Warren Buffet’s Economic Moats or Elon Musk’s Innovation?

Economic Moats

Warren Buffet of Berkshire Hathaway has helped popularize the concept of “economic moats” over the last 25 years. Morningstar, an investment research firm, was amongst the first to formalize and systematically leverage economic moats as an investment strategy (in mid 2000s). Based on their research, Morningstar identified five sources of economic moats (in descending order of their importance): (1) Intangible assets (patents, brands, etc.), (2) Sustainable cost advantage, (3) Switching costs, (4) Network effects, and (5) Efficient scale.

These economic moats mechanisms have been investigated and studied deeply from capital allocation perspective — esp. for investment into mature and late-stage companies. It has been shown that moats-based capital allocation strategy (esp. when combined for stock valuation) provides higher return on invested capital (RoIC).

Are economic moats (“defensive strategies”) relevant for the digital companies? Or, does the speed of innovation (“offensive strategies”) suffice? Not surprisingly, there are differences of opinions amongst the practitioners. This was brought out sharply by the altercation between Elon Musk and Warren Buffet after Musk referred to economic moats as “lame” and “like nice in a sort of quaint, vestigial way” during a Tesla’s quarterly earnings call [May 2018; link]. Musk asserted that “the pace of innovation” is a stronger predictor of long-term success because “if your only defense against invading armies is a moat, you will not last long”.

Not everyone agrees with Musk’s assertions. For example, analysis done by some Venture Capital firms (VCs) has indicated that digital firms do benefit from the defensibility provided by various mechanisms. Interestingly, it turns out that the digital firms also have the similar four mechanisms for defensibility: network effects, scale, brand, and lock-in; however, the order of importance of these four differs vis-à-vis non-digital firms.

In the first three decades of the Internet era, network effects were found to provide the best defensibility. For example, consider Visa and Mastercard, which provide a payment network for customers, merchants, card “issuers” (credit / debit cards issued to customers) and payment “acquirers” (payment devices provided to merchants to accept card payments). This is a lucrative business: Visa generated $23B in revenue on $8.8T total payment volume (TPV) in 2019 — which is 0.25% of the payment volume. Mastercard is roughly half the size of Visa but with similar service fee structure [link].

There have been several attempts to disrupt this network — for example, AT&T, T-Mobile, and Verizon tried to build a payment network along with the large mobile phone manufacturers such as Samsung, Motorola, HTC, LG, etc. (called Softcard; originally, Isis Mobile Wallet; link) starting 2010 using the NFC technology (along with a secure hardware component). It failed to disrupt Visa and Mastercard networks and was acquired by Google Wallet in 2015. (As an aside: Alipay and WePay have been able to disrupt Visa and Mastercard networks in China (with combined TPV of $36T, which is 2.5x of Visa and Mastercard combined global TPV; link). More impressively, UPI in India has also been able to disrupt these networks — both in terms of number of transactions and TPV — within 3 years of its launch [link].)

This can be seen observed from Google’s ability to defend its search engine dominance despite spirited effort by Microsoft Bing; or, Facebook’s ability to defend and grow its social network despite Google’s numerous efforts to build independent social networks. This can also be seen from the continued relevance of services such as Craigslist and Monster.com despite limited innovation and upgradation in their services.

What about other three defensibility mechanisms? There are several examples for each of these: examples of defensibility provided by scale would be companies such as Expedia and Booking.com; brand examples would be companies such as eBay and Flickr; and lock-in examples would be Sybase (part of SAP) and COBOL (more than 25,000 companies still use COBOL).

Though investors (especially public markets and late-stage investors) might desire defensibility, entrepreneurs crave for growth. This is because growth is the magic potion that energizes startups and drives them to innovate faster during their journey. Therefore, till startups achieve maturity, one can visualize that growth is more important than building moats for defensibility for startups and entrepreneurs.

Given this, the question becomes: what are good drivers for growth? How can startups create value faster? How can entrepreneurs identify the most suitable growth strategies for their startups?

Value Stack

Based on our experience with hundreds of category-creating and category-dominating companies across consumer, business, health, finance, etc. categories, we have found that there are four value creation drivers:

• Scale (to refer to both supply-side and demand-side scale),
• Habit (includes stickiness and retention for categories such as health & finance),
• Brand (includes intangible assets such as patents, regulatory approvals, etc. — esp. for pharmaceuticals, finance, etc. categories), and
• Network effects.

You would notice that these look almost identical to the economic moats! So, are we just playing around with words? No — when we look at these from value creation perspective (instead of defensibility lens), we must look at these from product-first perspective. In other words, we must ask: how can startup build relevant products that directly help drive scale, create habit, build brand, and unlock network effects?

Here’s the key insight: the first three growth drivers can be unlocked by adding products with distinct characteristics to the product mix:

• Scale value boosters require products/features that have high frequency of usage and low importance.
• Habit value boosters require products/features that have medium to high frequency of usage and low to medium importance of activity.
• Brand value boosters require products/features that have high importance and low frequency of usage.

We refer to this product-led value boosters that help build Scale, Habit and Brand as “Value Layers”. Why Value Layers? Products with the right characteristics, once built, provide ongoing benefits to the company. Not only the value of these products increases as number of users/customers increase, but the products themselves can be improved as well. The additive nature of these product-led value layers helps to build the full Value Stack. We have observed that the Value Stack is unlocks disproportionate value for the companies and, for Value Stack, the sum is greater than the parts (corresponding to the individual Value Layers).

Value Boosters are product-led value enablers for scale, habit, and brand.

The additive nature of different layers is highly prevalent in nature, as can be seen below [link]:

The above spirals are known as Hemachandra-Fibonacci Spirals corresponding to eponymous number series. Hemachandra number series is additive as can be see from the first few numbers:

1, 1, 2, 3, 5, 8, 13, 21, 34, and so on.

After the initial two numbers, each number in the series is the sum of the previous two numbers. There are fascinating stories about the origins of these numbers in Ancient India and how Pingala, Varahamihira and Hemachandra — Indian mathematicians — used them to define rules of Sanskrit poetry, music, art, astronomy, etc. (and, thereby, providing mathematical foundations to art; one example of this can be seen from the use of “Golden Ratio” in design and arts even now) [link].

Coming back to additive Value Layers, the product-led engagement boosters are not targeted towards building moats; instead, these product-led boosters help companies to create value and benefit from the ongoing additive nature of the products.

Engagement Graph

How can companies build products in order to leverage Value Layers? Is there a framework that can be used by startups to explore and build these value creation engines in a systematic manner?

Towards this, we first define Engagement Graph that depends on two parameters that underlie the value creation drivers. We then use the Engagement Graph to outline how startups can build Value Layers in a systematic manner.

All activities done by people — whether in personal or work context — can be viewed from the perspective of the “Frequency of activity” and “Importance of activity”. Let’s start by defining the scale for “frequency of activity”. Tasks that correspond to daily (or a few times a week) use-cases are considered to have “high” frequency of activity; weekly (or a few times a month) use-cases have “medium” frequency of activity; all other use-cases have “low” frequency of activity. The scale for “importance of activity” can be defined likewise. Tasks that have large implication and, therefore, require consultation with other stakeholders (such as family members or corporate committees) can be classified to have “high” importance of activity; tasks that trigger users to diligently evaluate pros/cons amongst alternatives as “medium” importance; utility-like tasks that can be performed without much thought are “low” importance tasks.

Following diagram shows the Engagement Graph with the characteristics of the three Value Layers:

The figure above provides an indicator towards how startups can leverage Value Layers: products/features are more amenable to scale-based, habit-based, or brand-based defensibility mechanisms based on the characteristics of user activities they cater to. There are different zones corresponding to different activity characteristics; the figure above shows the Scale, Habit, and Brand zones.

Conversely, a company can scale faster, increase stickiness or strengthen brand by consciously building products/features that cater to user activities with the desired frequency and importance of activity characteristics. In other words, by adding products/features with the relevant characteristics to their product mix, companies can strengthen and sustain their growth using scale, habit, and brand value drivers.

Let’s consider a few examples to make this more concrete.

Let’s look at Amazon to explore the first two Value Layers. Amazon’s primary focus area (i.e., commerce) has a weekly or bi-weekly usage frequency. Typical purchases range from low to medium importance — from buying household goods to purchase of fashion (say, apparel) products. In other words, Amazon’s e-commerce product falls under the “Habit Zone”.

In strengthen its position in the Habit Zone, Amazon launched Amazon Prime in 2005 to address two primary concerns: delivery charges and the speed of delivery. Amazon Prime removed minimum basket size requirement and promised “two day shipping” — that is, any product order that is covered under Prime gets delivered within two business days. (Amazon Prime removed 2-day shipping promise in 2015 — approximately, 10 years after launching the program. However, most people still believe that Amazon Prime guarantees free 2-day shipping!) By reducing both mental and emotional effort associated with ordering online, Amazon was able to increase frequency of usage. The graph below highlights not only the growth of Amazon Prime but, more importantly, the fact that Amazon Prime customers spend more than twice as much as the non-Prime customers ($1400 versus $600 per year). Amazon Prime, therefore, is a great example of Habit Value Booster.

Amazon Prime Videos is a wonderful example of Scale Value Booster. Amazon Prime Video addresses a user activity that has daily usage frequency and has low importance. By making thousands of Prime Video available to users at zero cost, Prime Videos is able to acquire customers. Subsequently, customers can not only consume free videos but also see TV episodes and move on a pay-per-use basis. Moreover, customers can subscribe to more than 100 premium channels with Prime Video Channels.

Prime Video increases customer’s engagement with Amazon platform, which, inevitably, results in higher frequency of e-commerce purchases from Amazon stores. This is highlighted in the graph below: 6% of Prime customers make daily purchases; 18% of Prime customers purchase 2+ times per week; another 22% purchase once a week. In other words, 46% of Prime customers (as opposed to 13% of non-Prime customers) purchase at least once a week!

Another beautiful example of Habit Value Booster is the Zestimate tool / product launched by Zillow in 2006. Zestimate provides an estimate of the value of every house that is listed on the Zillow website. The goal is to enable users to assess not only how their own home is trending but also to provide voyeuristic pleasure of assessing the wealth of their friends and colleagues (based on the value of their properties).

Zestimate attracted more than 1 million visitors with the first three days of its launch. Subsequently, Zestimate helped Zillow grow its traffic to more than 200 million visitors per month. More than 80% of US houses have been viewed on Zestimate; in order words, Zestimate has helped Zillow attract users even if they are not actively looking to buy or sell a real-estate property. If we assume Zillow’s target audience to be around 200 million (out of the total population of approximately 350 million in USA), we can see that (on average) every person amongst the audience visits Zillow once a month.

It is interesting to note that Zestimate unlocks Brand Value Booster from homeowner’s perspective,. This is because home ownership is (clearly) a high importance activity from customer’s perspective. Towards this, Zestimate encourages homeowners to provide data about major upgrades and repairs to their homes so that Zestimate algorithm can compute the price more accurately. Homeowners have provided details about prices and upgrades for more than 80 million homes. Zestimate, therefore, has contributed in a significant way to build Zillow into the largest and the most well known real estate brand in the USA

As another example of Brand Value Booster, let’s look at AirBnB. AirBnB (short-from of AirBed & Breakfast) started off as an organized and better version of the traditional Bed & Breakfast lodging entities. The AirBnB brand took shape when the team productized the tourist’s desire to “travel like a human”. Towards this, they not only facilitated emotional connect between the hosts and guests (via rich host & guest profiles) but also worked to get hosts involved to provide personalized local experience to the guests. By enabling guests to get authentic local experience (instead of shallower and commercialized touristy experience), AirBnB leveraged the product itself to amplify the emotional connect between guests and the hosts. This was captured brilliantly in the company’s “Belong Anywhere” brand marketing campaign.

In this context, emotional design or emotion-aware design can be treated as an element of Brand Value Booster — products that reflect / complement customers’ emotional and non-functional needs are able to connect better with the customers. AirBnB has been a leading proponent of emotional design and supported “Wish List” feature to capture the aspirational aspect of the customers’ wants. Within four months of its launch, AirBnB reported that 45% of AirBnB users were engaging with it! A small A/B testing experiment reiterated and emphasized the importance of emotional design: changing Wish List icon from “star” to “heart” resulted in 30% increase in engagement! [link] Therefore, it is important to leverage products to establish emotional connect with customers (instead of limiting products only to functional aspects by focusing on features and tasks).

At this point (and using AirBnB’s successful “Belong Anywhere” campaign as an example), it is important to emphasize that efficient and sustainable brand marketing campaigns are often based on the Brand Value Boosters. In other words, before running brand campaigns, it is important to ensure that the product (or product mix) supports high importance activity. Since brand marketing is a “linear” activity (reach / awareness increases in direct proportion to the spends), it is important to build non-linear boosters within the product so that the company can maximize the impact of the brand marketing spends. In the absence of Brand Value Layer, companies need to sustain brand-marketing campaigns to ensure that the brand recall doesn’t atrophy quickly. For example, Nike spent approximately $3.7 billion on advertising and promotion costs in 2019 while Coca-Cola spent $5.8 billion on global advertising and marketing in 2018! (Can Nike and Coco Cola build Brand Value Layers? If yes, how? This is an interesting question — but, for the sake of brevity, we defer exploring this right now.)

To summarize, companies can build different types of Value Layers in a systematic, efficient and sustainable way by building products/features with relevant frequency and importance characteristics. Value Layers are self-sustaining: once products with the right frequency and importance characteristics are built, they yield results whenever they are used. In other words, product-led approach helps to continuously support and grow the Value Stack.

Network Effects

We have talked about Scale, Habit, and Brand so far. What about Network Effects, which have helped create the most value in the Internet era?

Here’s an interesting insight we have uncovered: Network Effects are Value Boosters with one important addition — direct user involvement. If product can get users directly involved in the Scale, Habit, or Brand Value Boosters (i.e., in the products/features that correspond to these aspects of the product), it super-charges the three Value Boosters! This provides the compounding benefit: not only the scale/habit/brand improve due to the use of Value Layers but the Value Layers themselves improve due to direct user involvement. As a result, Network Effects become stronger as a result of user growth. This is the lure and the strength of the network effects: they promise ever-improving product and customer experience!

Network Effect helps to increase the importance of frequent activities and/or helps to increase the frequency of medium importance activities. Figure above shows the “Network Effect Zone” in the Engagement Graph.

Direct users involvement in the three Value Boosters results in three different kinds of network effects:

• Direct user involvement in Scale Value Boosters gives rise to “Viral Networks”
• Direct user involvement in Habit Value Boosters gives rise to “Exchange Networks”
• Direct user involvement in Brand Value Boosters gives rise to “Connected Networks”

We will look at them each of these in more depth subsequently; for the time being, we outline their main characteristics:

Viral Networks are built when current users invite new users to join the network. There are two types of viral networks: (1) acquisition-based viral loops and (2) engagement-based Viral Networks.

Exchange Networks are built when current users engage with each other to improve the experience for everyone. There are three types of Exchange Networks: (1) marketplaces & market networks, (2) platform-based networks (including metadata networks and SaaS-enable Marketplaces — SeMs), and (3) platforms with n-sided network effects (including content & data networks).

Connected networks are built when current users help to build and deepen emotional connect for everyone. There are three types of Connected Networks: (1) social & collaboration networks, (2) community-based networks, and (3) marketplaces with collaboration & same-side network effects.

Even in the absence of direct user involvement, weaker forms of Network Effects are possible. For example, users generate valuable metadata during the course of their engagement with products. This metadata (aggregated over current and past users) can be used to provide better experience to new users. For example, based on past buyer journeys, companies can improve their ability to attract customers, manage leads more efficiently, and to onboard them more effectively.

Network Effects that arise due to indirect involvement of users correspond to the weakest form of network effects and can be referred to as “Indirect Networks”.

Viral Networks, Exchange Networks, and Connected Networks are progressively stronger forms of Network Effects. This is because these Network Effects correspond to three levels of users’ direct involvement in the product. Across these three kinds of Network Effects, user involvement progressively becomes deeper — resulting in increasingly strong Network Effects.

Viral Network Effects

Dropbox grew rapidly due to Viral Network Effect that was based on getting current users involved to invite new users. Dropbox had a very effective two-sided referral program that augmented the inherent virality with additional referral incentives. If a user got a new user to signup, both benefited by getting additional free storage (25MB). In any case, all non-users received a URL that pointed to the files uploaded into Dropbox by the sender. Also, as additional users signed up to use Dropbox, the frequency of engagement increased — users were accessing Dropbox more often (and, therefore, had some elements of Exchange Network Effects).

Viral Network Effects helped Dropbox to quickly acquire more than 500 million users after launching the initial version of the product in 2008. This enabled Dropbox to generate $1B ARR within 8 years of its launch — at that time, it was the fastest SaaS company to hit $1B ARR. [link]

Exchange Network Effects

Google Waze, a navigation app, provides real-time traffic updates and directions to users (travelers) based on inputs provided by fellow travelers. Google Waze was visualized as an Exchange Network right from the inception.

Is it possible to add Exchange Networks on top of existing products? The answer is yes: by adding Habit Value Layer and getting users involved in them. For example, Intuit’s TurboTax consciously added community support to make it more interactive. Scott Cook, chairman and cofounder of Intuit, mentions: “With TurboTax, we’re getting customers to answer people’s tax questions. We’ve created the largest and best source of answers on taxes — if you go to Google and put in a tax question, the link at the top will often be our answer. This is tapping a newer habit from the digital age: participating in online communities.” [link]

Brand Network Effects

PinDuoDuo stands for “shop more together” [link, link, link]. PinDuoDuo launched e-commerce service in September 2015 in the competitive Chinese market to compete with market leaders such as Taobao (Alibaba’s China-focused e-commerce platform), JD.com (part of the Tencent group), Vip.com, etc. The markets had grown rapidly over the last 7+ years to become the largest in the world (at $600B GMV). As a result of hyper-growth over the last several years, the growth-rate of e-commerce was tampering down a bit (though still growing at 30–40% year-on-year rate).

Despite entering a competitive market, PinDuoDuo was able to become the 2nd largest e-commerce player in China within 3 years. It IPOed in Jul 2018 in the US markets with almost $24 billion valuation.

How was PinDuoDuo able to crack open the Chinese e-commerce market? How was it able to compete with Taobao and JD.com? This is because it layered in Viral Network Effects and Brand Network Effects natively in the user experience.

PinDuoDuo focused on making it easier for users to create a new group for purchasing specific items. PinDuoDuo encourages users to form “shopping teams” (a new group) by prepaying for the selected items; after this, they can send a link to invite their friends and family members and encourage them to participate in order to buy the products at a group-purchase price, which is much lower than the normal price. To enrich the shared shopping experience, PinDuoDuo’s product has added many elements of gamification to commerce; for example, users can play games with friends and family to win a shopping coupon. Users are also provided with discounts if they pay for a friend.

This mechanism also allowed PinDuoDuo to unlock powerful Viral Network Effects. PinDuoDuo incentivized social sharing via WeChat, which resulted in rapid adoption and wide reach of the platform. This is the reason why PinDuoDuo’s CAC is $2 (vis-à-vis $18 for Vip.com, $39 for JD.com, and $41 for Taobao). In fact, PinDuoDuo’s CAC has reduced from $5 to $3 to $2 as they grew from 110M active buyers to 240M to 340M [link].

Note that, unlike Groupon, PinDuoDuo prefers teams of people known to one other (instead of teaming up with strangers just to get discount prices). Since “tech-savvy” early-adopters were able to onboard their friends and family members (often in Tier 3 or lower cities), PinDuoDuo succeeded in onboarding a large number of users in Tier 3 and lower cities. Almost 57% of PinDuoDuo users are from Tier 3 or lower cities (compared to 44% for Taobao and 53% for JD.com).

PinDuoDuo also worked to add Habit Value Boosters: there are limited-time offers and lucky draws that seek to get users to visit the app every day. Users can spin a wheel in order to win shopping coupons. Users are provided with cash rewards for checking-in daily.

Amazon has built strong Exchange Network Effects with the help of the user reviews and ratings platform. PinDuoDuo took the network effects to a different level by unlocking Viral and Brand network effects. Together, these mechanisms helped PinDuoDuo to unlock the elusive network effects in the e-commerce experience.

It is important to emphasize that, so far, it has been assumed that Network Effects are dependent on product’s category in the sense that a company can leverage Network Effects if and only if the category intrinsically is dependent on marketplace dynamics, community-based interactions, etc. By identifying different types of Network Effects and their dependence on different types of direct user involvement, we have explained how companies — even those that don’t have natural marketplace mechanics or social networking dynamics — can thoughtfully and systematically craft various types of Network Effects into their products/services. In other words, any company can overlay Network Effects (over their core products) to amplify value creation and to strengthen defensibility.

Summary

We have observed that the three Value Layers (Scale, Habit, and Brand) have distinct product characteristics and, therefore, can be unlocked by adding relevant features / products to the product mix. Value Layers not only enable startups to create more value in a sustainable way but also help to strengthen their defensibility.

We have shown that product-based Value Boosters can be built by identifying relevant category-specific tasks (undertaken by target personas) with specific frequency and importance of engagement characteristics: high frequency and low importance tasks for the Scale Value Boosters, medium frequency and medium importance tasks for the Habit Value Boosters, and high importance and low/medium frequency tasks for the Brand Value Boosters.

In addition, we have highlighted that Network Effects are variants of Value Boosters that can be unlocked via direct users involvement. Product-led user involvement in the three Value Boosters gives rise to three different types of explicit network effects. Direct user involvement helps convert Value Boosters into Network Effects, wherein the product itself continually becomes more energized and stronger with growing number of users.

Entrepreneurs can use the framework to build the Value Layers and Network Effects in a structured way and, thereby, create more value and build sustainable competitive advantage in a systematic, efficient, and sustainable manner.

‍

Also Read:

Looking beyond Silicon Valley for Global Innovation

Creative Innovation & Disruptive Innovation

‍

‍