talked about all these issues for many decades now, and thoroughly enjoy doing so. If I can use YouTube to impart knowledge, I think that is incredibly important, and is one reason why I am quite willing to reiterate on YouTube because my experience suggests that reiteration is key to learning.

Putting these themes together, what I look to do is find current topics that I can explore which are likely to attract an audience, but which then provide me with an opportunity to explore a subject in more depth and from a different angle to that which the mainstream media, in particular, will offer. If that does, for example, mean that I return to the theme of Trump quite regularly at present, that is because he does provide plenty of narratives to use to explore political economy, governance, politics, and economics and I am not going to turn the opportunity to tell a story down.

The use of stories is also, in my experience, essential to the learning process. When I was lecturing regularly I had students who sometimes complained that I told too many stories during my lectures, but whenever I asked them about which story they were complaining they could always recall not just the story but the point that I was seeking to make by telling it. That, I explained to them, was precisely why the story was told. Without contextualising the issues that I was exploring, they might not understand why I thought they were relevant, and that is what justifies the approach that I adopt here and on the YouTube channel. I am not just seeking to make political points when I make a video, and my aim is never to be party political. What I am interested in is why things are as they appear to be, why some of them might work, why others might not, and what we can do about them.

This might sound like a long winded introduction to a description of how we make videos, but it is key to that subject. I spend a lot of time thinking about the subject matter of any videos that we might make, and note down potential titles quite often as part of that process. Those that I have reflected on for a while tend to work best.

Whilst mentioning titles, we have discovered that asking questions within the thumbnail title tends to lead to higher traffic than any other format that we can use, which explains our liking for question marks whenever we can use them. Number chasing is not key to what we do, but on the other hand if this channel is to succeed in the long-term it is going to need to generate sufficient revenue to cover more than one salary plus equipment costs with, perhaps, some left over for me, and so we do have to take that factor into account.

Having decided on the title, unless there are some very obvious facts that I need to check before filming, or some numbers that must be worked out before starting to film, little more preparation is done. The videos are never scripted. We do have a teleprompter for use on educational videos, which we are also making for work linked to the University of Sheffield, but for the channel I find using one really unhelpful and lacking in spontaneity. It is, therefore, the case that once we start to film, Thomas turns the camera on and we see where we go.

Very occasionally we have recorded what feels like a completely clean script. It does not require any editing. More often than not, though, I do make a mistake of some sort whilst recording, even if it is rare that we stop recording as a result. I can usually hold the last two or three sentences that I have said in my head, and if I make a mistake, I simply go back and start from one of them again. As a result well over 90% of all the videos are made in a single take.

For the technically orientated, we use a Sony A7iv camera for most of the recording with a Sony 24 to 70 GM2 lens usually set at around a 30mm focal length and an aperture of F2.88 to get the blurred out background.

We use at least three studio lights with the result that I am wearing make up on all the videos that you see, because otherwise my skin would look quite blotchy, as does anyone's in a studio lit environment. Applying this is something I have had to learn.

We tried a number of set-ups for microphones, and it probably took longer to find the right solution for sound than it did for visual purposes. We use Rode Lavalier Wirless Pro microphones, with 32 bit floating sound, but use other options on occasion. The normal set-up is recorded straight into the camera, but with a separate backup audio file always being recorded, just in case. So far we have very rarely needed to use it. Careful sound setting on the camera usually avoids that need.

At present we do not record in S log 3 but are considering doing so in the future to improve lighting control, and so the colour on videos.

A recording session usually involves making three or four videos, one straight after the other, but not all of them are necessarily used. If either Thomas or I feel uncomfortable about the quality of something that we have recorded, which is usually the consequence of me going off topic, then we will always abandon a video. However, in that case we usually return to it and make it again in a subsequent session, when they very often work quite well.

Recording sessions are usually done fairly early in the morning, but if there are a big events, like speeches or budgets, that we know will attract interest, they slot in whenever that is possible.

As is obvious, I always record in a black polo shirt. Right now these go on to record and come off straight afterwards. We don't keep the house that warm. This singular choice might seem boring, but I note that most male YouTubers do seem to work on the basis of consistently wearing very similar clothing in every video, and it seems to work with audiences. It also saves time that would otherwise be spent wondering what to wear, which is not something I am greatly interested in.

For me, most of the work is done once the camera is turned off. Thomas then takes control of the process. At present we use both Descript and Camtasia for video production. In particular, Descript is used to create a transcript of what I have said, which I then check.

We have developed a colour coding system for the edits. Because we always put subtitles on the videos because we are aware that many viewers like to use these, and that those generated by YouTube are not as good as the ones that we can generate in Descript, most of the edits are grammatical. Descript does not always get punctuation right, and a lot of of the changes that we make are to ensure that such things are corrected both in the subtitling, and in the subsequent transcript put on this blog. I make the edits in Word. Thomas works out how to reconcile the needs for a good transcript with the flow of the written word. Sometimes we let them differ because that is necessary, but they are closely related.

Every now and again a word has been omitted from the transcript, or on occasion I simply forgot to say it. We ignore how often this happens in conversation. Video is not so forgiving. These require some care to ensure that they are managed, which is a process that Thomas is largely responsible for. If I think a word is missing I indicate that fact in my edit, and we can usually generate a single additional word if it is required using the AI facilities in Descript. Anything longer than that is problematic, so I have to get things right as far as is possible in that first take.

It is much more likely there will be some things to delete from the video. In particular, there will be duplication if I know that I've made a mistake. That explains the small breaks that happen in most videos. More frequently, but far from unknown, I will simply decide that an argument that I included in the video did not add as much value as I hoped, and on occasions I will delete a hole chunk of the video from the final cut. It is, quite literally, like editing a paragraph out of the manuscript.

Once this edit is completed, Thomas then adds graphics if that is appropriate, corrects the sound quality if that is required, rebalances the colour if we got the white balance wrong, for example (although we are getting much better at checking for such things now), and then prepares the video for publication. I tend to write the video titles and the blurb that goes with them, but Thomas also contributes at this stage. He sometimes sees angles that I have missed and if so he will make changes.

Thomas (who is known as Tom by everyone in the world, excepting his parents) also designs the thumbnail . Most people think that thumbnails are very important on YouTube, and they spend a great deal of time designing that opening image that is used to attract someone to watch the video. This may be something that we need to return to, and in the early days of this year when we were started remaking videos, we certainly got these wrong most of the time. At present we have settled into a formula, which is a picture of me, the video title in a box against a burgundy background, and the background image which is a picture that either Thomas or I have taken somewhere on our travels. The approach is formulaic, and it might need improvement, but we also think that there is some merit in the predictability, and we are not going for neon p-colouring, flashy graphics, or other such things because we know that these are unlikely to appeal to our audience, which is older than the YouTube average.

After that, final checks take place. Thomas is a bit of a perfectionist (which is incredibly valuable) so always tells me if I need to doublecheck something. The third member of the team on this occasion is my wife, Jacqueline, who is involved in quite a lot of editing on the site, but only at the post production stage with videos.

Then it is publication time. The main version is always that which goes on YouTube, but because an audio version is incredibly quick and easy to produce, that is also published on Podbean, which then sends it to about 10 podcast channels. If I'm honest, we have very little idea as to what the traffic is on these. We know a lot more about the versions that also go on Instagram and TikTok, which are made from the YouTube version in a vertical format. So far Instagram has not worked very well for us, but TikTok can. Videos either do very modestly there, or very well, but in the last month we have had more than 1 million views on TikTok, which is a bit over 50% of the number on YouTube.

It would be great to have time to respond to the comments that are posted by people on the YouTube , of which there are usually hundreds in a day, but I simply do not have the time to do so. What is refreshing is that the vast majority are positive, and our like percentage is very high. So too, overall, is our viewer retention rate i.e. the number of people who actually watch the video to the end.

It would be good to get more subscribers. Only about one in six people who watch the videos actually subscribe to the YouTube channel even though it is completely free to do so. That is something we will be working on.

Finally, we know that some videos will always do better than others. Referring to Rachel Reeves, Donald Trump or the Bank of England in the title will always boost our traffic. Kier Starmer does not help in the same way. Why others work is harder to tell, and something we need to learn more about. In particular, I need to spend more time finessing the YouTube descriptions. That said,  I sometimes just want to make a video on a subject that I know will not be particularly popular and that does not stop me from doing so.

Where do we go from here? We definitely want to increase production quality, and diversify the way in which we make videos. So, for example, I would like to do some podcasting to include other voices on the channel. We also want to revisit the use of shorter videos to explain economic topics as short introductions to subjects. This would, in effect, be a YouTube glossary. This particularly appeals to me, but YouTube is going through a weird process of changing the definition of its short videos at present, and until they work out what they're doing, we will not be doing this.

In addition, if I do, as I hope, free out some of my other work time I would like to make some longer form videos as well, explaining subjects in more depth, possibly using more graphics as a way to explain more complex issues, which approach might, I hope, be of interest. That does, however, depend upon my successful scaling back of my academic obligations.

In all, this process has been a massive learning curve for both me and Thomas, which we have both enjoyed more than we expected. We are now committed to continuing to produce videos, and the decision to monetise our channel, which was difficult to make, does ensure that this is possible. We have also been approached about the video sponsorship possibilities but this is, again, a difficult topic about which we have not yet formed an opinion, but which might allow for greater investment in what we are producing. That is to be considered in due course, and would, very much, be a matter of proceeding with care.

What I am pretty sure about is that this exercise has worked, and in many ways better than I dared expect. It continues to surprise me that so few of the views on the videos come via people reading this blog. It appears that people who like blogs like transcripts, and text, but not videos. I expected a heavy overlap when I started this process, and have been very surprised to find that is not the case. However, what that does mean is that almost all the video traffic comes from new audiences, and if that helps the process of change in which I am most interested, that has been really rewarding. It is also why I will keep making videos. There is a lot more to learn and do with video, and I want to do a lot more.