Amazon and Managerial Incentives

Aug 04, 2020

Today’s newsletter features a list of questions you can use when discussing this essay with your team.

Plus, I will send a second email with links to some articles that I think are valuable (or at least interesting) for readers this week.

—Edward

Battling with billionaires

Last Wednesday, CEOs from four of the five most valuable companies in the US (Apple, Amazon, Google, and Facebook) testified before the House Judiciary Committee's antitrust subcommittee. For six hours, politicians grilled Tim Cook, Jeff Bezos, Sundar Pichai, and Mark Zuckerberg about their companies’ business practices.

The House Committee accessed internal emails from the companies. In some of the messages, executives communicated ideas and approved strategies that went against the public interest. Many media pundits and “thought leaders” viewed the hearings as confirmation of the companies’ aggressive tactics.

But we need to be careful when we anthropomorphize companies and think of them as single-minded organisms. Companies are, of course, made up of individual people. A LOT of people. As much as the four CEOs would LIKE to hold ultimate power over all aspects of their respective companies, anyone who has worked for a large organization knows that many individual people guide many different departments.

Yes, CEOs wield a lot of influence — they communicate vision, set direction, shift relative budgets, oversee senior-level hiring, and inspire corporate culture. Those are significant powers, but far from absolute control.

One example: Wells Fargo CEO John Stumpf wanted to instill a competitive culture that drove cross-selling of products to the bank’s customers. Stumpf did NOT, however, intend to build a corporate culture that motivated employees to open fake customer accounts (and break many laws along the way). How do we know — for sure — that Stumpf didn’t intend for these actions to take place? Because Wells Fargo (the company) was the victim of the Wells Fargo account fraud scandal! The employees involved in the scheme (and their managers, and their managers) were chasing the incentives established by the CEO, but they clearly violated the spirit of how the CEO (“the company”) WANTED them to reach their goals.

In the week before the congressional hearings, the Wall Street Journal published an investigation into the Alexa Fund — Amazon’s venture capital firm. The article claims, based on interviews with various entrepreneurs and investors, that Amazon strategically invests in companies to “help develop competing products.”

In some cases, Amazon’s decision to launch a competing product devastated the business in which it invested. In other cases, it met with startups about potential takeovers, sought to understand how their technology works, then declined to invest and later introduced similar Amazon-branded products

Here’s one example of an entrepreneur alleging Amazon’s use of invest-and-steal tactics:

When Amazon.com Inc.’s venture-capital fund invested in DefinedCrowd Corp., it gained access to the technology startup’s finances and other confidential information.
Nearly four years later, in April, Amazon’s cloud-computing unit launched an artificial-intelligence product that does almost exactly what DefinedCrowd does, said DefinedCrowd founder and Chief Executive Daniela Braga.
The new offering from Amazon Web Services, called A2I, competes directly “with one of our bread-and-butter foundational products” that collects and labels data, said Ms. Braga. [Emphasis mine]

Did you catch that detail?

Braga would have us believe that Amazon invested in her startup so it could learn more about her intellectual property and business plans — and then launch a competing product FOUR YEARS LATER! (Keep in mind that the average tenure for an Amazon manager is approximately 12 months.)

Let’s take a look at A2I, the Amazon-created product that Braga claimed was a direct competitor to her own company’s invention. Here is the description posted on A2I’s product page:

Amazon Augmented AI (Amazon A2I) makes it easy to build the workflows required for human review of ML predictions. Amazon A2I brings human review to all developers, removing the undifferentiated heavy lifting associated with building human review systems or managing large numbers of human reviewers.

In simpler terms, Amazon built a tool to facilitate the human review of machine-learning predictions. Amazon is not the only company playing in this space. Google has a tool. Microsoft Azure has a tool. IT Central Station lists 14 competitors to A2I.

Yes, Amazon MIGHT have used information about DefinedCrowd to build a similar product. Here’s the more reasonable explanation: Amazon motivated their investment team, with some form of incentives, to make “good” investments. For Amazon, investments were likely considered “good” if they not only generated venture-like returns, but also improved the commercial value of Amazon Echo devices. (Note: “The Fund's goal is to grow the ecosystem around Alexa.”)

In the face of allegations about Amazon’s investment strategy, the company insists that it “doesn’t use confidential information that companies share with it to build competing products.” The entrepreneurs cited in the WSJ article refuted that defence, claiming that Amazon spokespeople are lying and that Amazon DOES steal ideas from companies.

As I previously argued, companies are not single-minded organisms. A widespread strategy to invest in startups, gather relevant information, and launch a rival product would need to benefit Amazon AND ALSO benefit the individual humans that run all of the company’s various teams.

More specifically, this tactic would need to benefit Paul Bernard — the “Founding Director of Worldwide Corporate Development and Alexa Fund Lead.” His entire LinkedIn bio consists of roles where he helped companies grow “inorganically” (a fancy word for “helping companies buy other companies”). When Bernard searches for his next job, he will highlight the number of companies he bought for Amazon, along with information about their performance post-acquisition. He will NOT tell potential employers that he guided his team to interview companies, steal their ideas, and then pass that information over the “firewall” between the Alexa Fund and Amazon (contrary to company policy). In fact, this scheme would HURT Bernard — if executives in other Amazon departments had a shortcut to building new products, it would damage the return on his investments.

Even if Bernard were a Machiavellian villain straight out of a House of Cards spinoff, he would not have the incentives to break the rules in the ways described by the WSJ article.

Incentivizing unethical conduct

In a corporate context, individuals usually behave in ways that advance their own career. Sometimes — far less commonly — an individual person’s unethical actions also benefit the company. And when unethical individual incentives line up with unethical company incentives, you end up with situations like the Enron scandal.

In a post from last fall, I quoted an essay in The Correspondent that analyzed the effectiveness of eBay advertising. Authors Jesse Frederik and Maurits Martijn wrote the following:

It might sound crazy, but companies are not equipped to assess whether their ad spending actually makes money. It is in the best interest of a firm like eBay to know whether its campaigns are profitable, but not so for eBay’s marketing department. Its own interest is in securing the largest possible budget... Within the marketing department, TV, print and digital compete with each other to show who’s more important, a dynamic that hardly promotes honest reporting.
“Bad methodology makes everyone happy,” said David Reiley, who used to head Yahoo’s economics team and is now working for streaming service Pandora. “It will make the publisher happy. It will make the person who bought the media happy. It will make the boss of the person who bought the media happy. It will make the ad agency happy. Everybody can brag that they had a very successful campaign.” [Emphasis mine]

In multiple Marketing BS newsletters, I have articulated the idea that “individual incentives matter.” Marketers are generally happy to spend money (because their next job depends on budget size). As such, marketers search for any data that can justify expanding their fiefdom.

This sense of self-aggrandizement exists outside the marketing world, too. Other professional groups — like police officers, firefighters, and doctors — regularly advocate for greater resources, authority, and autonomy. (See “When the players are the referees”)

So, when we read accusations that an executive running one business unit would altruistically jeopardize his personal reputation and career aspirations AND violate company policy (and maybe the law) in order to help an executive in a completely unrelated part of the business, I think we can be skeptical.

What’s the more likely explanation for the fact that Amazon launched a product that directly competed with one from an Alexa Fund-supported startup?

The executive who went on to launch A2I probably wanted to work on some sort of human-review tool for years; there is nothing suspicious there, because this is an emerging field attracting lots of interest. Perhaps frustrated that Paul Bernard had not handed down a solution on a silver platter, the executive went ahead and guided their team to build its own product. Creating A2I not only satisfied the executive’s demands, but also contributed to Amazon’s ORGANIC growth — for which Bernard could not claim any credit.

In the hypothetical situation where an Amazon manager were intentionally breaking a policy, it’s far more likely that the actions would somehow benefit that individual manager’s career (i.e., a direct reward for their own department, not a different division). Consider the difficulty in convincing managers to FOLLOW a company’s directives to help unaffiliated business units, let alone hoping they do it when you explicitly caution them against such actions.

During the congressional hearing, Rep. Pramila Jayapal (D-WA) quoted another WSJ investigation that claimed Amazon used data from its third-party merchants to launch competing products. She pointedly asked Jeff Bezos, “does Amazon ever access and use seller data when making business decisions?”

The Amazon CEO’s response:

I can’t guarantee you that that policy has never been violated. We continue to look into that very carefully. I’m not yet satisfied that we’ve gotten to the bottom of it, and we’re going to keep looking at it. It’s not as easy to do as you would think because some of the sources in the article are anonymous.

Despite a corporate policy against using third-party data, Bezos acknowledged the possibility that individual managers responsible for launching new products could have looked at the data anyway. That information — which would benefit their specific business and advance their own career — IS the type of incentive that motivates individuals to break company rules.

Final thought

The key thing to remember from today’s newsletter: companies don't do anything. It’s the people who work at companies who do everything. To understand company behavior, you need to consider the behavior of individual managers. Companies may compete in the marketplace with other companies, but individuals compete both externally AND internally with other individuals.

When trying to analyze why something is happening, investigating the individual incentives often matters more than focusing on the company incentives — especially as organizations grow in size.

Keep it simple and stay safe,

Edward

Marketing BS with Edward Nevraumont

Discussion about this post