The Four Myths of Bundling with Shishir Mehrotra

In each HeadSpin Corner episode, our VP of Developer Products Jeena James talks to experts and thought leaders at HeadSpin and in the industry about what’s new in both the testing community and digital experience space, the latest industry trends, how to achieve digital business success, and more. You can find more episodes here.

 

We sat down with Shishir Mehrotra to hear his thoughts on the bundle – which ones work and which ones don’t, what we can learn from bundles that exist today, and what he believes bundles will look like in the future.

Shishir co-founded and currently leads Coda, an all-in-one doc that bundles together documents, spreadsheets, presentations, and applications to allow users to reimagine what they can do with productivity tools. Before Coda, he was responsible for Youtube products at Google, where he first began studying bundling and developing his set of principles that challenge common misconceptions on the topic.

Shishir combines mathematics and narrative to discuss these misconceptions, or myths, in this writeup, which he presents in detail in our conversation below. Read on or watch our interview below to hear his thoughts on bundling that will have you rethinking subscription products and your own approach to product development.

Tell us a bit more about yourself, what you have been up to, and an app you absolutely love now.

I’m Shishir Mehrotra. I am the founder of a company called Coda. We make a new type of document. It’s an all-in-one document. Bundling documents, spreadsheets, presentations, applications all into one new surface. The promise is anyone can make a document as powerful as an app.

Before this, I spent about six years at Google. I was responsible for the YouTube products there, which I assume we’ll talk more about in a bit. Before that, I spent about six years at Microsoft. I worked on Office and Windows and the SQL server, and then before that I started a company called Centrata, went to school in Boston (MIT), and I grew up in Virginia. 

The first app that came to mind was actually Tik-Tok. I don’t think I’m in the right age demographic for it, but I like following along with my kids, and I think it’s rare to see — especially coming out of my YouTube background — it’s rare to see a product change a media format that quickly and the level of engagement they get is sort of astonishing.

We’ll show up to a party or park, and even if the kids don’t know anybody, within 10 minutes they’re recording Tik-Toks. I like that it’s participatory and not leaned back, and I really like it. There are some Tik-Toks of me, but I wouldn’t brag about any of them.  

You mentioned Google, Youtube, and Microsoft. How has the journey been, from starting a company, moving on to Microsoft, then Google, then starting a company again? Tell us a bit more about one thing you unlearned and one new thing you learned.

I’ve been super lucky in my career. Every job transition has been a lot of fun. I’ve generally been drawn to something. I think one of the main things I’ve learned is each of my job transitions has been chasing an idea more than chasing a company or so on.

I started Centrata because I fell in love with the idea of what we wanted to do with utility computing. I joined Microsoft because there was a particular thing I wanted to work on in Office. I joined Google because I wanted to change how video worked. Each thing was less about the structure and more about the idea, and I started Coda obviously with this view that you can make docs as powerful as apps. I found that chasing an idea that you really believe in — usually the construct around it matters a little bit less. So, it’s maybe something I’ve learned.

What have I unlearned? As far as both sides, there’s things in big companies that you take for granted that as you go to a smaller place you realize you don’t have a team of 50 people to think about issues that might not be top-of-mind for you. It can be quite helpful, but it also creates clarity to not have that. 

I think probably the biggest thing was just clearing some of that away. You’re starting from scratch and saying to reinvent processes that otherwise you just sort of take for granted that you’re going to spend X months a year on performance reviews and you’re going to deal with these types of cross-divisional issues and so on, and you just start again and then you don’t need to do that.

We’ve been hearing a lot about bundling and unbundling, but also bundling and package value props that you can provide, and you’ve done a lot of research on this.

You also put together some thoughts and processes that businesses and individuals have. What gave you the idea of bundling? What was the opportunity that you saw that you felt you had to go deep into this topic? Is there a body of work that is actionable?

I like to say bundling is probably one of the weirdest hobbies I’ve ever seen somebody have, but that’s kind of what it has become for me. Just a little bit of history on it — when I joined YouTube in 2008, this was soon after the acquisition and job number-one was turning YouTube into a good business. It was a pretty good consumer property, but it was not a particularly good business at the time and, and every presumption was that YouTube would be monetized with advertising. So, that’s what we spent our energy on. Of course, being at Google, that was an obvious thing to do. 

So, we spent our time and energy on that, but in the back of our heads, there was always this thought that someday we’re going to turn this into a paid offering. It was never priority number one, but priority number four, or priority number eight. It was on the list somewhere, and we tried experiment after experiment after experiment and none of them worked. I can probably think of 10 different versions we tried, all of which were not like little failures, but generally sort of abject failures. 

There’s one in particular where we worked on it for six months. We shipped it. It made 100 dollars, not 100 dollars per person, but literally 100 dollars. We bought a few pieces for the team and that was it and shut it down.

In that process, I started thinking about what are we missing? Why is it so hard for us to get this done? 

One of the things I realized is that in every discussion about paid offerings, we had this built-in presumption, which was basically that we would do it differently. At Google, everything is do it different, do it better, do it Google-y, and there was this presumption that we would do it better than everybody else had done it. As I unpacked what we meant by that, I realized that there was one word that was most evocative of that emotion. It started with a ‘b’. It’s called “bundling”. Anytime we talked about what we would do with paid offerings, we’d say “Well, we’re going to do a new thing and the one thing we can’t do is we can’t do the bad thing that the world has already done”.

In the media space, that is quite common, but I’ve since come to realize it’s not actually really about the media space. I’ve given talks on this topic a lot of times. I’ll ask a room, and when I say the word “bundling”, what do you think of? Cable TV, right? Everyone’s first answer is cable TV, generally it’s Comcast, and generally it’s a very negative emotion. This is not seen as a positive thing. 

So in this period, nothing we did worked. I say maybe we’re missing something. I read everything I could find, finding everyone I could talk to, and just saying maybe I can learn a little bit more about how this works. I end up finding like-minded people, reading some good things, and I end up discovering that there was a set of things I believed about bundling that turned out not to be like a little bit wrong, but almost a diametric opposite of where I started. I refer to those as the “four myths of bundling”, and that’s what eventually became this paper.

This idea was pretty top-of-mind for me. We took some action on it at Youtube, but I left YouTube in 2014. A lot of this is developed afterwards. When I left, I started Coda, which is by itself a bundled doc, tables, apps, all-in-one surface, but it also became my primary investment thesis.

I’m a pretty active investor.  I invest in a lot of different companies, and I probably invested in something like 20 or 30 different subscription companies of various different types with some version of a bundling thesis. The most notable of which is probably Spotify.

I ended up helping Spotify early on, and then I joined the board a few years back. In that process of working with each of these different companies, I’ve gradually refined this theory and come up with a different worldview for it.

There’s a little bit of history where it came from. Once you see the world this way, it’s really hard to unsee it. It’s helpful for the most immediate thing you could imagine, say I want to build a multi-product bundle, but actually the same philosophies apply to many other things. It applies to general product development – how you think about putting together features in a product is also a form of a bundle and so on. 

What this thought process represents is a slightly different view of the business models for the internet. I like to call it the third business model for the internet that we went through, what I call the SuperFan period, the NonFan period, and the CasualFan period. We’re about to enter this new period which I think is very exciting. The last one, the NonFan period, which is really driven by ads platforms, including ones that we built at Google, led to all sorts of new businesses that wouldn’t exist without that business model or without the ability to monetize people’s attention and time. That led to publishers and social networks and games and all sorts of things that came out of that ability, for example blogs. I think that if you sort of buy into this theory, it will reshape your thinking on multi-product bundles. It may affect how you think about product development, but it may also affect how you think about what are viable services on the internet. So that’s a little bit of background on the theory.

It’s called the “Four Myths of Bundling”. You can find it published online. A lot of people helped me get this into good shape, and embedded inside this is a few different things, including a slide deck.

Terminology

  • Consumer: purchaser of the bundle (though in some cases, a consumer could bea  business, we’ll simplify terminology to consumer for now)
  • Provider: provider of a good within a bundle
  • Bundler: the organization that bundles goods together and sells them to consumers
  • Bundle: something a consumer purchases that includes access to multiple goods generally (though not always) from multiple providers. For shorthand purposes, we will assume all bundles in this doc are billed on a monthly subscription basis.
  • Myth-maker: our personified adversary for this discussion – an individual who will vehemently argue that all of these 4 myths are true

The whole paper is structured as a dialogue between me and this myth-maker. When you picture the myth-maker, you can picture anyone that you know that cringes when you say the word “bundling”. When I wrote this, the person I pictured was myself just a few years earlier, because everything in this is things I said and said. This paper is really a story about how my viewpoints on these things sort-of turned backwards.

Myth 1: Bundling is bad for consumers (as well as providers). 

This is the most generic and most common viewpoint: bundling is bad for consumers as well as providers. Let’s run a hypothetical exercise. Let’s imagine I have four products, and I have to make a decision. Do I sell them a-la-carte, or do I sell them in a bundle?

First let’s define a few terms precisely. 

  • Superfans are people who have two characteristics: They would pay retail for the good – they would pay the a-la-carte price for the good – and they have the activation energy to go find it. 
  • CasualFans lack one of those two characteristics. They wouldn’t pay retail or they don’t have that activation energy. 
  • NonFans ascribe zero or negative value to the product or the bundle.

This picture is meant to visualize these terms. I have a SuperFan for each product (shown in the blue areas), and there’s a set of CasualFans. What you’ll see from the picture is that what bundling does is it allows you to capture CasualFan value.

If I were to sell these products a-la-carte, the only products that I would get as a consumer are the ones that I’m a SuperFan of. From the perspective of the provider, the only customers would be SuperFans. What bundling does from a provider perspective is it allows you to address CasualFans. From the consumer’s perspective, it allows me to have access to goods I may only be a CasualFan of. 

One example is McDonald’s and the value meal. When one person stands in line at McDonald’s, and they think “I kind of want a Big Mac, maybe some fries, and if I buy those two things, then the drink is free.” The person right behind them could say “I kind of want fries. I kind of want a Big Mac or maybe I want the drinks and now the Big Mac’s free.” The person behind them is probably the most entertaining with the screaming little kid who says “Hey, I really need something to get my kid to stop screaming. There’s a Happy Meal that has a toy, and if I buy the toy, then I give something to my kid and the whole meal is free.”

Another example is UFC. When I ask a room of people, “Who here is a fan of UFC?”, generally between two to five people raise their hands. Everybody else is quiet. Those two to five people who raise their hand, raise their hands not a little bit, but all the way up in the air. If you’re a fan of the UFC it is hard not to be a SuperFan. When I ask the same room “Who here is a fan of the NFL?”, I get hands up at all levels. Some say they only watch the Super Bowl, some say they only watch when their team is winning, and some say they watch every single game. There are a wide range of NFL fans, but the UFC only has SuperFans. 

I believe the very simple reason is that the UFC is mostly an a-la-carte offering. If you’re a UFC fan, you pay 50 dollars to watch every Friday, and it becomes part of the viewer’s identity. It’s harder to get access to UFC. When I ask NFL fans how they pay to watch the NFL, many have no idea. Many say, “I don’t know. It just shows up. Sometimes it’s on these channels, sometimes Verizon gives me something. I just have it, and now I have to make a choice about whether to consume it or not, but I’ve already paid for it.”

If you then ask them which business they would rather have – the UFC or the NFL – most people would tell you the NFL business because it is 100 times bigger. The heart of that  is because it addresses CasualFans through bundling. The NFL has a much broader base of customers than the UFC does. When done well, bundling produces values not by addressing SuperFans but CasualFans. 

Thesis 1: When done well, bundling produces value for both consumers and providers by providing access for (and revenue from) CasualFans.

Myth 2: Revenue from bundles should be allocated based on usage.

A general response I get after presenting Myth 1 is “Yes, your example there makes sense, but they are not paying providers fairly. You need to get out of the bundle in order to be paid fairly.” When people say fairly, what do they mean? In Silicon Valley, “fair” is associated with usage, and usage is associated with consumption. 

This chart here is a fun way to look at this. I use a lot of media examples, partially because they are very relatable and partially because my theory was developed with that in my mindset.  

The History Channel and ESPN get about the same amount of usage.

They are very close in terms of how much time people spend watching each one. If you look at the way the cable bill breaks down, The History Channel gets about 25 cents per subscriber per month, and ESPN gets about five dollars per subscriber per month.

The y-axis is marginal churn contribution. Marginal turn contribution means if I remove one product from the bundle, how many people would churn? A number of surveys have shown that if you were to remove ESPN from the bundle versus History Channel from the bundle, about 20 times as many people would churn. That’s why ESPN gets paid 20 times as much, even though it gets about the same amount of usage.

This leads to a set of interesting equations as well. The paper walks through the math on this point.

Corollary 2.1: 

WholesalePrice (ProductX in BundleY)
= RetailPrice (BundleY) * MarginalChurnContribution (ProductX in Population of BundleY)

Corollary 2.2: The best bundling outcome will have

WholesalePrice (ProductX in  BundleY)
= RetailPrice (ProductX) * SuperFan% (ProductX in Population of  BundleY)

I would think of the second equation as a bit like supply and demand curves. The best bundling outcome will occur when this equilibrium is achieved.

We’ll use Spotify, or music services, as our product. Let’s say I run an apartment building and I decided I’d like to give Spotify to everybody in the building. I somehow get Daniel on the phone, and I say give me a price. This theory explains what that price should be. The retail price of those products are $10 a month. Let’s say for a minute that the Superfan%, the percentage of people who already subscribe to a service like that, is about 10%. The wholesale price of that product in my bundle, as a part of being in my apartment building, the price that I should pay Spotify to give you access to Spotify is about $1 a month. 

What this second theory says is that you can now calculate this contribution based on behavior outside the bundle, while the first theory answers: how do you assess marginal contribution and do you have to remove something from the bundle in order to do it? Usage is a very poor way  to pay providers. The better way is through marginal churn contribution, and there is actually a way to estimate that price. 

Thesis 2: The most fair way to distribute revenue to providers in a bundle is by Marginal Churn Contribution, not Usage.

Myth 3: Bundles will always feel like a rip-off to consumers since they represent a lack of choice. 

Corollary 3.1: A consumer C will see a bundle as being “a good deal” if: Price(BundleY) < Sum(RetailPrice(X)) for all Products X in BundleY for which C is a SuperFan

Corollary 3.2: A consumer C will see a bundle as being “a good deal” if: Price(BundleY) < 
IntegrationValue(BundleY)
+ Sum(CasualFanDiscountValue(X)*RetailPrice(X)) for all Products X in BundleY that the consumer sees as being part of the bundle

At this point, the myth maker says “That makes sense, but everybody is still going to think the bundle is a ripoff, because bundles, at the end of the day, represent a lack of choice. In order to buy X, I have to buy Y. So, clearly even if you get providers happy, consumers are still going to be unhappy.”

Now let’s look back at the example I started with McDonald’s and the Value Meal. Why does nobody think it’s a ripoff? Because there’s a price on their menu for all the individual items – there’s a price for the Big Mac, there’s a price for a drink. No one is forcing customers to buy them together, but the Value Meal is just a better deal. For a consumer to properly value a bundle, there must be a transparent and reasonable a-la-carte price for each of the products in the bundle. 

I get asked a lot: what is going to happen to Cable TV? Is Cable TV going through an unbundling? I do think the Cable bundle is one of the more broken bundles out there, and I think it probably needs to be reset, but I think what’s happening right now is an unbundling to re-bundle. The reason it’s happening is because people don’t have transparency on a-la-carte pricing.

Say you ask people what their biggest complaint about cable is, and they say they don’t want to pay for everything. Then you ask them, “What would you like to pay for?”, and they tell you ESPN. You ask them how much they’d like to pay, and let’s say they say 10 dollars (based on what they’ve read online that ESPN receives five dollars per cable bill). If you go back to the second equation in Myth 2, you’ll notice that five dollars is just ESPN’s wholesale price. Their retail price is a function of the SuperFan% – what portion of the bundle would churn if I removed ESPN? That number is about 10%, therefore the correct retail price is 50 dollars. 

This time you tell them that the price for ESPN is 50 dollars, but you will give them 299 other channels for free. All of a sudden your optics on this are completely reversed. The problem is people don’t think that ESPN is 50 bucks, so they think they’re getting ripped off. I actually think the cable bundle as it stands is one of the best deals on the planet (I do think it’s broken, but we’ll come back to that later). If you look at a bundle, your ability to access the value of the products in the bundle starts with whether you can access the a-la-carte price of the things you’re a superfan of. 

Thesis 3: For a Consumer to properly value a bundle, there must be a transparent (and reasonable) a-la-carte price for each of the products in the bundle.

Myth 4: The best bundles are narrow and have very similar products so they make sense to consumers.  

This myth is the most fun but most controversial, mostly because it’s counterintuitive. Why do bundles always bundle such unrelated things together? Why can’t I just get a bundle of related things, why do I have to buy things that don’t have anything to do with the main thing I’m looking at? 

When we were building up one of our monetization experiments in 2012 at Youtube, one of the deals that came up was NFL Sunday tickets. DirecTV had this offering for viewers that allowed consumers to get access to every NFL game except for the ones in your current market. DirecTV at the time was paying about a billion dollars a year for the rights to these NFL Sunday tickets. We got together and said we thought we could afford to pay a lot more than that. We decided to try to get the deal done. 

At the time, Larry had just taken over as CEO. Any deal this size (two billion dollars a year is a lot of money, even for Google) goes through all of Larry’s management team. So we went to the rest of the Google leadership team and we said “Here’s what we want to get done.” The first meeting was kind of hilarious. One thing to know about this leadership team is no one there really watched football. I’m going through it, and at some point someone says “Two billion dollars a year? Can’t you just start a sport for that much money?” So, my whole first meeting was distracted by what sport could you start with Ahmed that money. So we went back and started from the beginning, explaining what the NFL was, how many teams there were, what out-of-market meant, where all the rights were, here are the rights you want to buy, here are the rights you don’t want to buy, and so on. 

The first obvious question Google had was “How are you going to get all the other rights? How are you going to buy the rest of the rights of the NFL?” The first answer I gave is that you can’t, because some of these deals are locked up for 15 years. This was very controversial and unheard of at Google. At Google, you don’t do things kind of part way. You don’t map part of the world, you map the whole world. You don’t buy part of the NFL, you buy all of it. That was strike one, but I said we could bundle it with other things – not other sports, but other channels, from food to knitting, That got us laughed out of the room, and the deal didn’t happen. 

So what is this theory of Myth 4? I wish I had more to this language back then, because I think we would have probably gotten the deal. The reason for that is the part that people miss.

Thesis 4: The best bundle is one that minimizes SuperFan overlap and maximizes CasualFan overlap.

If you look at this diagram, it says imagine I have these three products: Product A, Product B, and Product C, and each one has a SuperFan base and a CasualFan base.

You look at this picture and you say, “I want to minimize the overlap between the solid circles.

Every time I have overlap between the solid circles, I now have a product I’m now offering in a bundle something that you would have paid for both products.” That’s an economic waste, but every time I get this type of perfect Venn diagrams, I get a really good bundle.

That is counterintuitive, but as you think about and look at the big bundles that have been created, you’ll see this pattern. You’ll see the reason things get put together is by adding the parts of the bundle, it’s not meant to super serve your SuperFan base, but rather to expand your customer base. That’s what bundling really does. This is another way to think about bundling as another way to expand your CasualFan base.

Summary of the Four Myths 

  • Myth 1: The way bundling produces value is by producing value for CasualFans.
  • Myth 2: Don’t allocate money based on usage, do it based on maringal churn contribution.
  • Myth 3: If you want your bundle to be perceived as a valuable purchase, then expose transparent and reasonable a-la-carte price for each of the products.
  • Myth 4: Minimize SuperFan overlap, maximize CasualFan overlap. 

Did you look at why [Myth 4] is? Is it possible that the consumer is not a single consumer but a unit of consumers, so you are actually maximizing CasualFans that can be CasualFans in a family or in a group of people consuming the same content?

Group purchases are very interesting, and that definitely applies to a lot of products. Everybody’s gone through that experience. It’s not really, “I’m a SuperFan of ESPN, and my spouse is a SuperFan of History Channel, or my kid is a SuperFan of Nickelodeon” and together, we have to make a collective decision. There is a lot of human behavior in that. For Spotify, the family plan has been very popular, and in music that works very well. 

The idea of treating the purchaser becomes a unit and if you go one step even further and think about it in a B2B context, it’s even more true. In a B2B context, a whole company has to make a decision to buy a product, and probably the most famous bundle and in that space is the Microsoft bundle. You buy an enterprise agreement to the Microsoft products, and you get Windows and Office but also things like SQL server and exchange and access to the dashboard development cycle. Completely different people in this company decide that this is the right way to buy these products. 

So, I think the idea of aggregation and of the customer purchase is a very powerful one and very important. At the end of the day, the same dynamic supply. Now I have a group where collectively they need to decide whether they’re going to purchase a bundle or not – and certainly within that group you need to get that SuperFan and CasualFan dynamic. Same thing applies. The thing to remember about it is if everybody is a SuperFan, then you actually are kind of wasting money, because if they all would have paid for their independent things then you might have been better off separating them off. You have to find this CasualFan overlap in order to really get the value of bundling. 

What are your thoughts on Netflix in that case, where you do have profiles and you don’t have to necessarily pay for a single profile, and you can have 5 profiles? Would you call that a bundle in some way, or would that be in its own category?

Netflix is one of the most successful bundles on the planet. If I think about what Netflix did is that they really surprised the world this way. Ignore for a moment what they’ve done in the last few years with original content. At Youtube, I was paired with a guy named Robert Kinsel, who actually built the online catalog for Netflix. Netflix used to do DVDs, and they built this online catalog that became their new business. That original offering didn’t seem that good. They had some things you kind of heard of and they had this big deal with Starz that’s often talked about, but then they licensed a bunch of French films, and then a bunch of Indie films and then a bunch of Spanish things, and then some documentaries. 

In no case did they get anywhere near comprehensive. They didn’t have all the French films and certainly didn’t have all of Bollywood or anything like that, but they just had little bits of each one. The reason this process happened was that was all they could get. The practicality of it was that these were undervalued assets and those were all they could get. What they ended up doing is they ended up building a coalition of different types of people. So, people were subscribing. At the time, they had a million subs. Some people were there cause it was the only place they could find some French movies, and for some it was the only place they could find these documentaries. They assembled this unlikely coalition in order to be able to do that, and that became the Netflix bundle. Over time I think they’ve done a very good job of that. 

As they started addressing kids (which is really why the profile thing really matters), they started realizing that they could take a person who might not have themselves been able to tip over and become a subscriber, but if they just add a little bit of kids programming, then the family as a unit will decide that their kid can sit on their iPad and watch this content (almost as a form of babysitting), so their value for it is very high. Then the family will notice another thing they kind of thought was interesting, so now they’re a CasualFan of that.

What happens over time is you start to see different behavior emerge. The family starts to  fall in love with the content – “Wow, I never knew about House of Cards and I never really knew about Stranger Things” – and now they start to fall in love, and all of a sudden the base shifts and now they become a SuperFan of these things. Then they drop Disney and they say they don’t need to pay for that for the kids anymore, because the parents are now SuperFans enough that it’s justifying this subscription.

Ted Sarandos sits on the Spotify board with me, and we talk a lot about the mechanics of the similarities between the Netflix bundle and the Spotify bundle. I think it’s very powerful that in the early days, it was a coalition of completely unexpected, unloved content. The Netflix content team literally went and bought things that nobody cared about. People were not even bidding on some of these things, so many of these things were just sitting on a shelf. You wonder why they would possibly take a bunch of French content and put it with a bunch of Indie content, until you realize it’s because of SuperFans and CasualFans. 

How does bundling apply to a SaaS platform? Where have you seen [bundling] work very well? Where do you feel like people could have thought a little more while trying to bundle or show a la carte?

This is a good lead into the predictions and implications. I think that bundling is a natural evolution for many Industries. I use mostly media examples or I either went through this, but as you think about these examples you’ll start to see everywhere. 

Predictions and Implications

Bundling is a natural evolution of many industries.

Credit cards are a very classy example. What does the Amex Platinum Card come with? If you ask anyone that’s a platinum subscriber, they’ll all tell you a different reason. It comes with access to a bunch of airline lounges, a warranty program that if you drop your iPhone in the toilet they will replace it for you, Early Access to concert tickets, and more. 

There’s a list of different things that have been stitched together into a credit card, which itself is a financial instrument, as another product. We see this in travel through frequent flyers systems and vacation clubs and so on, in enterprise software with the Microsoft suite and the Oracle Suite. 

Amazon Prime is another very commonly used example for thinking about bundles. When Amazon came out with free shipping, everybody thought that was great,  and then they added on music then movies, and everybody thought that was kind of weird.  What were they really doing? They were expanding their base and they’re now one of the most accessible on the planet.

I think one of the ones that probably is most counterintuitive is the insurance bundle. If you think about insurance that way then you think about Healthcare. Health insurance is a bundle at multiple levels. First off, it’s a bundle between healthy and sick people. You get a group of people together and the super fans of your health insurance plan are the people who are sick. Secondly, it’s often a bundle of medical, dental, life, disability. Probably most importantly health insurance gets bundled, at least in the U.S., with your employment. In most first-world countries in the world, it gets bundled with your citizenship.

If you think about the bundle in that way, you might shape it differently. You might say the debate about healthcare vs. government healthcare is not really so much about whether the government should pay or not, but rather should I have healthcare bundled with employment or bundled with citizenship. If you think about it that way, you can have a reasonable debate. Maybe it shouldn’t be with either and maybe it should be bundled with something else, but it’s just a bundle and you can choose where you want it to go and you can work through it dispassionately. I think this is a very natural evolution for a lot of industries – for journalism, for students, for transportation. 

This applies to more general product development as well. 

I once gave this talk at a company that didn’t yet have a multi-product bundle, and one of the people asked if this could apply to ignoring for a moment multiple products and to their own product development. As we talked about it their theory developed, and they since shifted to thinking about it this way. 

You start with a product that has multiple features and you have a pricing page. Some things are behind the paywall and some things are not. You have a premium product, but many products work that way. How do you decide which one goes behind a paywall, and which one doesn’t? How do you decide which one goes in which tier? All of that at some level is a function of assessing SuperFans and CasualFans of each of those parts of the product, and ideally you want to put things back there that strike the maximal number of people into that set. 

Even if you ignore the pay wall and say well, I’m a Google customer. What am I really a customer of? Well, I use Google for Google search, but I also use Gmail, and I also use YouTube, and I also use Maps. I was really there for search, but the mail was pretty good, and I was a CasualFan of that. Map was really good too. Overtime, Google has established a coalition of features that forms its own bundle. Each of us with a Gmail account is really a participant in this bundle.

Bundles of the future will be larger and more diverse than we have today.

There’s an enormous economy to scale in bundling, that going from zero to one million subscribers is much harder than going from 100 to 101 million subscribers. The biggest reason is that you get to distribute money over a larger base of CasualFans.

I think the bundles of the future will be much larger than we’re used to. Today we think some of these bundles are already really big, but I think the future will not only pull together your media and your news and your books, but will also pull together your dry cleaning and your Uber rides and your healthcare, your gardening service. All of these things will start to get bundled together. At some point, you’re going to join this membership service, you’ll open your phone, you’ll say I’m a member of this service, and a thousand different apps on your phone are going to light up to give you the premium versions.

One last point is a somewhat different way to think about business models on the internet. We have these two extremes on the Internet working quite well. The Superfan business model is you pay for what you use. Uber is a great example of this. There’s no way for them to make money if you don’t go take rides, and that’s a really good what I call SuperFan business model – you either want the product for I don’t help pay for it when I need it.

Nonfan business models are generally ones you pay with your time. You’re not really paying for access, you’re just happening to use it and then you’re taking my time, my eyeballs, and you’re selling it to someone else (generally an advertiser). That’s how Nonfan business models work and they support a lot of businesses. Think Youtube, Facebook, millions of games and social networks and blogs. This was all enabled by a set of ad networks, and that allowed for these below-scale things to go and monetize eyeballs, because I can be included in this broader network. It’s almost its own form of bundling. When this happened, new businesses, many not viable before, got created.

Bundling will allow for a “third business model” for the internet.

My view is that similarly, what we’re going to see with bundling, is it allows a new type of CasualFan business to get created, one in which I can’t possibly use it enough to make a NonFan business model work, but where I ascribe enough value to it that if it were provided to me, I will stick in a bundle because of it. 

One example I used earlier was History Channel. I think History Channel is a product that is mostly not viable outside of a bundle. It gets 25 cents per month. You can’t buy subscribers with that, and you would have to raise the price so much in order to find people that have the activation energy, and the product would not be successful.

If I were to draw a diagonal line through this chart, what I find is a bunch of products way down here where my usage trumps my MCC, my anchor value.

Those are products that I use all the time. A lot of Youtube is like that. I use it all the time, but if it went away, how would I feel? I’d feel fine, and I would use something else. 

Going to a sporting event is at the top of the chart. YouTube monetizes roughly 15 cents per hour of consumption time. If I go to a sporting event, I might pay $100 or even $1,000 per hour for that same thing. There are all sorts of products that sit up here at the top of the chart. I believe there’s a line of products that all sit in this same area of the chart that have not even been created yet. As we begin to embrace and think about bundling in this way, I think we’re going to see a whole new class of products be created.

What are your thoughts on how Google and Apple have been thinking about bundles (Apple Arcade, Google PlayPass, Google One)? 

I think there’s a number of great bundles being created, and I think they are taking a slightly different approach to it. There’s also many people starting with their own products. I think it is often quite tricky to get the mechanics of a bundle right. One of my observations is that even after the theory is clear, the application of it is not always that clear, so I think we’re watching everybody kind of crawl walk run through this. 

It feels very similar to the early 2000s when AdSense came out, and people are trying to understand ad networks.  People didn’t understand what it meant to be a part of the network and how they were going to get paid. Nowadays, it’s easy. You go take an ad tag, you stick it on your page, and you’re done. There was a time when that was a negotiation and it was much more complicated. I think we’re still in the early innings of it , and I definitely think there are a lot of people looking at it.

In terms of digital consumption, have there been massive or monumental shifts in [bundling] consumption with shelter-in-place and work-from-home going on?

I think this pandemic and global crisis have accelerated lots of trends, and I think it’s definitely accelerated the most obvious one, which is we’re all learning to work distributedly, and there’s lots of implications of that. We’ve seen it in a Coda usage in which things people are picking up and so on. 

I actually think subscriptions and how these services work is similar. Some of that’s driven by change, since we’re all operating differently right now and we’re all home a lot more and need entertainment in different ways and we’re seeing spikes in a lot of these different services. Some of its caused by economic hardship. One of the ways of bundling creates value is that it does get people goods at lower prices, which is one of the reasons why I believe it’s such a fundamental force of the internet. People would rather be paying for one thing than four separate things. Some of it is just change. Anytime the world changes, you see people try new things.

I definitely think we’re seeing a lot of resetting of expectations across services in this period, and I think we’ll see more, but it’s sort of a broad acceleration. Somebody said the pandemic has caused a 10-year acceleration of the work-from-home movement. I’m not sure I feel quite the same level on subscriptions and bundling, but it’s definitely provided some extra lift to that trend as well.

What does Coda stand for and why is it called Coda?

Coda, if you spell it backwards, it’s “a doc.” If you look at what Coda is, the way it starts – a blinking cursor and a blank screen – it’s just a doc. I like to think of docs as our trojan horse. You can make docs as powerful as apps, you can build everything from taking meet notes and doing your one-on-ones. all the way up to building CRM systems and inventory trackers, but the heart of the product is that it starts as a document. 

We were brainstorming names when we came up with it. That name was actually on our list for a while, and it’s kind of an interesting word. It’s a musical term as well. Then somebody said “Hey, if you spell it backwards, it’s a doc,” then once we got that visual in our head we couldn’t unsee it.

This article is a summary and transcription of a recent HeadSpin Corner episode and presentation by Shishir Mehrotra. You can watch more HeadSpin Corner episodes here, or listen wherever you hear your podcasts.