Who Has The Best Business Model (And It’s Not Google Or Facebook)

Eric Feng
13 min readJul 3, 2018

When a management with a reputation for brilliance tackles a business with a reputation for bad economics, it is the reputation of the business that remains intact.”

— Warren Buffett

Back in 2002, I was working at Microsoft during a time when the tech giant was facing competitive pressure from what would become arguably the most transformative technology platform in the history of the world: the Internet. Entire industries were shifting their business online where distribution was open and free. And end users were spending all their computing time inside a browser accessing applications and services that Microsoft had no control over. It would take a few more years for us at Microsoft to fully realize that the true danger was not the technology of the Internet, but the underlying business model enabling it.

Why Pay When Someone Else Will Pay For You?

For most of the company’s history, Microsoft has earned its considerable profits (billions of dollars each year adding up to a staggering $130 billion in cash reserves at last count) from a direct sales business model. Customers of Microsoft’s products were also the people who purchased them and paid Microsoft directly. Sometimes the costs were bundled along with other purchases, but they were always there (a Dell laptop with Windows was more expensive than a Dell laptop without Windows). And you got exactly what you paid for: there was no try before you buy, freemium, buy one get one, rent to own, timeshare, or any other clever permutation. It was a straightforward, direct transaction between the customer and Microsoft.

But the Internet brought forth an entirely different way to charge for your products: making them advertising supported. Instead of asking the actual users of your products to pay for them, advertisers (who were not your customer and didn’t use your product) bore the cost of the service indirectly. Advertising had long supported many forms of media content (mainly TV, newspapers, magazines, and radio). But one company took advertising to frightening new heights by using the Internet to funnel the half trillion dollar global ad market in ways previously unimaginable. And that was Google.

Leveraging the scale and reach of the Internet, Google brilliantly turned the advertising engine onto new product categories that had historically always relied on direct sales: email (Gmail), telephony (Google Voice), operating systems (Android), and more. Take in-car navigation as an example. With Google Maps on your phone, an entire industry evaporated within a year as consumers completely stopped directly paying for maps and directions in their cars because Google now offered it for free. Pretty soon, it looked like everything could be ad supported from helpdesk software to coffee to airline snacks and more.

While foolish in hindsight, at first the wave of ad supported companies was something Microsoft could justify ignoring. The Internet was a leisure platform for reading news, buying books, playing fantasy sports, and other small, entertainment activities. Certainly not something on which meaningful business would ever be conducted on. And when Google went public in 2004, the entire US paid search market was $2.5 billion for that year. A drop in the bucket for Microsoft who at that time was doing $3 billion in revenue… every single month.

Over the next few years though, it soon became clear that ad supported businesses weren’t just a competitive threat to Microsoft. They were an existential threat because consumer behavior was changing from expecting to purchase directly, to expecting someone else to purchase on their behalf. And then Facebook came along with growth unlike anything ever seen before to create an advertising duopoly with Google as all Internet activity flowed through the front doors of those two companies. The dominance of ad supported businesses was cemented. Companies that had the audacity to charge users directly would succumb to their “business with a reputation for bad economics”, as you just can’t out execute someone with a better business model. Simply put, how does a company that charges users directly for a product, compete with a company that finds someone else to pay indirectly for the same product?

But that’s not how I believe the story will end, with all businesses moving to ad supported models, because ads suck.

The Problems With Ads

It’s a fact that most people dislike ads. Why? If I’m being generous, it’s because ads are a nuisance. With the exception of maybe the Super Bowl, you aren’t trying to view an ad. You’re trying to accomplish something else and the ads are a distraction, an interruption, a roadblock to that task (ironically, “roadblock” is an official ad industry term). You then tune out, and advertisers respond by turning up the volume, which make you tune out more, and advertisers respond by screaming at you louder and more often, which makes you tune out even more. Rinse and repeat until you hate ads.

Putting aside the problem of ads for customers, ads also present a critical limitation to the businesses relying on them to make money: they treat all customers the same, and therefore do not maximize the potential value that your best users could bring you. Whether you’re an avid, passionate user of a service or an infrequent, casual user, the ad neither knows nor does it care. Advertisers are simply looking for a pair of eyes and paying the business the same in all cases for those eyes, which leaves valuable money on the table for companies. That’s the downside to being indirect, and abstracted away from the customer. Your best user and your average user are effectively the same thing to an advertiser: both are just warm bodies.

Back when I was at Hulu, our average user was watching a few hours of content each month and encountering about 40 paid ads. Our best users watched 10 times that much content, and therefore theoretically earned us 10 times as much money. But the math doesn’t actually work that way.

The issue has again to do with ads being a nuisance: it’s been proven that the more ads you see, the less effective they are. Advertisers know this so they restrict how often their ads appear (known as frequency capping). Businesses also know this so they too limit all the ads you see overall (known as ad load) from every advertiser. Add (pun intended) it all up together and the conclusion is that companies that rely on advertising simply cannot earn significantly more revenue from any particular group of users regardless of their usage of your product or service. And that’s what we saw at Hulu. Our best users were great for us, but nowhere near 10 times greater than our average users.

With other business models, you have VIPs, power users, whales, high rollers, big spenders, etc. But not with ad supported businesses. Ads simply don’t allow companies to maximize the appropriate value from their best users.

Your Best Customers Are Your Best Customers

I’m not trying to suggest that advertising is a bad business. Google and Facebook, after all, made $135 billion in revenue last year off of ads. Ad supported businesses exploited a user behavior advantage over direct sales businesses (why pay for something when advertisers will pay for it on your behalf?). But advertising has its own “bad economics” that can be exploited. What if instead you had a business model that could maximize revenue from your best customers, and then share that value across all your customers, while not annoying users in the process? Sounds good right? Not only does such a model exist, but it’s being used by many companies to great success. I’ll call this strategy shared-value transactions.

To understand shared-value transactions, let’s use free mobile games as an example. Free mobile games dominate the most downloaded app rankings each week, as well as the top grossing app rankings each week. In other words, free games are both incredibly popular (makes sense — they’re free) and incredibly profitable (how?). Because of in-app purchases. Users can optionally pay inside the game to enhance their gameplay. Less than 2% of free mobile game players end up making in-app purchases. And of these users who pay, the top 10% of them drive an astonishing 50% of all revenue for games. So an entire industry is mostly built off of a tiny fraction of a percent of its users. How? Because their very best users are delivering 1,000 times more value to their business than their average user.

Similar to an ad-supported service, the vast majority of free mobile gaming users play the game without having to pay, and another party picks up the tab. But because of shared-value transactions, that other party isn’t an impartial advertiser that doesn’t have any connection with the game; it’s instead the very best, most engaged fan of the game that’s paying the cost. An advertiser just wants to buy a pair of eyeballs and will only pay what the going rate is for an arbitrary fleeting moment of attention. The engaged fan wants more utility, more connection, more status with the game itself and is willing to pay a rate that’s orders of magnitude above an ad. And best of all the engaged fan is happy to have done so as they willingly volunteered to spend money and are overwhelmingly satisfied with their purchase. In other words, this most engaged user is actually glad they paid 1,000 times more than the average user. Can you imagine how someone would feel if they watched 1,000 times more ads than the average user?

Sharing Is Caring

What makes shared-value transactions powerful is that it combines principles of direct sales and indirect advertising, tied together through commerce. Your best users, who have the highest understanding and appreciation of your service, buy the maximum amount of product from you directly because of the value you bring them. But this isn’t just pricing segmentation where a business lumps customers into separate groups and charges them different different prices. A key to shared-value transactions is that you then use a portion of the spend from your best customers to indirectly fund products for your average users, thus sharing value and improving the customer experience across your entire user base.

Again, a company’s best customers are many many times more valuable than their average customer. If you can find a business model that allows you to take full advantage of that disparity, you unlock economic value for everyone: yourself, your current best customers, and all the rest of your customers who one day could grow into your future best customers. Direct and indirect purchasing working together. Rinse and repeat.

Who else takes advantage of shared-value transactions? Let’s look at QVC and HSN. The television home shopping industry has been around for more than 40 years selling a wide assortment of products through their television content. Only a small percentage of viewers purchase directly, but those who do, buy heavily to the tune of $11 billion dollars each year. Those purchases indirectly fund the creation of shows that every other users gets to watch without paying. Direct and indirect purchasing working together. The model is so effective that out of all the stations on your cable dial, QVC and HSN are the only ones that don’t demand a small cut of your cable bill (known as carriage fees), which all other cable networks receive in exchange for their content. In fact, QVC and HSN actually pay cable providers to give them television content, and not the other way around, given how good the economics of their shared-value transactions are.

Selling Shoes

You know who else is really good at shared-value transactions? Amazon.

Morgan Stanley estimates that the average Amazon Prime user spends 4.5 times as much each year as the average non-Prime user. That’s a huge delta in revenue performance, even when looking at an average of one large 100 million user population compared against the average of a 60 million user population. But what if you compared the annual spending of the best one million Prime shoppers against the average non-Prime shopper? There are plenty of examples of big Amazon spenders purchasing $5,000, $10,000, $30,000 or more each year, or greater than 10 times the average Prime user. Extrapolating out, it’s safe to assume the top 1% of Amazon Prime customers spent orders of magnitude more than the average Amazon user. (What do you think Jeff Bezos spends each year at Amazon?)

Amazon can then take part of the enormous revenue generated from their best shoppers and use the money to indirectly fund services that average Amazon users benefit from too, like unlimited digital photo storage, same day delivery, free ebooks, and of course a premium streaming video catalog. Amazon can afford to give these benefits to even casual users who only spend $100 a year with them, because shared-value transactions allows them to capture $100,000 a year in spend from their power customers. The value from these purchases can then be shared with their average customers, to eventually create more power customers.

Staying on the topic of streaming video, this is a relevant example of how shared-value transactions gives Amazon a potential structural advantage over the leader in the space: Netflix. Success in streaming video requires great video content, and Netflix will spend $8 billion this year buying video rights. The way Netflix funds this hefty content bill is that they have 120 million customers who pay them $10 each month directly, and then they take half of that fee collected from every subscriber and spend it on content. So every subscriber pays for content equally (about $5 per user per month) as Netflix earns the exact same amount from their best users as their worst users.

Amazon too will spend a significant sum buying video content (about $5 billion this year). But their content bill is paid entirely differently. Instead of only depending on a percentage of Prime membership fees (which are the same for every user) to fund their content budget, Amazon can pay for content using revenue from purchases of books, diapers, toilet paper, laundry detergent, and more (and this spend is most definitely not the same for every user). As Bezos has said: “When we win a Golden Globe, it helps us sell more shoes”. Amazon’s best users are able to purchase significantly more goods than their average user, and these funds can be indirectly applied to fund video content that everyone shares value from.

Faceoogle Strikes Back

The current eight largest Internet leaders — Apple, Amazon, Microsoft, Google, Alibaba, Tencent, and Netflix — represent almost $5 trillion dollars in net worth, and all have dominating positions in their respective markets. They’ve also all achieved that success very differently. Apple, Microsoft, and Netflix are primarily direct sales. Facebook and Google have mastered ad supported everything. And Amazon (power shoppers), Alibaba (power marketplace buyers and sellers), and Tencent (power gamers) are leading the pack at shared-value transactions, with their deep ecommerce expertise.

But those business model lines are not set in stone. Amazon has a multi-billion dollar ad business. Facebook and Google sell Oculus and Pixel devices directly to consumers. Still, there’s a lot more that both Facebook and Google could do to take advantage of shared-value transaction principles. For example:

  • Facebook could take their recently launched Marketplace (the classifieds-like section that lets their users sell buy and sell goods from each other) and start offering an in-app payment flow to capture the entire transaction experience. Buyers and Sellers would both get the convenience of Facebook handling the money to provide safety and reliability in the transaction. And Facebook would get access to their users’ wallets, the first step in offering more products that their best users can directly pay for.
  • Google could take their lonely Shopping homepage and fill it with great content for product discovery. Think of of how full, fresh, and relevant their News homepage is. Why should the Shopping section be any less interesting than the News section, given product pages are far more monetizable than news articles? Google could put together something equally content rich for Shopping using their extensive crawled product catalog and query data for trends and rankings. From there, it’s not much of a leap to imagine buy buttons on those product pages that uses Google Pay and directly integrates into various commerce backends.
  • Want a wild acquisition idea that would make sense for both Facebook and Google? Buy Shopify. Facebook could use Shopify to supercharge Pages with the three Cs: product Catalogs, shopping Carts, and Checkout. Google could make Shopify a unique differentiator for their Cloud offering and take a leading position powering ecommerce infrastructure for the entire non-Amazon web. And both could add Shopify Buy Buttons across all their properties and services? Imagine a product listing ad you could directly purchase from.
  • And if you’re going to add a buy button to anything, the holy grail would be Instagram and YouTube. The core idea of shared-value transactions is to take your engaging consumer platform and allow the best users on those platforms to generate 1,000 times (and sometimes 10,000) more revenue than the average user. What better way to achieve that than to turn on product purchases from within Instagram and YouTube, which are your best properties that are already filled with incredibly popular product content that already drives purchase decisions?

Let Them Fight

Warren Buffett once said: “When a management with a reputation for brilliance tackles a business with a reputation for bad economics, it is the reputation of the business that remains intact.” The management teams running Facebook, Google, Amazon, Microsoft, etc. are all unquestionably brilliant. And none of them actually run a business with bad economics. In fact, their businesses have some of the best economics in the history of business.

Perhaps the only thing greater than what these businesses have already accomplished is what they aspire to accomplish going forward. There seems to be nothing off limits for these industry giants, which means more competition is inevitable particularly with each other. As the fight continues for control of our phones, our homes, our cars, our wallets, our food, our health, our time, a key weapon on the battlefield will be what business model they wield. And I believe that shared-value transactions powered by commerce will be the most lethal weapon of choice in determining whose business reputation will remain intact.

Let the battle begin.



Eric Feng

Current: Co-founder of @cymbalxyz, Co-founder of @GoldHouseCo Ventures. Past: @Meta (via Packagd), GP at @KleinerPerkins, and CTO of @Hulu and @Flipboard.