Tag Archive | digital

Why digital may not take all of the ad dollars: that darned Pareto Principle

Digital advertising spending has grown over the last decade by stealing a lot of share from print magazines and newspapers and some other categories. But the odd thing is that ad spending in North America for both TV and radio advertising is NOT collapsing in the same way as print. In fact, depending on the year, both are still rising! I have a theory as to why.

This gets a little tricky, so bear with me! You may remember something called the Pareto Principle, also known at the 80/20 rule? It says that 80% of the effects come from 20% of the causes. This isn’t true in every industry, but it IS true a lot of the time. Many companies get about 4/5 of their revenues from one fifth of their customers, for instance.

It is also true of many digital behaviors, according to Nielsen. The top 20% of users are responsible for 76% of in home Internet usage on a PC, 83% of smartphone video, and a shocking 87% of streaming on a PC. (Is hyper-Pareto a thing?) On the other hand, the 80/20 rule does NOT apply to traditional TV and AM/FM radio. The top fifth of viewers/listeners still account for more than their share, of course…but they represent ‘only’ 52% (TV) and 48% (radio) of all usage. Much less than Pareto would suggest. Why is this important?

If you are an advertiser, and you want to sell a product with a very precise and targeted market in mind, digital is awesome! You can spend only what you need to, and ‘waste’ very little of your ad dollars. (It doesn’t mean they always buy, but at least you aren’t paying for ads to be seen by people who would never purchase your product.)

But what if you are selling a product that EVERYBODY might want to buy? In that case, if you buy (for example) display ads on video streaming on a PC, 87% of your ads will be seen by 20% of the market, and the rest of your market will see very few of your ads.

That suggests that some products are better for digital, but others are better for traditional broadcast media like radio and TV. If true, that would suggest that the spending gains seen by digital in the past few years may start slowing, TV and radio will stay resilient or even grow, and that their relative shares will stabilise for the next few years.

Interestingly, that is exactly what seems to be happening when we look at ad spending trends for the first half of 2016: TV and radio are doing better than expected, and digital is still growing, but a little less than expected.

The implications for the future of TV and radio advertising are intriguing. Yes, audiences are dropping a little, but ad spend may be more robust.


Ad spending has fallen off a cliff forever. Or not.

Advertising as a percentage of GDP has fallen in the US. Is that the new normal, or will it go back up one day?

The chart above fills my little data-geek heart with joy: it has 90 years of data, the data comes from literally thousands of independent sources, and it covers a large and very well-measured market. At a 99% confidence level, I am sure that the chart is showing a genuine and important trend: US advertising spending as a percentage of GDP was highly stable in a 1.1-1.5% range between 1947 and 2007 (60 years), but in the last ten years has fallen sharply to below its historical range. It has never been this low before, except during WW2.

But will this last?

Theory #1: Yes, this is the new normal, and will persist. If you have heard of “trading analog dollars for digital dimes” you will be on the right track. Digital advertising is more targeted, more effective, more measurable, and therefore more efficient. That means advertisers don’t need to spend as much to get ad effectiveness, and therefore don’t spend as much. We can expect ad spending to stay under 1% for the future, and may even drop further as more ad dollars move to efficient digital and away from inefficient traditional ads.

Theory #2: No, this will not last. Advertisers are like kids at Christmas playing with a new toy. Digital is novel, and does have some advantages, but the rates of ad fraud, bots, ad skipping and ad blockers means that advertisers are going to need to spend much more than they are today, on a mix of both digital and traditional advertising. The 1% level is not sustainable, because digital isn’t as effective as its proponents believe.

Theory #3: Digital advertisers (especially Facebook and Google, who share 55% of digital ad spend and 2/3 of the annual growth: see chart at bottom) are doing what all new entrants do: they are coming into a market, and low-balling pricing because that’s how you gain share as a new entrant. Once digital becomes 30-40% of total ad spend, and customers are entrenched in their buying habits, they will raise prices, and we will see ad spend go back into its historical range. It is clear that advertisers are more than capable of paying 1.1-1.5% of GDP for ads over the long term, so why shouldn’t digital players (once they are sufficiently established) charge all the market can bear?

I would be interested in any thoughts on the above. Theory #1 tends to be widely held by new media/digital media types, #2 is widely held by traditional media players, and (so far as I know) Theory #3 is original to me, and hasn’t been discussed elsewhere.

I kind of like #3, but everyone loves their own babies. 🙂



eBooks are not taking over


Digital eBooks are not taking over from print books in the same way that online newspaper websites are destroying print papers, MP3 files took almost all of the CD music market, or digital cameras and smartphones replaced film for consumer photography.

This last week saw two articles on this debate. The New York Times started us off with a piece based on data from the Association of American Publishers (AAP), which said that eBook sales were down 10% in the last year. The very insightful Mathew Ingram replied with his own article in Fortune, which said that the AAP data didn’t tell the whole story, and eBook sales from non-AAP publishers were doing much better. Mathew went on to say

…the overall trend in book publishing remains clear: Digital sales are going to increase, and print is likely to become a niche market over time, just as it is becoming in the newspaper and magazine industries.

I admire Mathew very much, and we almost always agree – but not on this issue. Two big reasons:

First, while I agree the AAP numbers are incomplete, that only means that eBook sales are likely to be down less than the headlined 10%. They might be down 5%, flat or even up 5%. But they are not out there stealing huge market share from print books! Based on my estimate for 2015 final US book sales, revenues for print books will be down about 11% from 2010 levels based on constant 2014 dollars. Print has been AFFECTED by digital, but only marginally so. (And the US is the market which has seen eBooks have the greatest success: on a global basis print books sales have been virtually unaffected. eBooks are under 5% of sales in most countries.)

Don’t get me wrong: down 11% has hurt print publishers and bookstores. But does this really compare to what we have seen in other industries that have been digitally disrupted? US print newspaper ad revenues are down 78% from their peak; CD sales are down 92%, and film camera sales have fallen 99.5% since 1999. When we compare how eBooks have affected print book sales…well, “one of these things is not like the other.” 🙂

Secondly, the shift in the growth of eBooks from AAP publishers to other publishers (see chart below) is very real, and not to be ignored. But equally, I do need to point out that the kinds of eBooks being published by those indie publishers do fall into certain categories. They tend to be what is called genre fiction, often romance, erotica or fan-fiction. Nothing wrong with any of those, but they tend to serve niche markets, with low price points, low volumes and an element of “disposable reading.”

If I compare what is going on with eBooks and the print book market, it would be very different from what we have seen in music, magazines or newspapers. Digital disruption did not mean that MP3 files were the preferred format for emerging acts and polka bands, while Taylor Swift still sold lots of CDs. Digital shifts didn’t mean science fiction fanzines went online, while Time and Newsweek kept selling tens of millions of print copies weekly. And the newspaper story is not that the Moose Jaw Times Herald was crushed while the Toronto Star kept thriving in print. 

Across music, magazines and newspapers the digital destruction has affected all artists, all genres, all titles and all publishers more or less equally. That is NOT happening in the book industry: eBooks are thriving in certain pockets…but they are not doing as well in the segments that most readers and book buyers care about.

Andy Weir’s The Martian (now a Major Motion Picture!) was originally self-published as an eBook. Yay! It sold incredibly well…35,000 copies in only three months! Double yay! If eBooks were truly the future of books, he and the book might well have stayed there.

But he didn’t. He took a $100,000 advance, submitted to professional editing and added more content, and launched it in print. Where it sold hundreds of thousands of copies in hardback and paperback, and continues to sell out of bookstores I visit around the world.

To be blunt, most people view eBooks as a reasonable place to start, but “REAL books” come in print versions, and usually outsell the eVersions by tens or even hundreds of times.

Unlike other print publishing businesses affected by digital, and based on the evidence so far, it looks to me like eBooks are the real niche market.

The Martian


This is terrible! How long have kids been stealing content?


Not to over-generalise, but more-or-less forever!

When I started writing this post, I went into deep thought to come up with an example or two from my own past that could be compared to torrenting TV shows and downloading MP3 files. After 15 minutes, blushing with remembered guilt, I had an entire catalog of thievery[1].

Cable splitters: Back in the 1970s and 80s, you had to subscribe to cable in order to get the US channels: an antenna didn’t work, at least where I lived in West Vancouver. I don’t know what the monthly bill was back then, but a coaxial cable brought us a list of around 20-30 TV channels. At first, when almost all homes owned only a single TV, this was fine. Those few homes that could afford a second TV and wanted to connect THAT to cable had to pay for another subscription. Which was no big deal, since we all assumed that only multi-millionaires had more than one TV; and they could afford the extra $10 a month or whatever it was.

But the price of TVs fell quickly, and even middle class homes bought a new one for the living room, and moved the older set into the basement. But the idea of paying the cable company MORE money (for a signal you were already paying for) seemed nuts. Why not just buy a cable splitter (photo at top), put it against the wall where the cable entered your house, and run a coaxial cable to the new set? I guess it was illegal, but it didn’t feel illegal.

And do another split to another TV set, and (while I am at it) that same cable also transmitted high quality stereo FM signals that I couldn’t get with an antenna, so I connected my FM tuner to the co-ax too. The problem is that each splitting of the signal causes the signal to attenuate (by about -3.5 dB, if you care) so that by the time I had a spaghetti monster of cables coming out of that single ‘line in’, the TV reception had gone to hell. Did I stop stealing and pay for each signal? What are you, crazy? I went out and bought a power amplified cable splitter, which were much more expensive. Not only was I stealing, but I was now purchasing special devices whose only possible use was theft[2]!

Cassette tapes: We were all happy to buy records in those days. My generation was a bit after the 45 rpm single craze, so we owned 33 rpm vinyl Long Play (LP) albums. I had a paper route, and all my money went on either my stereo equipment[3]  or albums from A&B Sound on Seymour St. at $5-10 per record. I probably had nearly 1,000 albums at one point. Not all my friends were so hard working or had such a large collection. Luckily the new generation of cassette tape recorders had better technology like high bias Chrome tapes and Dolby noise reduction, and that allowed me to make decent (but not great) copies for my friends. I didn’t charge them, except to cover the cost of the tapes. That was, yet again, not strictly legal. But since I had already paid for the content by buying the records, making the copies also didn’t feel like stealing.

Multiple phone lines: Remember how we were supposed to pay for each TV connected to the cable? Same for phones. It was illegal to run your own wires and create a new RJ-11 jack, and you weren’t allowed to buy your own phone in the States and connect it to the system. Only phones rented from the monopoly carrier, and connected to authorised jacks, were allowed. There was no way to enforce this, of course, so we cheerfully ran multiple lines and bought phones. To be clear, each phone had to share the same number (929-1939, if I recall) so we only had one line. Once again, this didn’t feel like stealing, but I don’t think BC Tel would have agreed.

Photocopied textbooks: OK, now we’re getting serious. University text books were as expensive then (in the money of the time) as they are today. And paying (in 1982 dollars) $60 for a text was killer, when I was only making $3.65 per hour working at McDonalds. Multiply that by 5-10 courses per year, and more than one text for some classes, and I just didn’t see how I could afford to pay for all of them. So I didn’t. We pooled with friends, or copied library books, or otherwise made our copies[4]  at Kinkos from books we paid for but returned.

Did we know we were stealing? Oh yeah. We justified it all kinds of ways. We did pay 5 cents per page for our photocopies, right? And we all knew that the profs just assigned new textbooks (that they had written) to make money, and the old ones were basically the same. And that professors made an awful lot more money than we did. But, in our heart of hearts, we knew.

Computer software: I stole. I stole left and right. I stole everything but MS-DOS, and that was only because they forced you to buy that when you bought the PC. I stole games software, I stole spreadsheet software, modem software, mapping software, graphics software. Word processing. Charts. Tax software. You name it: if someone had spent years of their life writing code, I copied it onto a 5.25 inch floppy disk and loaded it on my machine. #hangsheadinshame

Why? I am not completely sure how I justified all the theft to myself, but perhaps it was because the PC was this new thing? Growing up I knew that stealing a horse was bad (rustling), and bank robbery was bad. Heck, even a loaf of bread could get you in some serious trouble in Les Miserables! I knew all that from movies and TV and books. But I hadn’t seen a whole lot of information about how stealing PC software was wrong, so somehow it didn’t feel wrong. It may even have felt kind of cool? Hackers were heroes in the movies I was watching.

Computer Time: I don’t even think this is a thing anymore. Anyone who wants access to computing has their own device(s), access to one at work, or can get time at the library. But that wasn’t common in 1985. But my girlfriend did work at the Word Processing centre at UBC. There were a bunch of dumb terminals, and a central minicomputer[5] running WordPerfect. After hours, while Deirdre worked away, she tolerated me hanging around and playing an early text based computer game[6]. It was awful by today’s standards, but enjoyable enough that I asked her if I could bring some friends over. There’s no question that NONE of us were supposed to be there, and we were certainly “making unauthorised use of university property.” AKA stealing.

Gee, my parent’s generation must have thought we were a generation of thieves?


Not so much. I learned the cable splitting trick from my Dad – he’s an electrical engineer and taught me about the dB power loss. I doubt he or my mom cared about the cassette tapes: I had previously listened to hours of bootlegged music they had (illegally) recorded on an old 1958 Philips reel to reel recorder back when they were in their 20s! I am not sure if my parents did this, but teenagers used to sneak into drive-in movie theatres in the trunks of their friends’ cars. I don’t think they knew about the textbooks, and probably wouldn’t have understood the PC software theft, not having a PC of their own.

So I guess that generation, growing up during WW2, were the first generation of thieves of media? Nope. Before that shellac recordings were pirated, the original cylinders were pirated…even sheet music was pirated. The novels of Dickens were pirated in the US, and Twain was pirated in England; and Victor Hugo was pirated both places! I haven’t found anything about illegal copies of Player Piano rolls, but I bet they are out there. And people were stealing Shakespeare’s plays within days of their being performed.

Given that the history of the last 500 years indicates a large market for intellectual property theft, why are we labelling the kids of today as thieves?

A final observation:

Young people steal more, because they are relatively poor. When they are no longer poor, they may be more willing to pay. That may not happen with this generation, but I can only give my own evidence: I pay for EVERY copy of computer software today, and have for decades. Same with music and video: I know there are stolen versions available, but I won’t use them.

Interestingly, we are starting to see some of the same trends in young media consumers today. When we look at growth in PAID services, like Netflix, Spotify, and Pandora, we can see that – given a legal option – they may be willing to pay.

Which was the subject of my CMF guest post.

[1] I need to be clear here. The behaviors I describe below were not felonies. The stealing I am talking about was not against the Criminal Code of Canada, but was violations of agreements between me and various service providers or content providers. To the best of my knowledge, almost nobody was ever caught, charged, fined or even disconnected for the kind of stealing I did (sort of like how no one gets arrested for downloading MP3s, now that I think about it!) There were stories that the cable company would fine you if they caught you using a splitter, for example. But I don’t know if that ever happened.

[2] As an aside, this is much less of an issue today. Each TV set usually requires a set top box, so a simple device like a cable splitter doesn’t let you watch on multiple sets any more.

[3] I still miss my Dual 505 turntable with the Shure cartridge.

[4] While standing underneath signs that warned that copying textbooks was theft!

[5] I think it was a Digital Equipment PDP-11, but I could be wrong.

[6] If I remember, it was a variant of Dungeon, but this was 30 years ago. There were swords, elves, and jewels…if that helps. J

The 21st Century Monocle


Is it possible that Google Glass’ (or similar head mounted wearable computer) primary use case is as a fashion accoutrement for Digital Hipsters?

I was reading an article about how a New York restaurant was being punished in its online Google reviews (but not Yelp) because it refused service to someone wearing Google Glass for privacy reasons. The lady who had been refused service gave them a one star rating and wrote a note about her experience, which is fair enough. But one of the commenters on her post then suggested a bunch of people give the place a flood of one star reviews, which appears to have occurred.

That’s not fair play, but I was struck by a comment from one reviewer on the restaurant’s Google page: “It’s a real pity that such a great restaurant can have its online reputation ruined by a group of trolls sporting 21st Century monocles.” [Emphasis mine.]

I just Googled that phrase, and it seems to be the first time anyone has applied that term to Glass, or any head-mounted wearable. But as part of my search I came across this article: “One Part Mr. Peanut, One Part Hipster Chic: The Monocle Returns as a Fashion Accessory” from the NY Times of March 5, 2014.

There are so many relevant bits, that I ask forgiveness for quoting it so extensively:

“The one-lensed eyepiece, an item favored by 19th-century military men, robber barons and Mr. Peanut, is finding itself wedged anew into the ocular sockets of would-be gentlemen seeking to emulate the stern countenances of their stuffy forebears.

From the trendy enclaves of Berlin cafes and Manhattan restaurants to gin ads and fashion magazines, the monocle is taking its turn alongside key 21st-century accouterments like sharply tucked plaid shirts and certificates in swine butchering…

…Martin Raymond, a British trend forecaster, credits the rise to what he calls “the new gents,” a hipster subspecies who have been adding monocles to their bespoke tweed and distressed-boot outfits. On a recent trip to Cape Town, Mr. Raymond said, he saw such a group carrying monocles along with tiny brass telescopes kept in satchels.

“All of this is part of a sense of irony and a way of discovering and displaying old artisanal and craft-based technology,” Mr. Raymond said. “You see the monocle appearing in Berlin, parts of South Dublin.”

Toby Miller, a cultural historian, said: “Monocles have always marked people out as beyond the crowd, slightly different. On one hand you have the Prussian officer, on the other you have the effete English lord, and then you also have the New York and London lesbian in the 1920s.”

Aside from the Times publishing “accouterment” (what, they don’t have spell check over there?) the words that jumped out at me were trendy, hipster, irony and different. From the above mentioned swine butchering to the Edwardian beard, Pabst Blue Ribbon, Mason jars and Crosley USB turntables at Urban Outfitters, there is a whole thread of hipster behaviour that reaches back into an artisanal past (whatever the hell THAT’S supposed to mean) and embraces things that are deliberately, conspicuously and ostentatiously anachronistic.


And good for them…speaking as a guy who owns TWO Fatman valve amplifiers!

But it also occurs to me that hipsters can go back to the future, as it were. Can’t you consciously mark yourself out as being “slightly different” by going FORWARD too? Reject print media and traditional TV, pay for as much as possible with your smartphone, quantify yourself with a wrist-mounted wearable, and (as a final frontier, perhaps) slap on a head-mounted device that has been compared to both Geordi La Forge’s VISOR and to humans being assimilated into the Borg in the Star Trek universe?

I call it conspicuous digitalisation. Those are people buying and using technology, not as a utility, but as a fashion choice. There’s nothing wrong with that and tens of billions of dollars can and will be made.

But because Digital Hipsters are buying these things primarily to be ‘slightly different,’ it also means that they will never go mainstream. As I have said elsewhere, there are ubiquitous technologies. In the developed world, nearly 100% of the population has access to all of PCs, TVs, smart phones and tablets.

If Glass is indeed the 21st century monocle, it will never appeal to 100% of people, and will be limited to a mass-niche of 10-20%, at most.