Two (More) Hypotheses on Low Online CPMs

1. If low online CPMs were a result of oversupply, why haven't CPMs in other media dropped by the same amounts? If oversupply were why prices are so low, not efficacy, then online media would be a substitute for other media and ad spend would shift to the low-cost medium, equalizing prices. This is not what seems to be happening.

But if online ad spend has not yet caught up with online attention, that supply is temporarily exceeding demand. One way to test this would be to see if offline CPMs are increasing (because demand there would exceed supply, driving up prices.) If it were true, it means that online CPMs should increase over the next few years back to "normal" levels.

2. Since media sells attention but is paid by the impression, then the fragmentation of time spent on media might lower CPMs as the "bad money" of low-attention impressions drives out the "good money" of high-attention impressions.

Media has always been an attention economy

My friend John Krystynak finally got around to reading my old post, Supply of What?. In it I argue that a surplus of advertising inventory online is not the cause of low CPMs. He vehemently disagrees:

First you say "there is not appreciably more inventory". HA. Prima Facie ridiculous. Are you talking about online? THERE'S A TON MORE INVENTORY now online, maybe a factor of 500? or 5000? YouTube + Google alone would be enough to prove my point.

My point is that what we call advertising inventory (space on a page) is not really what advertising inventory is. This is confusing, so let me say it differently. I'll agree to call space on a page inventory if you agree that what media companies sell is not inventory, but attention.

That is: media companies do not sell advertising space, they sell access to the consumers of their media.*

I mean, they do sell advertising space, but not really. When Advertisers buy a page in a magazine that sells a million copies, they are not paying to get 1 million pieces of paper, they are paying for the attention of 1 million people. The advertiser is not buying space and the media is not selling space, they are buying and selling audience attention. They simply measure it in pages distributed.

Let's say the advertiser was paying $0.01 per page printed, or $10,000. Now lets say the publisher decided to print 2 million copies of the magazine, but still only sold 1 million copies (the other half went unread.) The advertiser would still only pay $10,000, right? So the price per page printed would halve. That's because the advertiser is not paying for space, they are paying for audience attention.

The supply here is not advertising "inventory", but people paying attention to the inventory.

So, is there more attention being sold now, or less? There is more on the internet itself, but this is certainly offset by less being sold in other media. If we are spending less time with media overall, then this has to be true. Let's assume hours spent with media as a proxy for attention available to be sold by media. If this is true, then we can say two things for certain:

1. There is less attention, and
2. That attention has become extremely fragmented.

People used to spend some hours a day reading newspapers. They now spend some of that time on the internet. But if they spend 10 minutes on YouTube, 20 minutes looking at Facebook messages, 15 minutes reading their Gmail, 5 minutes searching on Google and 10 minutes looking at a friend's vacation photos on Flickr, they still will only pay attention to the same amount of advertising as if they spent an hour reading a newspaper.

So all of this brings up two other possibilities about low online CPMs, but I'm going to break those out in a different post.
----------------
* Advertising is a two-sided market. Consumers and advertisers interact through the platform of the media. There's been some really interesting analysis of this, like Anderson and Gabszewicz's A Tale of Two-Sided Markets. But I have not yet seen a fully-articulated two-sided market model that provides any explanation of price levels. It doesn't seem like it would be hard to build a simulation, but I suspect the simulation would be very sensitive to the assumptions, whereas the real world--at least the old media real world--does not seem to be especially sensitive to changes in exogenous variables, like media consumption per capita, roi of marketing spend, etc. If this is true, the model would probably not be very useful.

What is it worth?

Maybe this is obvious.

Things are worth different things to different people. Companies, pork bellies, advertising inventory. Any things. If the thing is worth more to someone other than its current owner, the owner might sell it to that person, creating value for both.

The value to the buyer (b in the picture below) has to be greater than the value to the seller (a). At any value between a and b, both parties are better off. So what price, between a and b, will be paid?
One of my mentors in the valuation business insisted that any value greater than a was strategic value and that the seller did not deserve any of it: the price paid should be as close as possible to a. In an efficient auction, however, the price paid is pretty close to b.

Both the buyer and seller do best by figuring out what the value of the thing is to the other party, and negotiating to that value. If both buyer and seller know the value to the other party, the negotiation over price can be very difficult, because there is no right answer*. So people try to hide how much something is worth to them: whoever has better information ends up with most of the excess value.

Advertising inventory is worth next to nothing to the publisher itself. It's worth something to the advertiser. The excess value, the gap between a and b, is very large. The advertiser knows exactly what the inventory is worth to the publisher. But the publisher has no idea whatsoever what it is worth to the advertiser.

Guess who gets the excess value?

--------
* Non-iterated.

You Are Here

Shirky: Advertising Now Priced at its Real Value

Clay Shirky, in a fairly recent talk, observes

[Newspaper's making their money from advertising] was the historic circumstance, and it lasted for decades. But it was an accident. There was a set of forces that made that possible. And they weren’t deep truths — the commercial success of newspapers and their linking of that to accountability journalism wasn’t a deep truth about reality. Best Buy was not willing to support the Baghdad bureau because Best Buy cared about news from Baghdad. They just didn’t have any other good choices.

and

[Newspaper] advertisers were forced to overpay for the services they received, because there weren’t many alternatives for reaching people with display ads — or especially things like coupons.

and

The second characteristic of the happy state of the 20th-century newspapering was that the advertisers were not only overcharged, they were underserved. Not only did they have to deliver more money to the newspapers than they would have wanted, they didn’t even get to say: “And don’t report on my industry, please.”... Neither of those, neither the overpaying or the underserving, is true in the current market any longer, because media is now created by demand rather than supply — which is to say the next web page is printed when someone wants it to be printed, not printed and stored in a warehouse in advance if someone who may want it. Turned out that when you have an advertising market that balances supply and demand efficiently, the price plummets. And so for a long time, people could say analog dollars to digital dimes as if — well, when do we get the digital dimes? The answer may be never. The answer may be that we are seeing advertising priced at its real value for the first time in history, and that value is a tiny fraction of what we had gotten used to.

These are just the parts relevant to advertising. Read the whole thing. Clay is--as always--the smartest commentator on the newspaper business I know (yes, this is like being the tallest dwarf, but still.)

I wonder whether he's right about what the real price of advertising should be. He's the first person I've heard categorically claim that newspaper CPMs were (and thus are, to some extent) not good value. His assumption that advertising markets used to be inefficient and are now efficient perhaps gives too much credit to the current online system.

If newspaper CPMs were not good value--that is, if the ROI on newspaper advertising was less than the applicable advertiser hurdle rate--then why did advertisers buy them at all? Clay's answer, that they had no other choice, is non-sensical in terms of value (if we agree that newspaper advertisers were not each very stupid for a very long time.) Perhaps what he means is that the negotiating leverage has changed. But this is a rather weaker claim.

Skidelsky on Scapegoating

Here's something I've been trying to say for a year, poorly. Skidelsky in his new book, Keynes: The Return of the Master, says it succinctly:

Whenever anything goes badly wrong, our first instinct is to blame those in charge--in this case, bankers, credit agencies, regulators, central bankers and governments. We turn to blame the ideas only when it becomes obvious that those in charge were not exceptionally venal, greedy or incompetent, but were acting on what they believed to be sound principles: bankers in relying on risk-management systems they believed to be robust, governments in relying on markets they believed to be stable, investors in believing what the experts told them. In other words, our first reaction to crisis is scapegoating; it is only by delving deeper into the sources of the mistakes that the finger can be pointed to the system of ideas which gave rise to them.

As the crisis fades and everyone turns their attention elsewhere, I don't want to forget the lesson learned: what we know about economics is incomplete. Even more, no serious student of economics can now claim that any of the current "systems of ideas" are more than simplistic, directional models. No one knows nothing, and the people who claim to are fooling themselves.

Is the Time for Angels Past?

In the past month I've had two companies that I've committed to investing in come back and say a VC has decided to take the whole round. These were the best two deals in my funnel. The next two best companies are asking for non-standard deal terms. I don't do non-standard deal terms (after twelve years of professional venture investing, I've realized non-standard deal terms are just not worth the hassle.)

The last 18 months have been a great time to be an angel. Only the hardest-core company-building VCs--like USV and First Round--were systematically investing in seed stage companies. This left room for people like me to invest.

Now the VCs who sat out the last year scared have realized that if they want to put money into successful companies at reasonable prices, then they need to have invested in those companies before they became successful. And this option value means they can offer entrepreneurs a better deal than I can (my optionality is limited by my relatively meager investable funds.)

The people I've co-invested with over the past couple of years have been a huge resource to the companies we're in. I'd hate to see that strategic value squeezed out in favor of investors who are more money-manager than company-builder. But that, I think, is what is about to happen.