Patrick Doyle

Supported by Greenhaven Road Capital, finding value off the beaten path.

In a Real Vision interview, Marc Cohodes said he was now friends with Rick Federico of P.F. Chang who called him and said, ‘Whatever your issues, at least reach out to me first.’ Cohodes was impressed by this, by a leader who welcomes good arguments.

Who’s this Federico guy? I wondered. Thanks to the serendipity of search I found his talk with Patrick Doyle, former CEO of Domino’s Pizza and it was fantastic.

Doyle oversaw a revitalization of Domino’s Pizza, with stock market returns that rival any FAANG save Netflix.

Domino’s Pizza is a great story, starting in Michigan in the 1960s and pioneering home delivery. Before we get into the notes on Doyle we should note the half-century build up for food delivery. Christopher Payne, COO of DoorDash, called the switch to convenience from experience is “the biggest shift in commerce.”

The economic of everything is changing. Both ride sharing and meal delivery have affected restaurants said, Dave Chang. What took so long is that food delivery was too early until it wasn’t. Brian McCullough said, “The lesson of the early internet is that sometimes just good enough technology is good enough.”

Before mobile lots of business like Instagram, Airbnb, and Uber weren’t businesses, they were ideas. Timing matters.

But that cuts both ways. Domino’s Pizza had to change too. Doyle said in 2011:

“Are you willing to view the dislocation we had in our economy as an opportunity to drive change and innovation in your business, or do you take a very conservative view and say, ‘Times are tough so we’re going to ride this out until things go back to how they were.’ The people who are taking that approach are going to fail.”

After lackluster years of 2006 and 2007, Patrick Doyle and the Domino’s Pizza team needed to transform. Unlike school, there’s more than one way to solve a problem. Domino’s Pizza might have succeeded in any number of ways. What they actually did was these four things.

  1. Change the culture, remove silos.
  2. Improve the pizza, but don’t perfect it.
  3. Use social media, don’t hog the brand.
  4. Build the future around data and technology.

Listen on iTunes, Overcast, or Soundcloud.

It wasn’t that Domino’s Pizza was bad so much as badly perceived. “All the research told us that our pizza was as good as our national competitors. The bad news was that once you put our brand on that pizza consumers thought less of it.”

Domino’s had negative brand equity. People liked it less if they knew who made it. In third-party research, they also found out that “We were tied for taste with those fine purveyors of fine dining over at Charles E. Cheese.” As any sane parent knows, nothing good comes from Chuck E. Cheese.

Good feedback is hard to get and fatal if you don’t. Two stellar sellers are Jenn Hyman and Peter Rahal who each hounded their customers in test after test asking is this best?

Knowing about the pizza problem was only half the battle. Domino’s had to come up with a solution but they had the to a hammer everything looks like a nail problem.

Doyle’s predecessor, Dave Brandon “inherited perfect silos.” The company needed new ideas. They needed what Greg Lindsay spoke about, “Google started a beekeeping club so that engineers who are interested in beekeeping might meet each other and actually have discussions about unrelated subjects.”

Domino’s Pizza started promoting from within, transferring around, and changing the culture. “We’ve established a program bringing people in early in their careers, starting them in stores learning the basics of the business and then moving them around.”

Some solutions are intellectuizable – I think therefore I get.

Other problems you have to feel to understand. Robert Cialdini told Barry Ritholtz, “I realized that in a laboratory with college students I was missing the power of these (Influence) techniques.” Mike Lombardi wrote, “Blackboard coaching is a killer.”

Domino’s Pizza’s transfers, promotions, and scrambles led to better empathy, stronger teams, and a challenge culture. “We have a lot of really nice people who play well with each other, have a high level of trust, and challenge each other.” That challenge is important. James Mattis said to protect the mavericks in your service, the ones that upset people in the bureaucracy, “because if they are not nurtured in your service, the enemy will bring their contrary ideas to you.”

Good arguments, said Geoff Mulgan, are “one of the things which is essential to collective intelligence.” Warren Buffett likes Carol Loomis because she’ll tell him when he’s full of baloney.

But good arguments take time and respect, two things Domino’s cultivated as they grew from within. Doyle said, “If you trust the people you work with you can get twice as much work done.”

With the plan for new roles and experiences as well as employees that respect each other, Domino’s Pizza was ready for a decentralized command structure. “Decisions that come to me that have an 80-90% probability of being correct should not have gotten to me, that decision should have been made a long time ago.”

Alex Blumberg was in a similar situation to that of Doyle when he started Gimlet Media and immediately tried to arrange a workable structure. “I don’t have the right information to make every decision so you have to set up a process for who’s making a decision and how they are making it.”

Doyle said about a hypothetical one million dollar investment with a 50/50 chance of working out but a ten million dollar payoff, “We will do that all day long.” He wants the Domino’s employees to be empowered. He needed to provide top-down support. Here’s Kevin Arnovitz of ESPN on this idea:

Chris Douvos got this advice, “But David Salem said to me, ‘I want you to be unafraid of being wrong and alone because if you’re unafraid of being wrong and alone every now and then you’re going to be right and alone and that’s the box where fortune and glory reside.'”

When Domino’s decided to change the Pizza Doyle said, “The folks were told that you can change anything, everything is on the table. And two years they came back with the answer and the answer was that they changed everything.”

And he went with it. That’s what Netflix does too. Ted Sarandos said that they don’t use their data to make shows, they use their data to find directions and director but once something is done they put it on the platform as is.

Domino’s new pizza didn’t have to be great. Taste is not their point of differentiation, service was. The pie had to be good enough.

Quantifying ‘consumer preferences’ makes it seem real, tangible. It’s not. Humans are funky and full of contradictions interpersonally and intrapersonally. Quantifying leads to numbers and numbers lead to mathematics which interacts terribly with psychology.

The Wharton Moneyball often notes the non-transitive nature of sports. If Team A defeats Team B and Team B defeats Team C then Team A must be better than Team C. But based on the labyrinthian tie-breaker systems the different leagues have this can’t be true. Ditto for consumers.

Consumers ‘think fast’ and most decisions are binary, like pass/fail in college rather than graded.

Domino’s Pizza didn’t need to enter the top-quartile of pizzas so much as be more edible. Many cheeses, doughs, and sauces later they had a New and Improved Pizza.

But ‘New and Improved’, “…is wallpaper. People don’t pay any attention when brands say, ‘We’ve got this new product and it’s new and improved.'”

Okay?

“One of the lessons is that great communication uses tension.”

Domino’s Pizza needed breakthrough (attention) and persuasion (action). “We finally landed on honesty, to accept the criticism and play it right back to consumers.” And this, the most important thing in the Domino’s transition, “What we think about our pizza doesn’t matter. What our customers think about our pizza matters a lot.”

Doyle points out that Domino’s brand is like a Modern Monopoly, it is what the customers say it is. Brian McCullough said this about eBay, “They were the first company who succeeded with the business model of the platform, of whatever the users are doing.”

Domino’s Oh Yes We Did campaign was one of honesty, action, and tension. But the effects went beyond the marketing. “What it did culturally was probably more important than what it did for consumers. If you’re trying to create a culture where people are comfortable taking risks, the most powerful way is to do it publicly, to consumers, and that is who we are (publicly and privately) and that energized the organization.”

With a new pizza and a new plan Domino’s created this ad:

“We put this ad on the air and our sales were up double-digit the first week.” It was much bigger than Doyle and his team expected and it showed them something else. “People trust each other. They don’t trust big institutions.”

Doyle said in the old days they could spend money “and we could tell people what the truth was.” That’s not true anymore. Now, “What consumers say about the brand is the brand.”

Social media now leads the thinking at Domino’s marketing. But they still follow David Ogilvy‘s advice and talk to consumers.

When Domino’s expanded their menu to include artisan pizzas they noticed that it was a healthier offering. Could they sell that? Nope. “As we tested the different ways of marketing it, the taste as the primary focus was the most compelling.” People want pizza that’s easy to get, doesn’t cost too much and tastes good – whatever that means to them.

Part of ease means tech and Domino’s is looking more like a technology company. “Over half the people working in our headquarters work in technology.” Doyle also said, “We now base our decisions around customer behavior and we test everything.” That’s something Andrew Ng would be proud of, who noted that being a technology company means acting like a technology company.

Chris Dixon and Fred Wilson talked about the difficulty of finding the next killer app but with examples like Domino’s we get a glimpse of what it kinda-sorta looks like. It’s native and does things in a new way. Like ordering, “We will look back in ten years and say that thumbing things into a tiny screen is the most absurd interface.”

So Domino’s is invested in voice. So Domino’s has Hot Spots. So Domino’s will A/B test new ideas. They removed the internal silos and the company flourished by getting feedback, pushing ideas up, and embracing social.

 

Thanks for reading.

Kareem and Wooden

Supported by Greenhaven Road Capital, finding value off the beaten path.

When Kareem Abdul-Jabbar joined Tyler Cowen on Conversations with Tyler it was an odd pairing. Sure, Cowen’s interests are wide-ranging, but Kareem as one of the first guests? The only incongruous thing was my understanding.

Kareem Abdul-Jabbar has lived quite a life, growing up around Jazz greats in NYC, playing basketball during a golden age of college basketball and the early days of the modern NBA. He’s lived, reflected, and written about being an African American. He’s been awarded the Presidential Medal of Freedom, just like John Wooden, who was like a second father to him.

We’ll pull quotes, ideas, and lessons from the book Coach Wooden and Me.

Skill and Luck. The two-jar model is handy for identifying outcomes that tend toward luck and outcomes that tend towards skill. Kareem was lucky to be tall was not born to play basketball and preferred baseball as a kid. As a ten-year-old, “I had a better chance sitting on the ball and hatching a unicorn than making the ball swish through the hoop.”

The next two years, “I spent much of fifth and sixth grade riding the bench.” By the seventh grade, “I was no longer an XXL benchwarmer.” Yet, as a freshman, “my style of play reflected my personality: politely passive.” Eventually, Kareem became the NBA’s leader in career points.

Not so fast, John Wooden might say if he were still alive. In the two-jar model, outcomes = skill + luck, and prioritizing outcomes means relying on luck. That’s a bad feedback system. Kareem explained, “Coach Wooden’s most important lesson was that we should never focus on the outcome but on the activity itself.”

Success wasn’t about winning, winning was a byproduct of success. Wooden said, “Just do everything possible to prepare. As long as you know you have done everything possible and you have given your best self on the court, that is your reward. The scoreboard is meaningless.”

And, “If you get yourself too engrossed in things over which you have no control, it’s going to adversely affect the things over which you have control.”

This is why, Kareem writes, that Wooden didn’t like sports movies, because the team that learned the lesson often would win. “His (Wooden’s) point was that the life lesson is the success. The traveling is the reward, not reaching the destination.”

Kareem arrived after UCLA because he wanted to be there. John Wooden prioritized aligned stakeholders.

It’s not only rowing in cadence but recruiting people who want to be on the boat. Wooden told Kareem, “I wanted young men who wanted to play for UCLA and not one that I had to talk into playing for UCLA. I always believed that the way to build a great team is to find the kind of people you want to work with and tell them the truth.”

UCLA was able to recruit people who wanted to be there. This is what Josh Brown does through blogging. It’s what Warren Buffett does through letters. And the form matches the recipient. Here’s how Cade Massey explained one version of this:

One theme through Kareem’s life was the racial changes and lack of changes in America. Through his involvement, he met people like Muhammed Ali who told him, “When you saw me in the boxing ring fighting, it wasn’t just so I could beat my opponent. My fighting had a purpose. I had to be successful in order to get people to listen to the things I had to say.”

Kareem heard plenty of racial taunts at basketball games. He told reporters, “Bitterness gets in your way, you get involved in revenge instead of trying to create a change. I used to be bitter, but now I just play hard to win.” And after one game, “I’d already delivered my presentation of racial equality earlier that evening when we crushed their team.”

If bull markets conceal bad investors maybe the opposite is also true; times of stress can reveal people of remarkable character. Kareem Abdul-Jabar is one of those people.

Thanks for reading.

Streaming, Sharing, Stealing

Supported by Greenhaven Road Capital, finding value off the beaten path.

Michael Smith and Rahul Telang of Carnegie Melon University wrote an interesting book, Streaming, Sharing, Stealing: Big Data and the Future of Entertainment.

Smith and Telang wrote this book because of a “perfect storm” in the economic entertainment ecosystem. Traditionally a few firms dominated because the best business model for the last hundred years was a return to scale. The value was in finding/attracting talent, creating a market, and distributing the product.

screen-shot-2019-01-08-at-2.51.19-pm.png

This model worked. This model was logical. This model was successful.

Before 1995, the search for talent “was unscientific at best” and size helped to seek talent, sell it, and ship it. The Firms did this well and movies are one example. ‘Windowing’ is the process of a theatrical release followed by DVD sales followed by digital sales followed by subscription streaming followed by cable and broadcast rights.

Each stage of price discrimination maximizes revenue. People who value movies highly see them early, people who value movies less seem them later. Streaming, Sharing, Stealing is a book now because technology has evolved and the old windows are closing.

Blockbuster was founded in 1985 and The Little Mermaid began the ‘second golden age of animation’ in 1989. Shawn Fanning created Napster in 1999. Each of these was a locally logical decision at the time. It made sense to start Blockbuster, it made sense to make The Little Mermaid, it made sense to start Napster. Each is like a mile marker on the highway of entertainment history.

Smith and Telang note that the Netflix/HBO/Amazon/Spotify/etc. state of today was a sort of perfect storm. We should be wary of simple stories like Napster ruined everything, and be more like Sanjay Bakshi and approach things with a ‘part of the reason is this’ mentality. Smith and Telang propose three things.

First is broadband internet, “It became harder to delay a product in one channel versus another channel because people found it easier to infringe.” Gatekeepers Firms had to control distribution so they could control the windowing of theater to dvd to rental and so on.

Second is digitization. “Now you had technology change happening that eroded the role of a gatekeeper at these firms.” Casey Neistat is the vlogging example. E.L. James is the writing example. The digitiziation pendulum might have swung too far:

Our third change is the rise of platforms, because, “they didn’t have limited shelf space.” Frank Underwood kills a dog yet Netflix lives on because there’s always something else on.

At this point we should pause for a moment and recall something Andrew Ng said. Ng told the story about the early days of the internet and what it meant to be an internet business. It’s the difference between being a tourist and being a resident. “What defines an Internet company is whether or not you have an architect at your organization to leverage internet capabilities to do the things that the internet does really well.”

What things does an internet company do? Internet companies, Ng said, AB test, push decisions up (not down), and have short product cycles. We’ll see a Spotify example shortly.

Michael Smith is emphatic about this because it’s a perpetual point in his presentations. The genius for House of Cards was not knowing that Kevin Spacey, David Fincher, and Robin Wright were good at their jobs.

“The advantage Netflix got was from using its data was to do something that no traditional broadcast network could replicate. Netflix’s advantage didn’t come from knowing how many fans of Kevin Spacey were in the audience. Netflix’s advantage came from knowing exactly who they were as individuals and promote content to them directly based on their individual preferences.”

That’s what it means to be a digital entertainment company.

Netflix made nine trailers for House of Cards; some for Fincher fans, some for (former?) Kevin Spacey fans, some for fans of shows with strong female leads. None were for dog fans, dog fans won’t like House of Cards.

Smith recalled,  “Michael Eisner said that if he had tried to include a similarly violent scene in an episode of broadcast television he’d be out in ten minutes.”

Why could Netflix show House of Cards but ABC could not? Unlimited shelf space. ABC has a shelf space of one. This point is made by Ben Thompson at Stratechery too, pointing out the Netflix advantage when they signed a deal with Starz. Primetime evidence exists too, as we see shows about doctors, law enforcement, families, and more doctors, more lawyers, and more families.

Netflix serves niches.

Music is both like movies and not like movies. Hal Varian noted that “music is something you want to hear over and over again. Movies you want to see on an occasional basis with one exception, kids movies.”

They’re also different in consumer preferences. For music, live is the premium experience. For film, premium means polished. No one pays to see Daniel Craig run across a stage but many love to attend Phish shows just to be surprised.

Knowing that music and movies are different and knowing that Gatekeeper Firms and technology companies are different we’ll look at Spotify’s “discover weekly (Twitter)”.

Former Spotify engineer Chris Johnson said, “Discover Weekly wasn’t handed down from the top. There are news articles about how this is the Apple Music killer and we’ve been planning it for months. No. This is purely an organic bottom-up feature that a small team of four engineers worked on.”

In his talk, Johnson explains the iteration and small steps and, in hindsight, the obviousness of it. Their ‘that’s interesting’ moment was when “We noticed users spending more time on editorial playlists than they were on their personalized discover playlists. What’s going on here?” Didn’t people leave the radio DJ for Spotify?

It was 2014 and Spotify created the first personalized playlist. “It turned out that a month later millions of users are still listening to their personalized Play it Forward playlist. We’re like, ‘Holy crap, what’s going on here?’ We did not expect that.”

That’s when Johnson and his team realized, “The same content as the Discover Page but…a playlist.” People wanted, expected, used playlists more than pages. They tested and tweaked. They released internal versions. They changed the cover art of the playlist.

Then Apple pushed the U2 album onto iPhones. Uh oh. That’s kinda what Spotify was doing, pushing playlists onto the app. Johnson included a survey to Google Docs asking for feedback. Only three people out of a thousand brought it up so they kept building. “Users know best. AB test everything.”

Another part-of-the-reason Discover Weekly works is that it’s human curated. Not one person editorially but a person sporadically. “All the models we’re building are built using human data, on what users are streaming or curating. We’re just mining through the human element.”

In another talk, Mathew Ogle said “Every song on Discover Weekly is there because at least one other person on Spotify added it to their playlist. It’s not there because a machine thought it sounded good.”

Using Ng’s definition, Spotify is definitely a digital music company. Like Netflix, they’ve made it easy to pay for and listen to music and have provided added value using data and AB testing to the users.

But Spotify has other problems – “For Spotify the royalty payments also crush its gross margins.” – that Netflix does not. Marc Andreessen asked Ted Sarandos about sports and Ted said, “It’s okay for us not to be doing everything…The reason I don’t get tempted by major league sports is that the pricing power all belongs to the leagues.”

Netflix and Spotify exist because consumer preferences changed. In How Music Got Free we see part of the story about how this happened. For example, CD’s were over-engineered. “Where the (CD) sales literature promised ‘Perfect Sound Forever,’ (Dieter) Seitzer saw a maximalist repository of irrelevant information, most of which was ignored by the human ear. He knew that most of the data from a compact disc could be discarded—the human auditory system was already doing it.”

And the industry dominated distribution. “Consolidation in the radio industry also helped, creating a homogenous nationwide listening environment that could propel an album to a Platinum status almost instantly on the basis of a single hit. Controlling the airwaves was critical—if Limp Bizkit could go forty times platinum, then literally anyone could.”

In 1996 the compact disc retailed for $17, cost $1 to manufacture, and had approximately 1.8 good songs. Music was ripe for Christensen’s disruption. It was true that CD’s were easier to buy, sounded better, and had more ways to listen. But the MP3 was better in certain ways and the shortfalls could be improved.

It was technological advances that led to 1,000 songs in your pocket.

It was analytical advances that led to Discover Weekly.

It feels like circa 2019 there’s more room to improve analytically than technologically. Netflix uses analytics for strategy but not scripts. Ted Sarandos, “We’re way better off taking someone’s creative vision and putting it through the service than us trying to go in and retool it. At the end of the day if the creator says, ‘That’s my show.’ we put it up.”

Smith quotes Frank Underwood in House of Cards, “Power is a lot like real estate. The closer you are to the source, the higher your property value.” Netflix is darn close to the data source.

This approach is appropriate because Netflix has unlimited shelf space. Smith notes that they can invest in properties with niche audiences like those for Woody Allen, Adam Sandler, and Arrested Development. Netflix wants to offer one thing you love at any given moment. ABC wants to offer something most like at one particular time. “Netflix didn’t face the same scarcity that traditional broadcast networks do. In a traditional broadcast network, you can only deliver one show at a time so that show has to appeal to as many viewers as possible. But a Netflix subscriber who was offended by Kevin Spacey’s actions could choose from thousands of other shows on Netflix.”

“Netflix doesn’t have to find more viewers to watch the individual programs. Netflix has to find more programs that appeal to individual viewers.”

The stakeholders matter too. “Netflix wasn’t pursuing an advertising-supported business model so it didn’t have to worry about offending advertisers by including a potentially controversial scene.”

This trio of broad band, digitization, and platforms is the media subset of what Michael Munger writes about in Tomorrow 3.0. In the book, Munger explains the three T’s of transaction costs that prevent economic exchange; triangulation, transfer, and trust.

In other words, I’ve got to find what I want, I’ve got to get what I want (and pay the person for it), and I’ve got to trust the other person. That’s what the digital behemoths have done. Netflix make it easier to find what we want – even though we didn’t know we wanted it – because they make nine different trailers for a tv show Netflix is able to get us what we want legally by offering an automatic payment system. Netflix creates trust that they’ll provide something because they have so much.

Teaching this is fun, says Michael Smith. Students are digital natives and they get this.

“With higher education for example, we have nothing to worry about when it comes to technology changing our business and our power. The students do exactly what you all just did, they nod their heads then go ‘Wait a second!’.”

How so? “There’s a lot of parallels between what happened in the entertainment industry and what might happen in higher education.”

A transition from gatekeepers who succeeded thanks to the collection of talent, the creation of markets, and the distribution of content to digital companies with platforms and unlimited shelf space. Austen Allred has one theory of education. Our XMBA post is a looser approach.

Digitization changed some parts of the economy more than others. Information goods like newspapers, movies, and music all became non-rival as copies were cheap to make and share. Thanks also to network effects, ideas went viral and sharing was easy.

Ben Carlson observed this:

“Why 45 million ppl watched Bird Box on Netflix: holidays & ppl need entertainment, slow TV season before new episodes/shows come out in Jan, no good new movies released. So ppl basically had no other choice but to watch this I Am Legend/World War Z/Quiet Place reboot.”

Entertainment (and the economy) isn’t better or worse, only different. It’s like a hike where the terrain has changed but the direction towards good enough has not.

Sources: Smith and Telang gave a very nice talk at Google (https://youtu.be/ozdfYIALhnY), the book goes into more detail (https://amzn.to/2B8RV5O) and this lecture at UC Irvine about teaching this ideas also has some nuggets (https://youtu.be/u7f1RYXdUZI).

For Spotify, Mathew Ogle spoke in 2015 about music discovery (https://youtu.be/81OlAEnyTKE). Chris Johnson gave a talk in 2016 in Austin about Spotify’s Discovery Weekly history that was insightful (https://youtu.be/HWFYA6N9uA0).

Greg Lindsay

Supported by Greenhaven Road Capital, finding value off the beaten path.

Greg Lindsay’s talk, Innovation Doesn’t Happen at Your Desk is an attempt to get people to work together and better. I think he’d agree with Rory Sutherland in that many of the logical questions have been answered and we need to get a bit weird.

A more positive term for weird is serendipity.

Lindsay tells of a natural experiment where a building needed asbestos treatment. The  teams there had their offices shuffled about, and when their (innovation) results were measured, “Randomness was the best strategy for increasing innovation for these teams.”

Put different people near each other, let them talk, and watch the good ideas flow.

Part of the explanation is The Allen Curve, an observation that people talk more to those nearby, even when digital tools allow us to talk to anyone anywhere. Dr. Thomas Allen said, “Broadband communication isn’t a substitute for face to face.” That is, people communicate with those they see.

If proximity leads to communication then communication leads to serendipity. Examples include the Nike Waffle Shoe, Spotify’s Discover Weekly Playlist, and Reed Hastings (who got lucky serving coffee to a certain computer lab).

Lindsay said, “All three major types of artificial sweeteners came from lab accidents where the researchers then picked up their lab bench and stuck their fingers in their mouths.”

Jenna Fischer agrees with action as a solution, “…you’re much more likely to be in the right place at the right time if you’re busy doing showcases, plays, and taking classes. Chances are you won’t be in the right place at the right time if you’re spending your days eating Lucky Charms your couch. Trust me, I tried.”

There are four “black boxes” for serendipity Lindsay said:

  1. You
  2. The office
  3. The city
  4. The network

YOU

Building on ideas from James Lollies and Joey Ito, Lindsay said, “You know the classical line, the true sound of scientific discovery is not Eureka but ‘that’s interesting’.” And “Do you have the latitude to chase something?” It takes an internal curiosity and external circumstances.

That’s interesting moments surround us. Peter Rahal wondered why a Crossfit sold t-shirts but not protein bars. Eric Maddox wondered why certain people were still hanging around an area. Brian Koppelman wondered why Americans celebrate entrepreneurs. Jim O’Shaughnessy wondered if personalities weren’t the perfect proxy for well-performing investments.

THE OFFICE

Cubicle offices are too much like tree farms, orderly and great for optimizing harvests. Another way to approach Lindsay’s idea is via Christensen’s disruption model; optimized offices are best for sustaining innovations while open offices (and mandates) are better for disruptive technology.

“Google started a beekeeping club so that engineers who are interested in beekeeping might meet each other and actually have discussions about unrelated subjects.”

Pervading the office furniture is the office culture, and perks are not culture. Ben Horowitz said that culture is what people do when you don’t tell them what to do. Does your boss encourage serendipity? Do you?

THE CITY

Lindsay cites Geoffrey West that cities get better as they get bigger.

He also (seems) to like the work of Jane Jacobs and mentioned a Sante Fe panel/study along with this; Being Nicely Messy. While YOU and THE OFFICE are bonsai-like cultivatable, the city is more emergent. “Cities are the greatest serendipity environments of all so Google and Facebook are trying to design these environments that are more walkable and allow all these sorts of spillover effects.”

THE NETWORK

Org charts often differ from communication networks.  Albert-Laszlo Barabasi has done similar work to what Lindsay cites. Barabasi noted that the people best connected in a network are rarely the decision makers, it’s the safety management supervisor who has to visit each office and likes to talk.

Other examples from the talk include:

 

Thanks for reading and happy serendipitying to you.

Brynjolfsson and McAfee

Supported by Greenhaven Road Capital, finding value off the beaten path.

For a long time, there were only small changes in human existence. Then James Watt was asked to fix a steam engine. Watt fixed, tinkered, thought, experimented, and (is credited) with the invention of the double acting piston. Then the industrial revolution took off. Watt’s invention is a GPT (general purpose technology). Christened to remove water from mines, it was implemented in many other places. Subsequent GPTs were railways, the internal combustion engine, and electricity.

Technology becomes a GPT and not just a T when it’s applied to adjacent domains. In the early days of electricity, factories kept their layouts but changed their fuel. Electricity became an OG GP with the conveyer belt.

From there we get the Model-T, which benefited from another GPT, the combustion engine. Then we got the internet. As Andrew Ng noted, malls with websites weren’t Amazon competitors. What makes an internet company said Ng, is acting like an internet company. One example is Spotify and the Discover Weekly Playlist. AI may be another GPT, because Brynjolfsson said, “The core AI breakthroughs have applications in just about every part of the economy.”

This doesn’t mean the end of work. “It’s too soon to worry about the end of work…it may be that someday we will be able to make machines that do the full spectrum of what humans can do but that’s not the challenge we face today.” Ng said to think of tasks, of “anything a typical person can do with less than a second of thinking we can probably now assume automate.” Ajay Agrawal said much the same, “AI doesn’t do workflows, it does tasks,” and “Think of the recent advances in AI as advances in prediction; better, faster, cheaper prediction.”

Brynjolfsson isn’t worried about jobs so much as returns from work. Median income is flat since 1999 and though the economy has grown, the top twenty percent (based on wealth) earned more than 100% of the gains. They are the Apple iPhone of the economy.

What’s allowed this? People do more with less. Facebook bought Instagram when there were fifteen employees. When photography was a chemistry problem, Kodak occupied the lord’s estate and 145,000 serfs kept the kingdom running. When photography became an arithmetic problem Kevin Systrom & Mike Krieger stormed the gates with a band of merry men. Though it wasn’t only men, it was Systrom’s wife, neé girlfriend, who contributed an essential idea.

Instagram’s success is emblematic of the digitization bull that’s gone through the economic china shop. Along with easy copies are zero marginal cost distribution and the rise of the individual. Steve Jobs got this wrong, noted Brynjolfsson’s co-author, Andrew McAfee. Jobs mistakenly wanted the app store to remain closed. But the iPhone’s ascent accelerated when it opened, creating what Alex Moazed calls a Modern Monopoly.

Though we do more with less, “it would take the average American only eleven hours of labor per week to produce as much as he or she produced in forty hours in 1950,” this doesn’t mean the end of jobs is nigh. The Second Machine Age is an optimistic book about work.

AI, like other technologies, can supplement humans. Thanks to algorithms we’ll stick with base rates rather than have overconfidence in our abilities. McAfee said, “The single biggest failure mode that I see when I talk to smart people is that smart people tend to have exaggerated version of a failure mode, to be too confident of their own judgment.”

Then there are the jobs that won’t be too affected. In a study of 964 O*Net occupations, most had 20-30 tasks and most of the tasks were things algorithms won’t or can’t do. “When you look at all the tasks of a particular occupation, some of them were suitable for machine learning but many others were not. There was no case where machine learning just ran the table and was able to do the whole set.”

We talk about truck drivers being automated but truck drivers do a lot. Finn Murphy wrote that he was not only a driver but a manager, counselor, and in one case, an honorary Native American. Hal Varian makes a similar point when he points out that only one job has been automated away, the elevator operator. But, the tasks from that job have shifted to other areas, receptionists, concierges, and hostesses all do more.

Education might help. “There’s an increasing need for interpersonal skills…It’s not that work is disappearing but that there’s a whole bunch of tasks that only humans can do and we need to shift our skillset into those and then labor income is likely to go up.”

In another talk Brynjolfsson explains, “Right now and for the next ten years training and education are probably at the top of my list and most economists…I don’t think our schools are doing a very good job teaching those, or worse yet, many of them are actually crushing them.”

Seth Godin wants schools to teach students how to lead and how to ask interesting questions. Bill Burnett at Stanford (https://youtu.be/34EuT2KH2Lw) said school cripples our creativity, “Something happens, mostly our educational system which promotes a massive fear of failure and a search for only one right idea.”

Rory Sutherland noted that the opposite of a good idea may be another good idea, which is contrary to the one-right-answer attitude in some schools. Pat Dorsey praised the liberal arts approach to education; “In a way equity investing combines a lot of different fields in that it is a set of ever-shifting problems to be solved. It’s not electrical engineering where there’s a right answer to everything. There is never a right answer to what the company is worth, or what is the competitive advantage to a business, or how much cash will they generate in five years.”

Which brings us to Erik Brynjolfsson and Andrew McAfee’s, Machine, Platform, Crowd. They wrote this one because “People who run companies kept approaching us in the hallways of places like Davos and kept on saying, ‘I believe your story, now what do I do?’ In some cases, I got the impression that there was desperation behind the questions,” explained McAfee.

Brynjolfsson and McAfee’s solution is in the title; collaborate with machines, harness platforms, and use the wisdom of crowds.

They wrote the book because “The narrative these days is that technology is killing jobs and it’s not too far from that narrative to ‘therefore technology is a bad thing’. We think that’s a really harmful road to go down.” Paul Daugherty agrees, “I’ve started using the term ‘collaborative intelligence’ rather than ‘artificial intelligence’. The problem with AI is it scares the public and it leads to the wrong discussions.”

Another stumbling point is making GPTs general. Malls with websites aren’t Amazon and record stores online aren’t Spotify. The issue is that “Tech progress rewrites the business playbook.” McAfee said, “I see a lot of companies underestimating the power of machines a lot of companies underestimating the power of platforms and a lot of companies not doing a great job of tapping into the crowd.”

Okay, so how can someone use these ideas?

For machines, we need to get away from HiPPO decisions (https://youtu.be/WyYaubtZuX8?t=391) and move to Geek decisions. This is sabermetrics in baseball like the Oakland A’s. This is running 10,000 experiments a day like at Google. This is who to market new shows to like at Netflix.

For platforms, the key is the interface. How can an organization create an easier way for customers to find X? How can an organization create easier ways for producers to create X? Successful platforms solve for Michael Munger’s three T’s; triangulation, transaction, and trust.

For crowds, the key is trusting weirdos. McAfee said that the core can be bested by the crowd. “I honestly just mean these hundreds of millions or very soon billions of complete stranger weirdos out there available on the internet that you can tap into where you can access their knowledge and their tenacity and their energy if you do it correctly.”

Yet again we’ll remind ourselves that technology is a tool like a shovel not a panacea like a silver bullet. It’s still up to the user not to hammer their thumb. Machines, crowds, and platforms are wonderful compliments for decision making. McAfee said, “The failure mode among really smart people is to trust themselves too much.”

No one knows how fast work will change. We’ll get glimpses of pieces first. The duo gives examples like a smart grid in a series of factories or an autonomous shipping route from Dallas to Los Angeles. While Brynjolfsson wants more training, McAfee wants changes to the entrepreneurship policy and physical infrastructure.

Neither seems to think work is going to go away in the next thirty years, and McAfee points out work is more than a job. “I’ve become a fanatic about work, about the value and importance of something like a job. The evidence is overwhelming that when work leaves a community bad things happen.”

Roger Lowenstein wrote about Warren Buffett, “(he) understood that most people, regardless of what they say, are looking for appreciation as much as they are for money.” Voltaire concluded, “Work keeps at bay three great evils: boredom, vice, and need.” Adam Smith suggested that man wants to be loved and be lovely. Andy Grove wrote this:

“I felt the frustration that comes when the things that worked for you in the past no longer do any good.”

“Businesspeople have emotions, and a lot of their emotions are tied up in the identity and well-being of their business.”

“In many instances, your personal identity is inseparable from your lifework.”

Change, especially when your personal identity is at stake, isn’t easy. Brynjolfsson and McAfee cite the work of Angus Deaton and Ann Case who coined ‘deaths from despair’. “I think about which of those social problems will be fixed by a magical check from the government showing up every month,” McAfee said, “and my answer is basically none.”

Jobs give more than money.

We don’t know how the future will be different only that it will be different. In that spirit we’ll take our final word from Andy Grove:

“You need to plan the way a fire department plans: It cannot anticipate where the next fire will be, so it has to shape an energetic and efficient team that is capable of responding to the unanticipated as well as to any ordinary event.”

 

Thanks for reading.

Final Jeopardy

Supported by Greenhaven Road Capital, finding value off the beaten path.

Final Jeopardy by Stephen Baker is the story of IBM Watson, the Jeopardy champion who defeated Brad Rutter and Ken Jennings.

In December we looked at AI with posts from Andrew Ng, Paul Daugherty, Hal Varian, and Avi Goldfarb and Ajay Agrawal. The book Final Jeopardy is a polaroid for AI. It’s a precursor to those blog posts. A sort of this is how things were.

In the late 1970’s IBM crested the technology wave and then big blue began a descent as Microsoft rode the new wave of PCs. IBM’s mainframes were a vertically integrated business model and the era of PCs, horizontal integration ran roughshod over a prostrate Big Blue.

By 1992 the company was saddled with a 5B$ loss and their shift to software and consulting began. In 1997 IBM’s Deep Blue defeated Kasparov in chess and in Harry Friedman was promoted to Jeopardy producer and began to include more popular culture and current events in the show.

He thought he was going to figure out the computer. That BBC video demonstrates something Rory Sutherland rails on; that creativity means solving problems in new ways. Kasparov resigned after 19 moves becuase the computer solved the problem in new ways.

Aside from Trebek shaving his mustache (2001) and SNL Celebrity Jeopardy, Jeopardy, America’s favorite quiz show stood stolidly with Wheel of Fortune as American staples.

In 2003, the five-show limit was lifted and Ken Jennings took the game by storm, drawing 15 million viewers – according to some – igniting a question at IBM. Could we beat him?

Watson was part curiosity, part marketing, part interestingness. Baker wrote, “IBM’s biggest division, after all, was Global Services, which included one of the world’s largest consultancies. It sold technical and strategic advice to corporations all over the world. Could the consultants bundle this technology into their offerings?”

Internally there were some high hopes. They just had to make it and that proved damn difficult.

“The biggest obstacle (early on in AI) was language.” In addition to that, there was no consensus on how to approach trivia questions. Some researchers want general AI, others craved specific. Reading a book written in 2011 about events in 2009 in 2018 gave one perspective but it helps to remember that whatever is being done at the time doesn’t seem easy to the people doing it. We all Google all day, but Baker does a good job articulating all the issues the team faced.

In 2005 IBM starts considering the challenge in earnest and one engineer is tasked with an off-the-shelf approximation. We see this advocated in companies from IBM to IDEO, there’s huge value in prototyping. Before jumping in, assemble a simple solution. IBM discovered that their first internal version was much better than off-the-shelf software. They moved on.

The IBM team used stuff they had worked on but also invented new ways to parse data. They also created new ways to work together. One key to Watson was parallel inquiries. Once the system figured out what a question asked; the sausage celebrated every year since 1953 in Sheboygan Wisconsin, it would try to come up with a collection of answers.

Ideally, some of the answers would be the same and quantity lent credence to quality. It was a sort of wisdom of crowds, what does the collective algorithm suggest? And the team needed to work this way too. “Ferrucci decided to take the same approach with his team. He would cluster them. He found an empty lab at Hawthorne and invited his people to work there.” They had to be together.

This might be true beyond teams like the Jeopardy engineers. What if schools work so well not because of economies of scale but network effects? I give you a good idea and you give her a good idea and so on. What if we should prioritize the gestalt rather than the CBA.

Vicinity also kept values aligned. Small groups don’t need mission statements so much as communication. The Watson team didn’t necessarily want Watson to Win. They wanted Watson to demonstrate IBM’s prowess. That included not looking bad, the team worried that “a humanoid Watson might frighten people.” Watson needed to represent IBM well, and not embarrass them. It could be funny, but not profane. Early answers examples like “This is the type of diet grasshoppers eat – Kosher” brought laughs. But in the category of Just Say No, ‘This is a four-letter German word.’ Watson’s gave the answer ‘Fuck’.

Oops. IBM staffers chuckled and began work on a profanity filter.

Straight facts were the easiest questions. “There were factoids, each one wrapped in the most helpful data for Watson: hard facts unencumbered by humor, slang, or the cultural references that could tie a cognitive engine into knots.”

But Jeopardy is full of puns and ambiguity. “At low levels of confidence, I think we’ll just have it say it doesn’t know, sometimes that sounds smarter,” said Chu-Carroll. Watson needed to look like a helpful took, not an unhelpful HAL.

What’s clear from the book is that Watson took a lot of work to work well in a limited domain. What’s also clear since then is that technology has gotten quite good at working well in many different domains. From grammar to adaptive cruise control, it seems like technology mostly works and mostly makes life a lot better. One step along that way was Watson.

 

Thanks for reading.

Brian McCullough

pablo-13.png

Supported by Greenhaven Road Capital, finding value off the beaten path.

Brian McCullough spoke at Google and podcasted with Chris Dixon about his book How the Internet Happened. McCullough said he wrote the book  because “I feel like anybody would want to know who did this” and “It just sort of bothered me that no one had done a book about the internet going mainstream and I just got tired of no one writing that book.”

These glitches in the Matrix where something confuses us, enlightens us, or bothers us are where a lot of products begin. It’s the catalyst for products like RX Bars, Panera Bread, and Spotify’s Discovery Weekly. “So like any other startup I’ve done, I was like, ‘Well someone’s gonna do that so why not me?'”

McCullough found some interesting things at the library, like Al Gore’s contribution. During interviews, he also noticed some media manipulation. “To a person, all the Mosaic/Netscape engineers made the point that this sort of work hard, sleep under your desk, startup culture was from the media. These were corn-fed Midwesterners.”

The media distorts things all the time. Ed Catmull noted the caricature of Steve Jobs. Annie Leibovitz wrote about the manipulation of scenes for war zone photos.  Andre Agassi never planned on saying image is everything.

But maps distort things. So do stories. Memories too. Nothing is a perfect replication.

As countermeasures, we can adopt Tyler Cowen’s approach and avoid the philosophy of once-and-for-allism and heed this advice from Jason Bateman:

“Baked into the cake is the assumption that the people reading this are reading this for fun. You kinda know that with those magazines there’s more than a grain of salt, more like a kernel of truth to it, but its super hyperbolic to get the magazine sold or the click-bait on the computer.”

Netscape’s browser was the first major software to be distributed digitally. McAfee and Brynjolfsson and Smith and Telang researched why this shift was important. McCullough reminds us, “Before Netscape, you would still put software in boxes, shrink wrap it, and put it on a shelf.” I remember getting Warcraft in a box and thinking, that’s a lot of space for one disc and a manual.

Netscape was also the first internet company in the sense they acted like – A/B testing – an internet company. Andrew Ng said, “What defines the internet company is whether or not you have an architect at your organization to leverage internet capabilities to do the things that the internet allows you do really well.”

Part of the reason Netscape did this was out of fear. “The research was a surprising reminder to me how much everything in the 90s was in relation to what Microsoft might do.” Read anything about technology in the 90s and Microsoft comes up. Tim Wu compared Microsoft to Kronos, eating the young. Jerry Kaplan wrote “If we so much as threatened to sue, it would rain lawyers on our office like the plagues of Egypt.” “Googlers told me that when AdWords finally took off they didn’t tell anybody until they were far enough along so that Gates can’t catch us,” noted McCullough.

Gates’s mistake was in thinking TV was the portal and broadband was the means. “The thing that Gates missed was that the internet was good enough.” McCullough told Dixon, “The lesson of the early internet is that sometimes just good enough technology is good enough.”

This was true for Twitter and the fail whale. Instagram too. Kevin Systrom said, “To be clear, there is no reason we should have succeeded. The server was down every other hour and people just kind of forgave us. They came back and they would share their photos. At the time mobile networks weren’t that great either, so people would blame their connection and not us – which was great.”

McCullough’s talk and book include many other companies; Facebook, AOL, Yahoo, and Amazon among others. eBay too, which did three things.

  1. “Taught normal people to trust faceless strangers across the country online.”
  2. Created “masses that are self-organizing,” and introduced a rating system.
  3. “They were the first company who succeeded with the business model of the platform, of whatever the users are doing.”

McCullough says that newspapers were interested in buying eBay but they had no inventory. Chris Dixon noted that what the publishers thought was a bug was actually a feature. Related was Alex Rampell’s Tivo problem.

At one point, 7% of all eBay listings were for Beanie Babies, and Zac Bissonnette wrote about the toy’s importance, “In order to lure people out of their comfort zone and into the idea of e-commerce—and, even scarier, peer-to-peer e-commerce—eBay needed to offer consumers something that was easy to ship, couldn’t be found anywhere else, and elicited passion among the people who were looking for it.”

More recently Michael Munger summarized the economics, noting that platforms reduce the transaction costs. He said, “The best proposition is where transaction costs are high but reducible and excess capacity is high.” Before eBay people used newsletters and local shows to trade toys.

McCullough also addresses the idea of a killer app. “I think the reason mobile didn’t take off for so many years was because there wasn’t anything for normal people to do with it. The iPhone comes out in 2007 and six months later Facebook opens registration to everybody.” The ‘opening’ was something Steve Jobs nearly got wrong.

Ben Horowitz addressed this idea too, “When asked ‘What’s the killer app?’ No one ever gets it right.” Horowitz didn’t give an answer but he did offer an approach. “If you think about the smartphone it was much worse than the PC. It had a tiny screen, it was far less powerful, etc, etc. But it had a couple of properties that weren’t in a PC. It had a GPS and it had a camera, so you could now build things like Lyft and Instagram.”

McCullough supposed, “Maybe you had to have Facebook to go mainstream for the iPhone to go mainstream and vice-versa.” There had to be ‘an app for that.’

6FBE94FF-927C-4FE2-876C-8D9468995B6F.png
Screenshot from iPhone 3G commercial.

We study history, said Michael Ovitz “Because past is prologue.” It also helps said McCullough, “how much can you learn by looking back at dumb ideas that will always be dumb ideas (inversion) or dumb ideas that were only dumb ideas at that time (timing matters).”

This a trope that this time is different. That’s true on some levels but not others. What never changes is an exchange, an X for Y. That’s the essential economic question. What always changes is the means, reasons, and explanations.

 

Thanks for reading.