undefined | Better HN

0 pointsChuckMcM11y ago0 comments

As I mentioned above, they have CPC erosion in both their AdMob numbers and their Sites numbers. This statement though I've heard a couple of times:

   > (Except that right now growth in the sheer number of
   > mobile ads is outpacing the growth in monetization 
   > of those ads.)

I don't think I understand it. I get that more and more ads are being shown (growth in sheer number) but this part I don't get (outpacing ... monetization). So Google is throwing more and more ads out there, and at some point they are going to discover how to make them more valuable to advertisers so that the advertisers will pay more for them? I get that this isn't like a commodity good which is racing toward zero margin, but I don't see mobile ads getting more valuable. I characterize what we do at Blekko as 'utility search', which is what a lot of mobile users seem to need. When they search on their mobile for a restaurant name you can give them "perfect" search results by returning, the home pages of the restaurant, a onebox with hours/phone/address, a link to a review site, and an advertisement for a coupon at that restaurant or a competitor in the same food pallete. Google gets that, its pretty much the formula they use, but it isn't special because pretty much a solid businesses database, set of navigation links, and an ad feed for restaurants is all you need to serve that up. And other people are willing to give the phone provider a bigger cut of the ad revenue.

That is why I'm having a hard time imagining a way that Google is going to pull higher CPCs here. It feels very commodity to me and that is a margin game.

As a margin game that 32% net margin which is funding free food, self driving cars, and moonshots comes under pressure.

0 comments

3 comments · 1 top-level

nyrulez11y ago· 2 in thread

I don't understand your equalization of CPCs with value. They are not the same thing. In the early days, CPCs were low and advertisers were getting good margins. It doesn't mean those clicks were less valuable, just that there was free lunch to be had and it didn't last. Ultimately the cost approaches an equilibrium where it is a balance between what a vertical or the sector is willing to pay and still make money off the traffic. This upper bound usually doesn't change that much per sector/market segment.

So I would say, the value from a click doesn't change for the advertiser as long as they are running the same business - nobody can make those clicks MORE valuable for the advertiser just by charging more.

Now, if advertisers are getting cheaper clicks with good margins, its a matter of time when the competitive forces drive the margins down to what's acceptable to survive for that segment. This is true for most advertising channels, not just online. This assumes that there is a fixed tolerance for margins for a sector/device combination. Unless all advertisers collude to keep CPCs down for themselves, the cost should reach equilibrium - think about it, why would other competitors not drive up the cost and get MORE traffic, until they reach their margin tolerance? It is a volume vs. margin question. More volume means less margin, and there is an equilibrium to be had.

As you have also stated elsewhere, CPCs going down is hard to understand without a distribution of CPCs bucketed by traffic. As volume grows and new markets are penetrated - it could either mean that it is touching more lower margin businesses OR equilibrium has not been yet established for those new markets. But it is hard to conclude that existing high margin businesses are suddenly paying less for those clicks unless their fundamental business properties are changing i.e. lawyers can't make the same money as they used to anymore and hence margins are squeezed. That may or may not be true - but it is not possible to deduce that from what's given.

ChuckMcMOP11y ago

   > I don't understand your equalization of CPCs with value.

CPCs are a proxy for value, especially in the Google case. This is because the way advertising works on Google advertisers bid for position, Google runs an auction, and the highest bid wins (at the next to highest price, it was at least when I was there a dutch auction[1]) So when CPCs fall, they do so because the highest bid has fallen, and the advertiser is setting their bids, so if they set them lower then it means they "value" being at the best position lower and thus "value" that ad placement slightly less.

Now flip your point of view around to that of the advertiser, you've got $1M to spend on internet advertising. Every quarter you advertise with different strategies and campaigns and you measure your business. If you are like many you are constantly tweaking that, some print, some radio, some Internet, vary the mix vary the message, these folks A/B test the crap out of this stuff. When you find bidding $5.00 a click gets you the same business activity as bidding $4.90 a click you say "Great, that gives us more money to spend on some other channel!" You have just reduced the CPC that Google sees from you by 2% and you are still getting the results you want.

When CPC is eroding both quarter to quarter and year to year it means that people as spending less money on buying ads from Google on a per ad basis. They may get more total impressions (and this is what I suspect Pichette thinks is the confounding variable in the paradox) but each impression is worth less. And the interesting thing that Marissa Meyer proved at Google early on was that people valued less intrusive advertising over what Altavista and Yahoo where doing at the time. I believe her insights still hold, and if Google continues to add more 'slots' into their properties they are going to erode value (ad blindness). It is a tough spot for them.

[1] New hires back in 2010 got to do a class called 'life of a dollar' which was perhaps the absolute best introduction to how Google monetizes search you could get.

nyrulez11y ago

Ok, I understand the view point. But here are couple of questions to consider:

1. How do you know that CPCs are falling because of existing advertisers paying less vs. new advertisers in new markets/mobile are paying less ? I.e. what evidence suggests that existing high value auctions are depreciating in CPCs/Value ?

2. CPCs are a proxy for value only if they have reached equilibrium. Since it is an auction as you have explained, it takes time for different participants to converge to the right bid configuration. That can take months or years to mature. So if the mobile market is new, how do you know that it has already reached equilibrium ? Desktop CPCs rose for a long time until they established that for many verticals.

j / k navigate · click thread line to collapse