← Back to list
Colossus (Invest Like the Best / Business Breakdowns)Podcast26 May 2020Source: traffic.libsyn.comHost: Patrick O'Shaughnessy

Shishir Mehrotra – The Art and Science of the Bundle - [Invest Like the Best, EP.175]

In plain words

This piece explains that bundling (selling things together as a package) isn't a rip-off—it creates value by reaching casual fans who wouldn't buy items individually. The author sees fixed pricing as healthier than raising prices, which signals growth is maxed out. Key examples: Spotify (student bundle $5/month with Hulu, actually a profit engine), Netflix (kept $8/month for years, forcing user growth), and Amazon Prime (held price, expanded from shipping to music/video).

AI SummaryAI-generated · may contain errors · verify against the original

Shishir Mehrotra (former YouTube product lead and Spotify board member) delved into the business strategy of "bundling" on the show, pointing out that it is a widely misunderstood yet extremely powerful business concept. He debunked four common myths: that bundling is bad for consumers, the superior

~12 min full read · 6 sections
Deep Analysis

At a Glance

Shishir Mehrotra (former YouTube product lead, Spotify board member, and Coda founder) systematically deconstructs the widely misunderstood business strategy of "bundling." Core thesis: Bundling is not exploitation of consumers, but a powerful mechanism that creates value by activating "casual fans"—benefiting consumers, suppliers, and bundlers alike.


Theme 1: The Four Myths of Bundling — From "Bad Deal" to "Good Business"

Myth 1: Bundling Is Bad for Both Consumers and Suppliers

Shishir Mehrotra argues that the core value of bundling lies not in serving "super fans" but in activating "casual fans."

  • Three user categories: Super fans (willing to pay retail price and actively seek the product), casual fans (lack willingness to pay or motivation to seek), and non-fans (assign zero or negative value to the product).
  • Mechanism: When sold individually, each product only reaches super fans; when packaged as a bundle, suppliers capture revenue from casual fans, and consumers gain access to products at the casual-fan level.
  • Example: Spotify replaced iTunes' $0.99 per-song purchase model — users can now listen to a vast amount of music they "would never pay for individually," which is precisely the value bundling creates.
  • Comparison data:
Business Model User Type Reached Value Creation Direction
A la carte sales Only super fans Deeply satisfies core users
Bundle sales Super fans + casual fans Expands user base and use cases

Myth 2: Bundle Revenue Should Be Allocated by Usage

Mehrotra points out that the correct allocation metric is "marginal churn contribution" (MCC), not usage.

  • MCC definition: The percentage of users who would churn if a specific product were removed from the bundle.
  • Example: ESPN and History Channel have similar usage levels, but ESPN generates approximately $4.50 per subscriber per month, while History Channel generates only $0.20 — because removing ESPN would cause 20 times the user churn compared to removing History Channel.
  • Formula relationship: Wholesale price ≈ Retail price × Super fan share. If ESPN's super fan share is 10%, its fair retail price should be $50/month (not the $5–8 users assume).
  • Mechanism: MCC incentivizes suppliers to build "indispensable" products rather than chasing high usage.

Myth 3: Bundling Deprives Consumers of Choice

Mehrotra believes the root cause of consumers feeling "ripped off" is the lack of transparent individual pricing, not bundling itself.

  • Core argument: To properly evaluate a bundle's value, consumers must know the fair standalone price of each component.
  • McDonald's analogy: Customers see a Big Mac for $5, fries for $3, a drink for $2, and a combo for $7 — transparent pricing makes the "free drink" perceived value work.
  • Cable industry dilemma: 40 years of evolution have made individual pricing completely opaque. Users mistakenly think "ESPN alone costs $7," when the fair retail price is actually $50 — so a $50 cable bundle suddenly seems to "give away 299 channels for free."
  • Implication: The cable industry will undergo a process of "unbundling → transparent pricing → rebundling."

Myth 4: Bundles Should Only Include Similar Products

Mehrotra offers a counterintuitive conclusion: The best bundle should minimize super fan overlap and maximize casual fan overlap.

  • Core logic: If three products have completely non-overlapping super fans, every user in the bundle is a "newly acquired" casual fan; if super fans heavily overlap, the bundle merely discounts existing users without creating new value.
  • Historical lesson: In 2012, when YouTube bid for NFL Sunday Ticket ($400/year), Mehrotra suggested bundling it with "Food Network" and "Knitting Channel" — Google executives rejected the idea ("Why not just invent a sport?"). In hindsight, this was precisely the strategy to maximize casual fan overlap.
  • Venn diagram thinking: Products have a "super fan core" and a "casual fan periphery." The ideal bundle makes the cores as non-overlapping as possible and the peripheries as overlapping as possible.

Theme 2: Bundler Strategy – Pricing Power and Growth Paths

Pricing Power as a "Negative Signal": Fixed Pricing Is Healthier Than Price Hikes

Mehrotra argues that bundlers raising prices often signals that user growth has been exhausted, while fixed pricing forces companies to continuously expand their user base.

  • Netflix Case: Reed Hastings long maintained a price of $8/month, refusing to raise it as content expanded. This forced the company's only growth path to be "acquiring new users" rather than "extracting more from existing users."
  • Mechanism: Fixed pricing + continuous addition of new content = incremental value from each new "ordinary fan"; price hikes = a saturated market of super fans.
  • Amazon Prime: For years without price increases, it continuously added seemingly "unrelated" services like music, books, and video, essentially expanding the base of ordinary fans.
  • Investment Perspective: Investors often view "pricing power" as a positive signal, but Mehrotra sees it as a warning that "growth headroom has peaked."

Bundler Strategic Goals: Reduce Super Fan Overlap + Expand the Ordinary Fan Base

  • Supplier Incentives: Suppliers should aim to enhance their own MCC (making users dependent on them), which naturally leads to "building better products."
  • Bundler Incentives: Continuously expand the bundling scope into new super fan domains while maintaining competitive pressure on existing suppliers.
  • Case: Spotify's expansion into podcasts – the overlap of super fans between music users and podcast users is small, but the overlap of ordinary fans is large, making it an ideal expansion direction.

Theme 3: The Future of Bundling – The Internet’s “Third Business Model”

Subscription Bundling Will Become the Third Major Internet Business Model After Advertising

Mehrotra predicts that bundling will give rise to a new wave of businesses that “can only survive through bundling.”

  • Historical analogy: In the mid-2000s, Google AdSense created the “non-fan business model” (ad-supported), spawning an entire ecosystem of blogs, podcasts, social networks, and more.
  • Now: Subscription bundling is emerging as a similar platform-level opportunity—technical barriers have lowered (OAuth, referral codes, etc.), and bundlers can set up with just “a few meetings.”
  • Future: Super-bundles far exceeding the scale of cable will emerge, covering seemingly unrelated areas such as media, logistics, healthcare, and dry cleaning.
  • Key judgment: Some products are highly valuable to ordinary fans but insufficiently valuable to superfans—their only path to market is reaching users through bundling.

Case Study: Spotify Student Bundle – The Misunderstood “Discount” Is Actually a Profit Engine

  • Content: Spotify + Hulu student plan at $5/month.
  • Mechanism: Among students, very few are superfans who pay for both products simultaneously → each new user is a “pure incremental” ordinary fan → both sides achieve high margins.
  • Insight: What appears to be a “loss-making” bundle is in fact a classic move to maximize overlap among ordinary fans.

Mentioned Positions

Position Analyst View Key Data
Spotify Bullish (as a bundling case) Student bundle $5/month (Spotify + Hulu); podcast expansion strategy
Netflix Bullish (pricing strategy model) Long-term maintenance at $8/month; original content investment diluted as user base scales
Amazon Prime Bullish (fixed price + continuous expansion) No price increase for years; expanded from free shipping to music, books, video
YouTube Neutral (historical case) Bid for NFL Sunday Ticket in 2012 ($400/year); ad-driven, subscription attempts failed
DirecTV Neutral (case) NFL Sunday Ticket annual fee approximately $1 billion; claimed during Weather Channel negotiations it would lose 1.6 million subscribers (actual target achieved with only 20,000–40,000 churn)
ESPN Neutral (MCC case) Wholesale price $4.50/month/subscriber; superfan share approximately 10% → reasonable retail price $50/month
History Channel Neutral (MCC case) Wholesale price $0.20/month/subscriber; usage similar to ESPN but MCC 20x lower
Coda Bullish (own company) Bundles documents, spreadsheets, apps; users view it as "better docs," "better spreadsheets," "better apps" respectively
UFC Neutral (comparison case) Pay-per-view $50/event; high superfan share, few casual fans
NFL Neutral (comparison case) Reaches many casual fans via cable bundling; Sunday Ticket annual fee $400

Judgments Worth Remembering

1. “Bundling does not create superfans; it creates ordinary fans.” (Mehrotra) — Every product has three types of users: those willing to pay and actively seeking it (superfans), those with demand but unwilling/unable to pay proactively (ordinary fans), and those with zero or negative value (non-fans). The core value of bundling lies in activating the second type.

2. “Bundling revenue should be allocated by ‘Marginal Churn Contribution’ (MCC), not by usage.” (Mehrotra) — ESPN and the History Channel have similar usage levels, but ESPN’s revenue is 20 times higher because removing it would cause 20 times more user churn. MCC incentivizes suppliers to build “indispensable” products.

3. “Consumers think cable bundling is a scam because individual product pricing is opaque — not because bundling itself is flawed.” (Mehrotra) — If ESPN’s fair retail price is $50/month (based on its 10% superfan share), then a $50 cable package effectively “gives away 299 channels for free.” McDonald’s meal deals succeed because individual item prices are transparent.

4. “The optimal bundle should minimize superfan overlap and maximize ordinary fan overlap.” (Mehrotra) — When YouTube bid for NFL Sunday Ticket in 2012, Mehrotra suggested bundling it with the Food Network and the Knitting Channel. Google executives rejected the idea (“Why not just invent a sport?”). In hindsight, this was precisely a strategy to maximize ordinary fan overlap.

5. “Price increases by bundlers signal that growth headroom is exhausted, not that pricing power is strong.” (Mehrotra) — Netflix long maintained $8/month, forcing the company’s only growth path to be acquiring new users; cable companies, constrained by geographic monopolies, cannot expand and can only raise prices. Investors should be wary of the growth bottlenecks behind price hikes.

6. “Subscription bundling will become the internet’s third business model — following advertising.” (Mehrotra) — Google AdSense created the “non-fan business model” (ad-supported), giving rise to blogs, podcasts, and social networks. Now, lower technical barriers make subscription bundling feasible, spawning a wave of new businesses that “can only survive through bundling.”

7. “Spotify’s student bundle ($5/month including Hulu) is not a discount, but a profit engine.” (Mehrotra) — Among students, very few are superfans paying for both products simultaneously; every new user is pure incremental ordinary fan revenue. Seemingly loss-making, it is actually a classic move to maximize ordinary fan overlap.

8. “Eigenquestion: Among many questions, find the key question that can answer all others.” (Mehrotra) — Taking “how to bring a teleportation device to market” as an example, just two questions (safety + cost structure) can derive four business models (human fax machine/airport network/freight network, etc.). This is the core technique of framing questions.