Note: I am gathering some of my better essays from the past and collecting them on Medium, as they are easier to store here than my Tumblr.
This one was written April 2, 2011. The Chris Dixon article, and bubble talk are obviously somewhat dated, but people often ask for a link to this for its analysis of the digital ad economy as a whole.
Okay, a few more thoughts on this Chris Dixon article and whether it’s a bubble, since going and reading the article, I am surprised to see Dixon arguing that not only is it a safer bubble (private investors, discussed in my last entry) but that perhaps it’s not a bubble at all.
That, to me, seems incorrect. It’s a bubble. Absolutely no doubt. Some thoughts on his arguments for it not being a bubble:
First, Dixon points out that bubbles are the decoupling of valuations and underlying economic fundamentals. This is true, definitely, but bear in mind it is the decoupling of individual companies, or assets valuations and the individual company’s economic fundamentals, not the economy’s as a whole. A strong or growing economy doesn’t make this less true. It’s on an asset by asset basis. Dixon correctly points out that it’s hard to pin the value of the current crop of assets, so he can’t say whether or not they’re overvalued. However, this lack of ability is actually a characteristic of a growing bubble. I don’t know Facebook’s economic fundamentals (aside from rabidly reading the rumors) but I would still love a piece of it. That should be of concern.
The argument that bubbles can’t happen so long as the economy is growing, because there’s always someone left to hold the bag, is incorrect. Bubbles, almost by definition, decouple from the rest of the economy, and you can very much have a bubble burst when the rest of the economy is doing just fine. In fact, this is more or less what happened in the first dot com bust, though like many bubbles, it got so big it effected the rest of the economy. This is an important distinction. This last housing bubble was inextricably tied to the economic meltdown that occurred in the banking sector. The two were linked, so it’s easy to say all bubbles are that way, but it’s totally not true. The 91/92 real estate bubble bursting (mainly in the south) was not tied to the larger economy. The Japanese asset price bubble in the late 80’s. The 97 financial crisis. In fact, by and large, bubbles happen in improving economies.
Okay, second point, where perhaps my trade and job might shine a unique perspective. Dixon says:
“The forces that drive the internet economy are strong and will probably only get stronger. I argue this regarding online advertising here so won’t repeat it. Since I wrote that post we’ve also seen a number of tech companies emerge that are generating significant revenues through non-advertising means — “freemium” (e.g Dropbox), paid mobile apps, virtual goods (e.g. Zynga), transaction fees (AirBnB), etc.”
I’ll leave the freemium/non-advertising revenues off the table since I’m no expert, but I get the strong sense they are still ancillary. The online advertising economy growing issue, however, I believe, is a red herring. And one I know a bit about.
Brands don’t actually want or need any more media channels. As far as they’re concerned, the internet can stop now. We have enough channels. We were happy when we had like seven (TV, print, outdoor, radio, in-store, direct and theater), got a little interested in the first few new ones. Urinals? Uh, okay. Banners? Interesting. Google? Yes. Groupon, Farmville, GroupMe? OKAY I AM GETTING TIRED NOW. Silicon Valley seems to think that advertising’s appetite for new media channels is unending. It is not. Marketers are changing. They are not the daft old man who doesn’t understand the new thing but knows he needs it and spends money on it. It’s a woman and she is getting smarter. Even she knows there’s a point where they’re reached their customers enough.
And yes, every time I talk to a Valley person about this they go on about how marketers want more data and they’ll pay for it, and they want to know everything about a person and oh man do you know how much advertisers would pay for this data? Well, yes, I do. But I guarantee you, YOU DO NOT KNOW. They have fixed budgets. I could tell you EXACTLY how much they will spend, because I spend that money. It is not bottomless.The endless quest for advertisers to know everything about their customers may never end, but its budgets will not increase forever. We will pick the best 3-10 data sources and stick with them. And in the meantime, the VCs will have effectively funded a massive R&D effort to radically improve those sources (THANK YOU) but in the end, we’ll still be paying about the same amount a year for the same 3-10 (say 100 if you’re P&G or WPP).
Then there’s the media money. The money agencies spend on placing ads. Though the valley can’t usually articulate it, this is the other big batch of money they’re going for, because it’s what Google and Yahoo! did. They got big brands to spend some of their giant media money on them instead of yellow pages or newspaper ads. You look at Google, and you think to yourself “oh man you can make a giant business out of advertising!” This line of thinking is usually accompanied by a chart showing the migration of advertising money from traditional channels to online channels, and they say “look how much more there is! It’s moving over at the rate of $6 Billion a year!”
Okay, stop. No it’s not. This is going to level out eventually. Do I have a chart to prove it? No. But it’s just common sense. Advertisers are going to stop advertising on Glee? Mad Men? American Idol? Viewership erosion has ebbed, TV has stepped up to the new competition and started making awesome shit. The first, easy places for innovation — yellow pages, classifieds — are levelling out, and while brand advertising spend is growing online still, it will never get all of it, since, well, the net still sucks for brand advertising* compared to TV and Outdoor. Pew Internet will track the changes now between web and tv in media consumption habits over years and decades — nothing will move fast enough to have an impact on this bubble.
Next, the olds. And that’s the other chart these people usually trot out — some chart showing how many old people are on the internet now. No they’re not. And they’re not going to be any more than they are right now any time soon. Look at the rate they’re increasing their use of the internet. It’s not increasing. It’s holding steady, yes, but there’s no big shift in the immediate horizon.
Plus, even if there was, the olds cost less to advertise to. Ads on Glee cost about $272,000for 26 million viewers. Penny a viewer. Those ads on 60 minutes cost $92,000 for 14 million viewers. two thirds of a cent per viewer. Daily show? $28,000 for 3 million viewers or so. .93 cents per viewer. And that’s cable. Advertisers spend less on advertising to the olds on a per viewer, per show basis.
Do you know how much those Facebook ads targeted to you cost vs those targeted to the olds? EXACTLY THE SAME. At first you think, “hey great, when they advertise on the web instead of TV, we’ll make more!” But no, it doesn’t work that way. First, it is a massive disincentive to migrate that advertising spend to the web, and secondly, it means we already got the good part of the spend. There is no reason to think that advertisers won’t just keep spending on the olds on TV until they die, then advertise on the web to the next set of olds who will be on the internet. But either way, nothing’s gonna happen with it in any big way in any time frame that’s going to impact this bubble.
I’ve mentioned this before, but I believe strongly that as money migrates from traditional to digital it will decline. Part of it is due to what I outlined above. Part of it is because of the perceived lack of inventory limitation. And part of it is because we told marketers for years we were cheaper.
It gets worse. IF (big if) Silicon Valley ever actually delivers on the snake oil dreams of perfect information and perfect targeting that they promise the ad world (and, by the way, for the record, I doubt I’m alone in advertising when I say I don’t want it, and I don’t need it, but that is totally a different article), why would I need to buy a bunch of ads on the 1,000 or so new ad-funded startups to reach my audience? The very efficacy of the products the Valley is building will utterly disrupt the economics they are purporting to claim will sustain them. Will it work for the better for the Valley or the worse? Just because it worked for the better for Google does not mean it has to happen that way again.
You know what? Fuck it. Hey Silicon Valley. You can have EVERY SINGLE PENNY the US spends on advertising. $200 billion. Forgive me, I’m mixing figures here, but it will still be broadly be true. You know what? TAKE THE WHOLE WORLD. Hey internet, take every penny spent on advertising in the whole, wide world. $450 billion. You already havejust over $100 billion of it. You only get 350% more growth. At all. Ever. Google’s revenue? Let’s call it $36 billion. YOU ONLY GET NINE MORE GOOGLES. Worldwide. If you get it all. And you will not, ever, get it all. Yes, the world economy will grow, yadda yadda, and ad spend will increase as the world economy grows, but not at the rate to make one bit of impact on this bubble, and not at any rate comparable to putting your money in a savings account.
If every penny of advertising in the world came to Silicon Valley, out of these thousands of new startups, there would be only nine Googles. Worldwide. In China. In Europe. It sounds like a lot, but how many companies are out there being funded on this premise?
One last way to look at the baked in absurdity: say everything I’m saying is too negative and every year, like clockwork, SIX BILLION DOLLARS of advertising spending arrives in the internet economy. I think that figure is off by an order of magnitude, but so what. Six billion bucks. That’s it. How many companies can that fund? How many are going to get their piece of that pie? We are fighting over a pie that is roughly the size of the cupcake industry.
I will leave you with a final thought, that’s really best for another post. Despite the Valley’s obsession with advertising, they still massively misunderstand it. They think it can be solved with algorithms, and not people. They also thought this about lawyers at one point. LegalZoom is doing great I hear, but so are lawyers. Not every penny of advertising will ever go to computers.
Okay, okay, I am done. But to recap:
1) A solid economy does not mean a bubble can’t happen
2) Marketers aren’t going to pay much more than they do now for all this awesome new customer data
3) TV ad money will not turn into internet ad money at the same revenues
4) The worldwide ad spend is not enough to fund a hundredth of the companies that are being funded to pursue it.
* this was an edit — it used to say “brand marketing” but I meant “brand advertising,” asNoah pointed out. Thank you.