188. How is AI disrupting your industry?

--- 

We love hearing from our readers and listeners. So if you have questions about the content or working with us, just get in touch on [email protected]

Say hi to Sophia on Twitter and follow her on LinkedIn.

Following us on YouTube, Facebook, Instagram and TikTok will make you smarter. 

Interview Transcript

Sophia Matveeva: Hello, Will Page and welcome to the Tech For on Techies podcast. I am so happy to have you here.

Will Page: It's a joy to be here and thank you for your fantastic work and for reaching out and inviting me as well.

Sophia Matveeva: Well, also happy new year because we're recording this. It is still kind of early January and we were just discussing how fabulous it is to spend New Year's Eve in Scotland because You guys in Scotland, you really know what you're doing.

So you're just saying that you guys get two days off to recover from New Year's Eve. Is that right?

Will Page: Yeah. The rest of the world has January the first as their public holiday in Scotland. We party so hard we get the second as well, which I think is a very. It's rare for the government to be generous, but in Scotland they're very generous.

They give you 48 hours to recover from your hangover. And in addition, we have this tradition in Scotland of not just partying incredibly hard for New Year's Eve, Hogmanay as we like to call it, but on the 1st of January we have the loony duke, which is where you go swimming in the North Sea. Now, if you have a terrible hangover and you swim in the North Sea, I can guarantee you this, A, your hangover will disappear, but B, it will be replaced by something far more painful, which is you come out looking rather blue.

So, You know, we have like swings and roundabouts the way that we approached Hogmanay, but for your listeners, if you are at a loose end for 2425's New Year's Eve celebrations, please consider a trip to Edinburgh. It's a fantastic five day festival and it really does take the whole notion of ending the year on a good note to a new level.

Sophia Matveeva: So there are so, I mean, there are so many things I love about Edinburgh. It is gorgeous. The education system, you know, Edinburgh University is wonderful. New Year's Eve at Edinburgh Fringe.
 

Will Page's Journey into the Music Industry

Sophia Matveeva: And so, Will, what are you doing in London?

Will Page: Well I was in Edinburgh up until 2006. And back then I had this rather bizarre Batman lifestyle.

You know, by day I was a government economist wearing a charcoal black suit, a blue shirt, a red tie, and doing really tedious subjects like local income tax reform. Send your audience to sleep with that one. And then by night I developed this sort of moonlighting career as a music journalist, writing for a very hip publication called Straight Notes Chaser.

And you may have heard of the DJ Giles Peterson, it's his publication and covering Philadelphia hip hop.

So you, my sister used to call me Batman because you had this kind of alter ego. And during those four years, I knocked on every door I could to say, would the music industry hire an economist? They never had one before.

And every door I knocked on said, no, I got interviewed by Sony in New York. I pointed out that Sony was creating the hardware to rip, mix, and burn music, at the same time suing the consumer for ripping, mixing, and burning music. You know, it's just a bizarre period. Napster had upended the business, the entire industry was in freefall, and I wanted to be the first economist to try and apply my discipline to turn it around.

Then I got my lucky break. Long story very short, I picked up a newspaper on a bus, read an article by the author, wrote to the author, correcting him on some mistakes in the article. He called me back. He was the chief executive for the Performing Rights Society and said, I want to make you the chief economist of the PRS.

And that was when economics entered this industry. Had I not sat in that bus I wouldn't be speaking to you now. It's a very bizarre story. We still keep the original newspaper cutting on our walls here. But literally, the message for your audience is never stop knocking on doors because one of them might just open.

And in my case, that door opened all far and wide.

The Importance of Initiative and Creating Your Own Job Description

Sophia Matveeva: I mean, there is so much to learn there from kind of a resilience and career perspective. And also, you know, hustle, side hustles are all the rage now. Yeah, you have this very serious job that your parents were probably really, really proud of, you know, government economists.

Oh, that's really good. And then you had your side hustle, which actually then turned into this really, really interesting career, which is kind of what has laid the foundations for today. And also what I see that is just being proactive. And I often, you know, get emails from people saying, Oh, you know, I really liked this episode, but do you have more episodes on XYZ?

And I'm like, well, you could just scroll down yourself and have a look. You could use Google to find out you don't have to email me to do this. And so I, I always just see this difference between people who are essentially proactive in it. They see an article in newspaper, they write an email and the people who are like, I don't know, just rescue me.

Like if you wait to be rescued, it's not going to happen usually.

Will Page: I mean, I get hired people saying, Hey, can you help me get this job at Spotify? And my first question is, have you actually applied yet? No, I was going to speak to you first. I now know that Spotify shouldn't hire you because you didn't bother to take the initiative to apply first.

But, you know, to build that out a little further for your audience's benefit, if I hop, skip and jump to the final words of my book, I say, don't wait for your job description. Create your job description. I've never looked at a job description in my life. I know what I can bring and I tell them this is what I think you need and I want to be the first person to deliver it.

And if you disagree with me, fine, no, no offense taken, but rarely do job descriptions describe what a company actually needs. And if I draw one lesson that I took from government working in the sort of quasi treasury function of the government under the chancellor of the exchequer, Gordon as we call them back then, he used to always say, As an economist, you need to strive for evidence based policymaking and avoid the temptation of policy based evidence making.

Look at the wordplay there. Evidence based policymaking. And I've used that in my career music. Every day I wake up, I'm thinking, you know, how are people framing their decisions based on policy deriving evidence to justify their belief versus my job is to swing the other way and say, well, let's look at the evidence and now let's devise a policy or let's devise a decision.

So. That, that would be, that's a great example of a simple rule that you got hammered into you in the boring world of the civil service that you can apply every day of your life in the sexy world of media, music technology as well.

Sophia Matveeva: I think your earlier point about creating your own job description, I think it is actually super relevant for the digital age because there are so many jobs that are really prevalent today.
Like I'm thinking product management that literally didn't even exist as a job. About 2025 years ago, whereas now it's, you know, it's a really popular path for saying, you know, the top graduates of the top business schools, and I suspect that this is going to carry on changing. So for example, you know, there is now a saying that it's not AI that's going to take your jobs.

It's people who know how to use AI are going to take your jobs. And so. Our job descriptions in every single industry are essentially being rewritten by us, and you can't really wait for people to do that for you because, essentially, I think. In a lot of cases, we're all kind of making it up as we go along.

Will Page: Yeah, a fake it till you make it approach to job descriptions.
The Impact of Digital Disruption on the Music Industry

Sophia Matveeva: Tell us about Spotify and the digital disruption in the music industry, because I kind of remember it vaguely. Like I remember, am I going to get sued for this? Anyway, I remember downloading stuff illegally at college. I remember getting notifications from the University of Chicago that, you know, if I carried on, if I carried on, you know, downloading, like whatever I was downloading, bad things were going to happen to me.

And, you know, I would feel bad like for a day and then I would just carry on doing it, you know, remember the first iPad. So I, I do kind of have a recollection as a consumer, but. It was a long time ago. And, you know, I don't, I don't really remember much. So can you tell us how, what was the music industry like before digital disruption?

And then what happened?

Will Page: Sure, it's a really well framed question. And let's unpack it in two stages. I would argue that if we look at the past two decades, for the sake of around numbers, the first 10 years was spent fighting change.

And the second 10 years was spent embracing change. And just to quickly pull back on the purpose of my book, Tars and Economics, retitled Pivot, is to help others avoid that 10 years of fighting change, avoid the suffering, and get straight to the recovery.

That's all I want to do when I teach economics. But if you go back to those first 10 years, let's start at the millennium, where life is easy.

Sophia Matveeva: Let's start with where they were, yes, because I remember you've got this great anecdote of Helicopters and private jets, and Yep. and oodles of cocaine, so

Will Page: Helicopters, jets, and cocaine.
And I remember record label execs having sushi flown from Japan to New York for lunch. I mean, that's how excess, excess, excess it was. To bring it back to street level, though, and this is a very true story, fact checked, it was so easy to sell CDs at the end of the millennium, that record labels would sell it to record shops by, and I want your audience to listen closely here, By the weight of palette.

So I say to Sophia, Hey, Sophia's Record Shop, I've got a palette of CDs here. You wouldn't say to me, What's on that palette of CDs? Is it Shania Twain? Is it Coldplay? You would say, How much does it weigh? And I'd say, About 30 kilos. And you'd say, I would give you this much. Now for all the students studying data science right now, let me just be clear, there's not much data science in a palette of CDs sold by weight.

But as much as we love. The business then was worth way more than the businesses now, with or without inflation adjustment. So they had to be doing something right. But as with any industry, when you get it that easy, when Robson Jerome can sell five million CDs to the British population and they couldn't sing, dance, perform, they had zero talent.

You know, when Texas works with Method Man and it just doesn't make any sense, but it still sells millions. When you have it that easy, You get fat and you get complacent. And on May 1999, American students woke up to a device called Napster that took the word copyright, copyright, and removed the right to control copying.

And it upended the concept of intellectual property and the business of the music industry overnight. All of a sudden is way more convenient to steal all the world's music than it was to purchase a CD in a plastic case that if you remember those things used to break your nails as you try to open them.

So roll forward. And what did we see during those piracy years? We saw the industry spend millions on fighting change and losing billions in revenue and three piracy services, which none of us have ever used, correct? Were Kazaa, μTorrent and the Pirate Bay. Now, what was interesting about these three services is they all came from the country of Sweden.

Sweden was producing the most powerful piracy applications, especially μTorrent. My god, that was fast. But along comes Daniel Ek, who says, wait a second, they're not stealing copyright because they want to stick it to their man, or they hate intellectual property laws, or they want to be a rebel. They're stealing because it's more convenient.

The Rise of Spotify and the Recovery of the Music Industry

Will Page: And I'm gonna build an even better piracy app, and I'm gonna go to the record labels and ask for a license. You know, the action is the same. Now it's licensed, it's no longer illegal. You can bring a horse to water, or water to a horse. Let's take the copyright industry to where the consumer is, instead of asking the consumer to go to where the copyright industry is, and launch Spotify.

So, by that point, the business had more than halved in nominal terms. You know, firing of headcount everywhere. The place was on its knees. And all of a sudden, the door, after knocking so many times, began to open. Which is, customers didn't want to pay for ownership anymore. And Suing them isn't going to make them pay for ownership anymore.

Suing you or sending you threatening letters at the University of Chicago isn't going to make you go back to Tower Records and buy CDs. And by the way, Tower Records was closed, so you couldn't. That's a very important point. But would you then pay for convenience? Stop selling ownership, stop selling convenience.

Roll forward to today, you know, as Spotify started accelerating around a while, 2009, 2010. America launched in 2011, went public in 2018. Hilarious story when Spotify went public on the New York Stock Exchange. A tech unicorn from Europe lists on the New York Stock Exchange. They flew Swiss flags outside the building, not Swedish flags.

Oh my God. We can give you the image to post on your website. It's just a very funny moment. I was with the team when we had to phone them and say we're a Swedish company and the official apology from the York Stock Exchange said, Hey, Spotify, everyone makes mistakes. And let's face it, Switzerland's got great chocolates and they swapped the flags around.

But you have this story now where Rolf Ward from staring into the abyss between 2000 and 2010, we now have over half a billion people on planet earth. paying on a monthly basis to access music. And it's voluntary to pay. Me and you could never pay for music ever again and continue enjoying it. It is voluntary to pay.

And when you look at the success of what music's achieved in this first to suffer, first to recover story, The important thing for your audience is to press pause, look around and think about how many other industries are staring at a Napster moment. The newspaper industry is staring to the abyss right now.

What could they learn from our journey? And that's where the story really gets its spice.

Sophia Matveeva: Isn't the newspaper industry already sort of evolved because we can look at the New York Times or the Guardian, you know, they were basically driving themselves off a cliff with just ad supported media. And then they both experimented with lots of different models and so New York Times is now really subscription model.

And the Guardian has somehow miraculously managed to make donations work for it. And so, and you know, the Daily Mail is thriving. Yes, there are organizations like Mashable and Buzzfeed that essentially couldn't support itself by advertising or subscription. And, you know, they got absorbed into other entities.
Is it, why, why are you saying that newspapers are still going through that moment?

Will Page: I think newspapers are papering over the cracks. You're right to point out to those successes, but let's look under the hood, as we say, and have a closer look at what's going on there. Firstly, the Guardian has a voluntary tip jar model, by the way, not dissimilar to what Radiohead did in 2007 with the album in Rainbow, something I documented and worked on in my book.

That kind of removes the need to commit a payment because it's voluntary. Can you run a business of that? Or do you run a business of the Scott Pension Fund that's been subsidizing the Guardian all along? Is that a standalone success story or is that a subsidized success story? Question mark. I'm not saying for or against, I'm just saying we need to look a little closer.

The New York Times, you know, they have. 8 million subscribers. They acquired The Athletic for 500 million. There was a joke, which is, did you hear the New York Times bought The Athletic for 50 million? No, Will, it was 500 million. Sorry, that was next month's headline because the company was running out of cash.

But even still, you've got huge challenges there with churn and subscription hopping. When people join Spotify, they stay and they stay for the rest of their life and the lives that they bring into this world with family plan. When people join newspapers. It wouldn't be an exaggeration to say you're good if one of two stay.

That is churn on an annualized measure is hovering between 40 and 50 percent for most titles. And with the athletic, as an example, there's a lot of churn hopping. This is not just the athletic, it's across the board, but I will join for this discounted premium rate to get me back into the service and leave before my year is up and join again.
people don't do that in music as well.

Sophia Matveeva: And then the final thing was actually because, you know, first there was Netflix and everybody had Netflix and that was what you needed to have. And now there are so many different things that I, I'm sure I do the same thing as lots of. listeners I would subscribe to basically watch a show.
And then once that show is done, then I end that subscription and then there's a show on another service. And I'm like, okay, I'm going to go and binge watch that. Yeah.

Will Page: That's key. And that's again, like the headlines might be quite persuasive, but underneath there's a lot of mush going on. And that's, that's a really, really important point of just how people hop from subscription to subscription and other platforms, but the loyalty that they give.

Apple, Amazon, YouTube and Spotify streaming services is phenomenal. That's a genuine success story. Churn of music is so low, churn elsewhere is so high and we have to explore the reasons why. But the last thing I want to say about the newspaper industry, it's really important, is the numbers aren't that big.

I just told you and your audience that half a billion people are paying for music. Do you know that Joe Rogan in America gets to more Americans with his three hour show two, three times a week than buy newspapers. And Joe Rogan employs four people, newspapers employ thousands of people, have fixtures and fittings, health and safety, unions, God knows how much regulation to cope with.

In terms of the business of getting my voice to an audience, it could be the New York Times who are authoritative voice with copy editors and fact checkers, if they haven't fired our fact checkers that is, or it could be Joe Rogan. But in terms of battling for attention, I really want to stress the newspaper industry is relatively small given the rest of the distractions that they're competing with.

Sophia Matveeva: And so now that we've discussed the music industry and people can kind of remember what happened there.
The Future of AI in Various Industries

Sophia Matveeva: Can you tell us now that you know generative AI has taken the world by storm or the beginning of 2024. What's your prediction? Which industry is going to get really, really disrupted by it? And why do you think that is?

Will Page: So the way I approach the AI debate, and it is the dinner table topic du jour at the moment, which is, you know, to prepare, it's a Friday night, we're recording this podcast to prepare audiences for their dinner parties, is to think about how I was first taught what capitalism was. Which goes back to a Federal Reserve Governor called Alan Blinder, who once said to me, Capitalism is when you employ a gardener to cut your grass because you can do something more productive with your time.
the comparative advantage. Don't do what you're good at, cutting grass.

Sophia, do what you're great at, teaching students. So it's, his point was, you might enjoy gardening, you might love the methodical process of cutting grass, but the day that you can do something more productive with your time is capitalism in action.

I love that analogy of you can relate to it, it's non technical, just hits it on the nose for me. So then you look at AI, and I just, I encourage your audience to think about AI as just one huge F off gardener. That is, it's going to do all the donkey work for us so we can do something more productive with our time.

And all the fear and clickbait headlines need to be calmed down, cooled off a little, and just appreciate how this thing could boost productivity. There's one profession I think AI could really make an impact in. It's the legal one. I think lawyers, lawyers work with words and meaning of words. AI can work with words, meanings of words, teachers, doctors, they work with people with policies, but lawyers is a word business.

And I think that AI would look at the amount of donkey work. in the legal profession, the amount of heavy lifting, the amount of administration in the legal profession, and literally drive a tank through it. It could be your 24 seven, always on, most read lawyer of all time, that is free. So I think if you look at, there's a great line in my book, which is if you look at a legal department in an org chart, Quite often you see the job title assistant to assistant.

You'll never see that anywhere else. You might see EA to the boss, but in legal you have assistant to assistant. When you see words like assistant to assistant, you have to think AI could take out the first assistant. And secondly, AI could probably take out the second assistant. And there's companies out there that I can point to as one called Fastcase, Damien Real, who'd be a great guest for your show, which is basically reading and training on every legal case there is in the world.

And acting on it. So let's say we have a dispute and we can resolve the dispute on Clause A, whether there's been lots of litigation, or we can resolve the dispute on Clause B, whether there's been no litigation, AI would say, take Clause B because you'll get to a dispute resolution without having to fight.
Lawyers wouldn't say that, would they? Let's say, let's go down to dispute A, charge you a fortune, fight like hell, and after two years, settle. So think about the donkey work that AI can achieve and the cost efficiencies it can produce for the client. It can democratize access to law and serve law at far less cost to the client that needs it.

Sophia Matveeva: I've actually heard lawyers from big law firms say that clients are asking for this because essentially, you know, we all know about chat GPT. And so clients are saying, okay, where can you use it to cut my cost? And I actually saw a case study.

There's this massive law firm called White and Case, and you know, they represent Meta, WhatsApp, you know, and so basically that SoftBank, the big, big, big tech tech companies in DCs, and they've been experimenting in, in one of their teams with using AI, and they actually trained an AI to essentially be an assistant, and they, I might be getting the numbers wrong, but it's, From what I recall, they said that essentially this research work that traditionally takes about seven hours has now been cut down to about two and a half hours.

So, You know that this is a really this is a huge, huge reduction in that kind of grunt work, and it's a huge reduction in cost. And it's interesting because literally white in case put it on their website and they're talking about it. And so I find that an interesting case from two points of view that It is happening.

And, you know, it is happening because of market pressures, right? It wouldn't be happening if clients weren't like, okay, how can we, how can we use this stuff to reduce my legal bills, but also it's happening.

Also, the fact that they're talking about it and that they are saying, yes, we are this big law firm.
We are still specialists. You still, you need, you know, you need us as humans. But we are using these technologies to essentially serve our clients better and, you know, frankly, to reduce costs. I think it's interesting that market pressures. Are already showing themselves in at least some of these firms, but also not in all of them.

I was speaking to a lawyer yesterday who told me that and this is not a white in case this is a, this is somebody entirely different. He told me that some of his colleagues still use word perfect. To start, I don't even like, how do they get it. Where do they download. I don't know.

Will Page: Yeah, no, it's a big one.

I think the expression for your audience to hold in their heads when we discuss a disruptive technology on one side and an incumbent establishment on the other side, is do turkeys vote for Christmas? They don't. And there'll be a lot of resistance to embracing this technology because turkeys don't vote for Christmas.
Equally, there's a first mover advantage in the first turkey that does.

That is, if White and Case can say, we're fully AI automated, our fees have dropped by 70%, they could clean up a lot of market share. And that's a really important thing to think about. And we saw it in music in terms of the first record label to license Spotify was another first mover willing to give away control of their copyright to instead price access.

There's this whole tension where something clearly makes sense to the consumer. doesn't make sense to the producer, and there's going to be a break point where it snaps and all of a sudden the floodgates open. Yeah, but I just reiterate that company Fastcase is one of many in this space, but for your audience have a look at what they're doing and scale it.

That's just the US legal system. What about international law? What about Dutch law? What about Scots law, which is different from English law? You know, why don't we just get machines to train off the law so the consumer can access it? Is that too much to ask? Common sense to me.

Sophia Matveeva: And so aside from law, what other industries do you see are going to be really, really disrupted by this new technology of generative AI?

Will Page: Well, I think what we discussed in law was back to Alan Blinder's lesson that he gave me when I was, you know, a young wild living student, which is We're going to have a useful gardener, you know, we're not going to get rid of lawyers. We're just going to have a gardener to cut the grass of law. That's all we're trying to do.

And I think if I pick on two others, I think education and health are worth investigating and the Kahn Institute for Education, first of all, which is we know of the two sigma problem in education. If there's two students in the class, Will and Sophia, and Sophia gets one to one tutorials, there will be a delta in her performance, like for like compared to mine.

One to one tutorials help.

The Potential of AI in Education

Will Page: There is a debate about whether class size is actually a variable in education. Some people say it is. Other people say it's got nothing to do with equality. But additional one to one tuition definitely helps. Why can't a ChatGPT large language model be that one to one tutor?

For everyone, rich and poor, developed and emerging markets. You know, there's an ambitious North Star goal for us to aspire to. And you think, well, it's kind of doable because the way that you would deliver that one to one tuition presumably is app based. So the delivery mechanism that hindered past technologies has been removed now.

The adoption time for this would be the send timer on your phone which says your app is uploaded. So I think enabling everyone to have one to one tuition is a nice North Star goal to think about how this could help education.

AI in Healthcare: A New Frontier

Will Page: Flipping it over to health, I think, you know, you look at the donkey work, the grunt work, as you described it, that's involved in health care.

You know, how much of patient checking, symptoms diagnosis, could be done on AI as well? Hospitals are stretched, they're going to get increasingly stretched. How much of that work, that grunt work, as you described, could be offloaded to AI to become your gardener so you can do something more productive with your time?

Pre emptive healthcare, for example, and I think those two, I don't think, to wrap it up and put a pin in it, Sophia, I don't think we discussed the benefits of AI enough. It's too much fear and clickbait headlines, and we rarely discuss the benefits to public services. I just want to switch the dialogue in this conversation to what could I do for public services already stretched, lacking technological advancement to make the world a better place.

Sophia Matveeva: You know, it's really

AI in Mental Health: A Case Study

Sophia Matveeva: interesting what you brought up about AI, because I literally have a client who I'm hoping will join me on this podcast. They run an AI company and this AI company works with mental health clinics and essentially they help therapists take notes. So take notes from sessions and also they help therapists spot whether somebody in a group therapy session is in trouble.

So, you know, If there's a therapy, if there's an outpatient clinic and there are group therapies, and you know, somebody is recovering, but some, and somebody else is actually on a downward trend.

Yes, theoretically, a doctor should be able to spot that and you know to put somebody on suicide watch, but in reality, there's human error, you know, if you're a doctor and you've got a group of people.

Unfortunately, it is all too easy. to miss that somebody really is in trouble. And essentially their system in the first three, in the first, I think, six months of being just trialed in a, in a clinic had already prevented three suicides.

So they literally said, okay, watch out for, you know, these three people and.
Which the doctors wouldn't have spotted. And actually, yes, those people were found to be in serious trouble and essentially lives were saved. And you know, this company, it's still, they still haven't yet gone out to everybody in the public saying that this is what we're doing.

So I can't disclose more, but I think what they're doing is really, really wonderful.

Yes, it's technologically really interesting because essentially what they need to do is they need to combine data on what symptoms, what behaviors do mentally really unwell people. exhibit with technology.

So it's really interesting technologically, but also what they're doing, you know, they could have chosen to use this technology to analyze stock markets and analyze what chief executives are saying.

And, you know, to see if somebody sounds a little bit nervous and then sell that insight to traders and, you know, Arguably, they could be making a hell of a lot more money doing that.

But actually, they decided that like, no, this is where we want to start. Because for us, Like, we still believe we can make a lot of money saving human lives, and this is more important for us.

So it's a really beautiful story.

The Challenge of Measuring Preventative Actions

Will Page: Now, in my book, I have a chapter called Big Data, Big Mistakes. And one of the challenges I deal with there is when we discuss preventative actions, how do you measure what you prevented? And if I can just give your audience a very good rule of thumb for this, how do you measure zeros instead of ones?

It's, I call it the crime stats dilemma. So we both live in different parts of London. You're out West, I'm in the North. Let's say I say that in your borough of West London, crime stats have increased. What's your gut reaction to that headline? I

Sophia Matveeva: would say it's not true.

Will Page: And then double lock your doors

Sophia Matveeva: I would say, you know, people's car is being stolen. I don't know. It's West London.

Will Page: So a lot of people would say that means crime is on the increase. Intuitive headline. Crime stats on the increase.

There's more crime. Or does it mean we're getting better at catching criminals?

Or does it mean we've changed the definition of crime? Or even have we created a new crime reporting app that's really popular, which makes it easier to report crime? It does not necessarily mean what you first Your first gut instinct is, which is, oh my God, crime's on the increase in Sophia's part of West London.

It doesn't mean that. And I think when we look at preventative actions, like this commendable company that you're working with, it's just that extra challenge of proving the benefits are harder when you're trying to measure zeros instead of ones.

Sophia Matveeva: Well, it's like Minority Report. Remember that movie? when it was all basically in minority report, there would be estimates of who is going to commit a crime.

So they would go, that person would be arrested basically before they're going to do something, which, you know, In a way you're like, well, this is great. Like, don't we want to arrest the person as they are like planning a murder? Like that would be really good because we don't want the murder to happen.

But then inevitably you know, in the movie, watch it. It's, I think with Tom Cruise, it's actually a really like good, thoughtful Tom Cruise movie, which is really rare. And. you know, there's this dilemma of like, okay, most of the time the system works.

So in the movie, most of the time the system works and most of the time it prevents crimes.

But sometimes it doesn't, and in the times that it doesn't, a person who, you know, may have been, like, I've been definitely angry with people and thought like, God, I really want to kill them, but it doesn't actually mean that I'm going to murder them. It just probably means that I'm, you know, going to put the phone down and not speak to them for 24 hours.

Will Page: What a beautiful anecdote. Well, A, I know what my Friday night viewing is going to be. So I'm going to go back to that classic. But that's brilliant. That's a really, really good illustration of. Yeah, I love it when you can see these real life dilemmas playing out in cinema. It helps bring the story to life even more.

The Impact of AI on Business and Society

Sophia Matveeva: So as we are wrapping up, I would love your thoughts on what people on the business side, what they can do or what they can learn to succeed. In this world that is being changed by technology and by AI specifically. And the reason why I ask this is because I find that, you know, people who listen to my podcast tend to be very high achieving people and very high achieving people are really good at punishing themselves.

And because they're really good at punishing themselves, they tend to be like, Oh, I don't know anything about technology. Let me go and you know, learn C plus plus. They don't know what C plus plus is. They're like this, there's this course, this is what I'm going to do. And then they take it and then they end up being like, okay, I still don't know what it is, but I can do it.

Yep. So my aim is to help those people learn so they can succeed. So they can really be knowledgeable in companies going through digital transformation, but also, you know, to direct them towards something sensible, as opposed to something just pointlessly hard. So what would you say to these people?

What should they learn? What should they do?

The Importance of Consumption Analytics

Will Page: I've got two lessons I'd like to convey in this podcast for that audience. And the first is to do with consumption analytics. The second is to reframe how you think about attention.

There are high level points, I won't go into technical jargon here, but you can bring it back into the technical jargon of what you need to learn in terms of coding skills, data science application, to achieve these goals.

But let's, let's just take example number one, which is consumption analytics. I mentioned at the start of our conversation how record labels used to sell CDs by the weight of pallet to record shops back when it was easy.

And that was a transactional business. Sophia bought my CD. Did she open to it? Did she play it more than once?

Did she play all the songs more than once? We didn't know. There was an upfront transaction and a vacuum of no knowledge downstream. Now we have streaming.

We don't have that upfront transaction. We have complete knowledge of everything that happens downstream. Streams grow over time. Sales decay over time for a reason.

If you like that album, you'll listen to it more. I worked with the band Imagine Dragons on demonstrating this. But the story I tell at the start of the book is so applicable to the rest of business.

The first thing I was asked to do at Spotify, I was an early joiner there, was create a chart of the most popular album for 2012.

And the most popular album was an artist called Gotye, who had a famous song. Can you remember what the name of the song was?

Sophia Matveeva: Oh, it's in your book, but I, this, I love this example, but I forgot the name. It was the one I used to know. Somebody that I used to know. You're someone that I used to know. Oh yes, I have danced to that many times.

Will Page: Hollow song. It sounded great when you played it through your old Apple Mac, which is how it was produced for. Now, the question I always ask people is, we now know who Gotye is. We now know what his famous global hit was. Can Sophia name a second Gotye song? Yes or no? Absolutely not. I asked his publisher and he said no, and that made me really worried.

So I've got 100 percent track record here, which is everyone thinks Goethe was a huge global artist, 2011 12. He had this huge song, but they can't mention or something. I did this chart which showed for the best album that year Goche was number one and Lana Del Rey was number eight. Lana Del Rey we can name lots of her songs Blue Jeans, At The Races and so on and so on and so on.

So then what I did was I said well hold on that's just simply lumping the album together. Let's do a median rank. There's 11 songs in an album. Let's take the sixth most popular song and re rank this album chart that way. We're asking how much of the album did you go into as opposed to what was the most popular album.

Remember CDs we knew nothing, now we know it's something. Very quick statistical exercise. End result, Goche went from the most popular album to number 57. Lana Del Rey went from number eight to number one, which tells you that Gotye is just a one hit wonder and all of Lana Del Rey's fans love all of her songs.

We never knew that for 70 years of selling vinyl, cassettes, CDs and downloads. We now know it in an age of streaming. We now monetize consumption. Not transactions. Pivotal moment in our business to realize that. Quick commercial application. If Sophia was running a festival and she's going to book Gotye, you should give him four minutes, 18 seconds on the stage.
Cause that's all the fans want. If you're going to book Lana Valerae, book in an extra hour for Encores!

Cause all her fans love everything she's ever written. Great, we understand that. That's music. That's my little 40 billion business of music. You deal with much bigger businesses and your audience much, much bigger businesses.

Have a think about this, if Rishi Sunak was told car sales are up 4 percent this quarter compared to the same quarter prior year, that's a transactional data point. Doesn't Rishi Sunak, the Prime Minister of this country, need to know how those cars are being consumed? How are they being used? Are they all Toyota hybrids, which A, has an environmental implication, B, being used for Uber, which is not an ownership application, it's an access application.

So imagine how government would work if they had consumption statistics as opposed to transaction statistics.

And for your audience, have a look at your business, audience listeners, and think, am I an expert in dealing with defunct transactional data, and a complete novice at dealing with. Advanced consumption data, and if you are Then you've got to swing from that old vine to that new vine yesterday, because that's the way the world is going.

Even my own discipline, economics, is at risk. Economics is often based on transaction analytics. Data science is consumption analytics. So I've got to eat my own dog food here and say maybe I'm redundant as an economist, because all of our theories are based on transactions. Not consumption. That's lesson one.

And let me just pause if you want to just explore that further before I

Sophia Matveeva: well, I wanted to say that. So my previous company was a retail tech company. So we were analyzing what people were actually wearing and how they were wearing it and how they were combining it and how they were making those decisions.

Like why somebody has to buy something or not buy something. And we're essentially the only company that was with the woman in the dressing room, because getting in, getting into the dressing room as she is, I don't know, in Zara, like everybody wanted to know, but obviously privacy, right? We essentially created this experience where she would take photos and ask our stylists in the community, like, should I buy this? Should I wear this? Or like, it's eight, my favorite ones were like, it's 80 percent off, but it's two sizes too small. What do you think? And like, I've definitely been in that situation. I'm like, it's 80 percent off, but it's Dolce Gabbana, but it's lime green.

Like, shall I just buy it and dye it? And What I find is that the retail industry, so fashion apparel is actually really interested in consumption because they, yes, they're interested in transaction, but what they want to know is why did the purchase happen, because the more they understand why it happened like, was it an impulse buy, because she was walking.

Past the shop and she saw a shop window or was it because you know she's got a date coming up with a guy that she really fancies and she's been planning this whole thing because if you understand in retail how the consumer is making that decision then essentially it can really change your business.

So out of, I don't know, I think from what I see media and retail they're really interested in consumption data. I'm not really sure about everybody else. I think everybody else is still very much transaction based.

Will Page: I know, I know. And even when companies try to be consumption based, they often fail.
And there's no great example here in the United Kingdom and supplies globally of how you measure TV and radio consumption in 2024.

Sophia, it's still diaries based on recall. What radio station do you listen to on Tuesday? I don't know. Well, just tick the box. That's not how to measure consumption compared to Spotify, which knows the age, gender, operating system, app being used, time of day analytics, city location of every single one of the half billion users.

This second, Mm hmm. And you're still using diaries to say, what did you watch on Tuesday evening?

That's not fit for purpose. So the consumption is huge everywhere. I guarantee every listener to your podcast will be able to sit back and think about this for a second and think, wait, 80 percent of my decisions are based on transactional data, which do not give me the answer I'm actually looking for.

How do you find consumption analytics? Second point.
 

The Role of Attention in Business

Will Page: Is more kind of builds out of the first, which is if you're going to address a problem and your audience will be well equipped at problem solving, what is the problem you're trying to solve? I argue that the first problem you should try and solve is one of attention.

It's scarce and people want more of it. So if Netflix wins our attention, then everyone else has to lose, which means everyone else has less attention to compete for. And there's a funny story to that, which is. I wrote a piece in Billboard Magazine, our official trade publication, which said we're not competing against each other.

It's not Taylor Swift versus Katy Perry. It's not Universal versus Sony. It's not Spotify versus Apple Music. We're all competing for scarce attention that could otherwise get lost. I sent it to my colleague at Netflix investor relations there and said, what do you think of this piece? We've been discussing attention economics for years.

What do you think of this piece? And he wrote back two weeks later, he said, well, it's great. And Reed

Hastings responded to it. It's like, wow. And the headline was sleep is our biggest form of competition. And at the bottom of the email, he wrote PS FU that is. Netflix is going to win and Spotify is going to lose.

So I wrote back to him and I said, that's fantastic, man, to have Reed respond to your op ed. You know, I really get the point. Sleep is a form of competition, but by the way, Spotify's got sleep music playlists as well.

So F you too. So we can get to you when you're asleep, Netflix can't. But then you think about like how attention affects our business.

And a really good example from our world is what the rules of attention are doing. Our rules in streaming say that you don't get paid unless you've been played for 30 uninterrupted seconds and you don't get paid a penny more unless if you've been played for a second more. That's just our rules, our attention rules.

What happens? Songwriters write shorter songs because we don't compensate for duration, and they put the chorus at the front to hook you in for that first 30 seconds. This has been going on for like eight, nine years now. Avicii was one of the first artists to do it. But then you can see how the tail wags the dog, the rules affects the game, the business affects the show.

So I encourage people to think about attention as a scarce property, where there's Uber competition to get what's left of it and how it affects how the game is being played. We've seen it in music and I'm sure your audience will be able to relate to that and see how it's affecting their world. If you're a lawyer, how long should that memo actually be?

Or should it be a tweet length paragraph because that's literally the attention span of the client? You know, how do you, how do you fit this attention framework into the scarcity of the consumer's mind is a challenge which is universal.

Sophia Matveeva: This is really interesting.

The Influence of Economics on Creativity

Sophia Matveeva: I think that also the idea of how do economics impact creativity.

You know, I think that's a, that's an age old discussion because, you know, when I found out that, for example, you know, some of the world's greatest composers essentially had to go and like teach rich people's kids the piano in order for them to compose wonderful sonatas or like that, they were writing things that, you know, I now know and love.

And essentially it was basically something that they wrote for, you know, a rich guy's wife's birthday. And you think, Oh, did he, it sounds like you really put his heart in it, but it's like, actually no, he just needed to pay his bills.

Will Page: Yeah. And Beethoven didn't have copyright either. So you had to perform as much as possible before people worked out how to copy him.

Sophia Matveeva: This has, this has been an absolutely fascinating discussion.

Final Thoughts and Resources

Sophia Matveeva: I could keep on asking you more questions, but that would be cruel. So what I'm going to do is I'm going to invite listeners to listen to your wonderful podcast, Bubble Trouble, and to read your book Pivot, which you very kindly sent to me.

And I thought it was really, really good. So Are there any other places that people should follow you on so they could learn more?

Will Page: Sure. Well, thank you for name checking the book, but if I can quickly tell your audience, the book was called Tars and Economics as a Hardback. Mm-Hmm. and then Retitled Pivot.
And the back story to that was so we had this situation where W eight Miss. Our main travel agent at airports and Hudson Travel in America said to our publisher, we love Will's book, we really want to get behind it, but books with economics in their title don't sell at airports. So get rid of that title, get rid of that brand, come back with something different and we'll put it in.

So the publisher came up with the term pivot. It's got a really nice binding where the each letter of pivot rotates 90 degrees. You know, it's quite clever cover, all fine with air. I flew to South by Southwest last year in March and my flight was delayed by six hours. I thought, you know what, I'm going to go around every W Xmas in Heathrow and see if they've kept their promise.

And there it was on the table and it's selling in Korea. So one key thing is the website to support the book is intuitively titled pivotaleconomics. com. And I built that as a resource for people like your audience, students, professionals, where if you are interested in this first to suffer, first to recover story that music's been through, if you are seeing yourself stare at your Napster moment, there is a wealth of information on that website to help, to educate, to teach about how to handle this type of stuff.

So I'd say that's the best resource. And again, a huge shout out to our previous guest viewers and the co host. of Bubble Trouble, Richard Kramer, for teaching me so much about understanding what companies and markets don't tell you, which is what the purpose of that podcast is. Why do we always seem to get fooled by bubbles?

And there's another one coming up right now. So it's the more the stock market goes down, the more our audience goes up.

Sophia Matveeva: Excellent. Well, I love your podcast and my conversation with Richard Kramer was one of my favorites. So audience, if you haven't listened to it yet, go and check that out. Well, thank you so much.

Will, I've learned so much from you and also it's been really fun talking to you. Thank you.
Will Page: Thank you very much for having me on your show.