Podcast Blurbs [Background on Google’s DoubleClick acquisition, PPG Industries, Quantifying stuff]

Recode Decode (10/30/17; Why I’m not afraid of Amazon, David Rosenblatt, CEO of 1stdibs)

(David Rosentblatt, CEO of 1stdibs, former CEO of DoubleClick)

(on DoubleClick)

“DoubleClick invented the display advertising industry.  It started out life as an ad network, I was hired to unbundle the technology that drove the network and sell it to publishers who competed with the DoubleClick network like the Wall Street Journal and others, and that business really ended up surviving the ad network business, which we shut down and sold and became the core of the company…We went public in 1998, we had a $15bn market cap by 2000, and we hit the wall like everyone else.  Unlike most other companies, we were fortunate in the spring of 2000 to raise $1bn, so we had the capital to remain solvent when others went bankrupt.  I was promoted in 2001 to be the #2 person, between 2001 and 2004, we both survived the dot-com meltdown and attempted to diversify away from our core business, which was advertising, into things like database technologies and data itself.

What happened in 2004 is the equity market for internet stocks came back.  We, however, did not come back with it because by that point we had diversified away from being what everyone thought we were, which was an internet advertising company, and the Board ended up deciding to sell the company through an auction.  We ended up being acquired by Hellman and Friedman, they asked me to stay on as CEO, and our strategy from then on was basically a kind of back to the future approach where we reversed our way out from the diversification we had followed.

The big issues [with internet advertising in the wake of the dot com bust] were the industry was not standardized, it was incredibly expensive to operate both as a seller and as a buyer, there was no standard definition of what an ad unit was, there was no standard way to price it, there was no standard way to evaluate the efficacy of the advertising, and in that kind of environment a big advertiser like P&G aren’t going to buy at scale…in that sort of fallow period from 2001-2004, a lot of the infrastructure that ended up forming the basis of the industry was created.  The ultimate event though which allowed the display business to become as big as it is today was the creation of the DoubleClick ad exchange and the conversion of the display ad business from a sort of old fashioned, 3 martini lunch negotiated ad buy to one in which ad inventory was advertised through a marketplace of tremendous scale.

What ad exchanges allow is for the advertiser to buy the underlying audience.  Publishers at the end of the day are simply a proxy for the underlying audience, so our theory in launching the ad exchange was well, if you could somehow separate the underlying audience from the publisher and allow the advertiser to buy the audience directly, then it becomes a much more efficient process.

We understood that Google was aiming to enter our business…if you could create an ad exchange, then all the stuff that Google was building mattered a lot less […]  The basic idea behind the merger with Google was that what we offered was an ad platform which publishers would use to manage their own advertising…by the time we sold the company, we were delivering somewhere between 3bn-4bn ads per day.  The idea was we would create this marketplace to help publishers monetize unsold inventory…so the simple ideas was let’s marry our publisher install base with their monetization machine and basically take the DoubleClick strategy as it existed then, but power it with all of the demand that Google could bring from the advertisers they already worked with.

At the time, Google was functionally organized, so they had very few people who were focused on this business and who we thought really understood it.  Like a lot of companies, they thought about this new business in terms of the old business, meaning search, and none of that stuff applied or was right.  Secondly, big publishers were threatened by Google, yet at the same time Google needed those publishers for inventory because at the time, Google didn’t have a whole lot of display inventory themselves.

Microsoft and Yahoo! both bid for us.  We ended up in a bidding war between those three…Microsoft under Balmer at the time was hypercompetitive and reacted [to Google winning the auction] in the way that hypercompetitive companies do, which is they bought what they thought was the alternative to DoubleClick, which was a company called aQuantive for twice the price, they paid $6bn, only to then realize they bought the wrong company and they wrote down the full $6bn….Yahoo! bought Right Media to develop an ad exchange and they never invested in it and that too died a slow and painful death.

(on 1stdibs, online marketplace for luxury goods; furniture is the largest category)

“We think we’re in a pretty different segment [than Amazon].  As we describe it, the race for the $50 order is over, it’s been won by Amazon.  The race, however, for the $5,000 order has not been won…What makes the luxury business in general different and harder [for Amazon] to access than most other industries?  It really has to do with the comfort level of the seller with the environment in which they sell…we just added two contemporary furniture brands to 1stdibs, those brands had been selling on Wayfair, but they withdrew from Wayfair even before they started selling on 1stdibs simply because they didn’t want the adjacency of their brand with the lower-priced products.  There’s a reason why you can’t buy Chanel in Target.

The lesson of the internet is power shifts from those who control distribution, meaning supply, to those who control the buyer…those who meet the buyer on his or her own terms, and to the extent the buyer thinks about the world in a seamless way between digital and analog, one has to be in both channels, and so we will be too.”

Rationally Speaking, Ep. #197 (Doug Hubbard on “Why people think some things can’t be quantified (and why they’re wrong)”)

Doug Hubbard:

“[Kahneman’s] point was that there are fundamental misconceptions about random sampling, that everyone gets this all wrong and that it has a big impact on actual research.  And so he surveyed a bunch of published scientists and students, and the fact is there are profound, persistent misconceptions about how sampling actually works and what it tells us.  People kind of remember some things and they’ll throw out words like ‘that’s not statistically significant’ and they didn’t really do any math to make that claim…they’ll also say, ‘well, correlation is not evidence of causation’ and that’s not actually quite true.  Correlation isn’t proof of causation but I can show you Bayesian proof that it is evidence of it…people are just winging it all the time.  They’ll say, ‘well, there’s this potential bias in the survey and because this bias exists, no inference can be made.’  […] are you saying that there’s some randomly assigned variation in the population and that unless we account for all possible variations, we can’t make inferences?  Well, then, all science is wrong.  Every controlled experiment in the world doesn’t actually control for every varying factor […]

…people conclude that unless you do all these things perfectly that are extremely difficult to pull off, I can make no inference whatsoever from observations.  Well, that’s not how you live your life…if you can’t make inferences using scientific method and statistical inference, then how are you doing it with just your life observations?  Because you’re doing that with selective recall […]

The claim that you don’t know the exact probability and therefore you can’t put a probability on something, if that were true, then how come probabilistic models do better than you?…The fact is their subjective decision making is routinely outperformed by standard statistical models in areas that they would have insisted only a human could understand.

I did a model once in the movie industry for forecasting the box office receipts for new movies…[script readers] were convinced that there was no way you could quantify their sophisticated judgment process where they consider, in their words, hundreds of variables in a holistic artistic network.  They really had a very fancy image of their mental processes.  When you compare actuals to [the script readers’] original estimates, the correlation was zero.  So, you and I picking the industry average every time would have done just as well as the experts…I came up with a really crappy model, probably the worst regression model I made that had a correlation of 0.3, an r-squared of 0.1 […]

There is a fundamental issue though with the word ‘probability’ because even statisticians don’t agree on it.  Most people for practical decision making need to treat probability as a state of the observer, it’s not an external thing that you’re measuring, it’s not a state of nature, it’s your state […]

People have this negative connotation to just quantifying things to begin with…so, quantifying things is a really interesting method of looking at the world that seems to be rare among species.  Language may not be that rare among species, but math does seem to be rare among species.  Maybe we should think of [quantifying] as a fundamentally human thing, so it’s not reducing someone to a number, you can elevate someone to a number too.”

Julia Galef:

“I was teaching a class on trying to quantify your own uncertainty, put a probability on your beliefs or predictions, and someone in the class kept insisting that you can’t know what the right probability is.  And I kept trying to get him in the mindset of how he actually makes decisions in real life.  And I’d be like ‘well, you know let’s say you eat a sandwich, if you eat it, that means you probably put a low probability on it being poisoned.’  And his response was ‘I’m not worried about it being poisoned, but there’s no way of knowing the probability of it being poisoned.’  The fact that he made that distinction suggested to me that people have this compartment that they put anything quantitative in where there’s this super high standard and you’re not allowed to make any estimate unless it’s completely rock solid, whereas in your day-to-day life you just do whatever seems sensible […]

There’s some cases in which we are implicitly putting a value on something that we want to be able to say is invaluable, like the value of human life.  So, we might say there’s no limit to the value of human life, it’s infinite.  But in fact, from our behavior, we don’t actually believe that because if we did we would, you know, set the speed limit to 15mph or something that couldn’t possibly kill people.  And we don’t do that because there’s a trade-off – it makes our lives slower and less efficient and gives us less autonomy…and that’s an implicit sign of how we value human life, but we don’t have to talk about it.  Once you start asking people to put a number on human life, then suddenly we’re violating this sacred taboo.

Mad Money w/ Jim Cramer (11/16/17)

(Michael McGarry, Chairman and CEO of PPG Industries)

“Even though the environmental pressures on the companies in China are causing a lot of supply disruption, it’s actually very good for us.  We make and have the world’s leading technology is water-based coatings.  So, we have more and more people shifting from solvent-based to water-based and that’s helping us grow share […]

Our overall refinish business, so when you get in an accident and repaint the car, it’s up in all our segments…we help the body shops repair cars faster, we help them color match faster and better…we’ve converted more shops from solvent-based to water-based than the whole industry combined, and that includes China, Europe, and the US.  Refinish, I might add, is driven by miles driven as well as congestion and congestion is driven by employment.  So, for us, this has been a very good business.  We are also helped by distracted driving.

We’re working with all the [electric car battery] companies.  Without getting into the particular customers, there’s actually more coatings on the battery than there are on the car because what they try to do is insulate the battery from the rest of the car, from a cooling perspective and also prevention of thermal events or fires…in fact, we sell equipment for the solutions as well, so we’re working with all the electric car companies, whether they’re in the US or in China…now obviously, this is many years away, but when it comes, we’ll be on the forefront.

We are the #1 global [auto] OEM supplier, whether you’re in US, Europe, or China.  We work in China with all the global OEMs as well as with all the local companies, we have #1 share with both [groups].  We’re going to be well-positioned regardless of who’s winning, whether it’s electric or whether it’s hybrids or traditional gas powered.”