The Janus Rule, the Misjudgment of Solomon, and a critique of Aggregation Theory

Anthony Bardaro
Jul 30 · 24 min read
👆 choose GREEN to annotate this article or PINK to get a summary 👆

Two’s company but three’s a crowd: Shopify, Amazon, and Etsy (in no particular order)

Shopify has finally attracted mainstream attention of late, due to the company’s remarkable (and remarkably quiet) success in one of this millennium’s hottest, most competitive markets: ecommerce. That should come as no surprise, as the Shopify platform had strategically occupied a blindspot in Amazon’s territorial map, as discussed in “Shopify and the Power of Platforms”:

Aggregators tend to internalize their network effects and commoditize their suppliers, which is exactly what Amazon has done… Amazon benefits from more 3rd-party merchants being on its platform because it can offer more products to consumers and justify the buildout of that extensive fulfillment network; 3rd-party merchants are [atomized, modularized, commoditized, and thus] mostly reduced to competing on price.

That, though, suggests there is a platform alternative — that is, a company that succeeds by enabling its suppliers to differentiate and externalizing network effects to create a mutually beneficial ecosystem. That alternative is Shopify.

At first glance, Shopify isn’t an Amazon competitor at all [but the] difference is that Shopify is a platform: instead of interfacing with customers directly, [its] 3rd-party merchants sit on top of Shopify and are responsible for acquiring all of those customers on their own.

Janus (or, the two faces of ecommerce: Amazon and Shopify)

In other words, Amazon built a behemoth by aggregating commoditized suppliers under the Amazon banner, and that naturally provided a market opportunity for a complement — a mirror opposite — which Shopify launched with its platform for differentiated suppliers, who operate under their own banners, using the ecommerce infrastructure of an otherwise faceless Shopify. That essentially describes the “Janus Rule”…

The Janus Rule: Markets like these tend to accommodate two complementary businesses that are mirror reflections of one another — one being an aggregator of commoditized suppliers and the other a platform for differentiated suppliers.

In other words, two’s company; three’s a crowd. As such, juxtaposed against those two faces of ecommerce — who are nestled neatly in their respective moats — Walmart has largely failed in its multi-decade assault on Amazon. The retail giant originally came-up short due in large part to the classic Innovator’s Dilemma, as discussed earlier in the piece excerpted above:

[A]fter years of trying to leverage its stores in e-commerce, Walmart realized that Amazon was winning because e-commerce required a fundamentally different value chain than retail stores… the proper response to that recognition was not to try to imitate Amazon, but rather to focus on areas where the stores actually were an advantage, like groceries, but it’s worth understanding exactly why attacking Amazon head-on was a losing proposition…

[While Walmart was taking 19 years to mount its assault], Amazon.com effectively bifurcated itself into a retail unit and a fulfillment unit: The old value chain is still there — nearly half of the products on Amazon.com are still bought by Amazon at wholesale and sold to customers — but 3rd parties can sell directly to consumers as well, bypassing Amazon’s retail arm and leveraging only Amazon’s fulfillment arm…

Walmart and its 20 distribution centers don’t stand a chance [against Amazon’s 300], particularly since catching up means competing for consumers not only with Amazon but with all of those 3rd-party merchants filling up all of those fulfillment centers.

Once Amazon had reached scale, the cat-was-out-of-the-bag. At that point, we know an outsider cannot play offense or defense against an incumbent on the incumbent’s home turf — you cannot attack them half-heartedly or head-on like Walmart did Amazon — per “The Carrot and the Stick”:

New entrants can only sustain were they to aim for “new market disruption” in which they occupy incumbents’ blindspots — completely new vectors where incumbents can’t squash upstarts for fear of cannibalizing their own, preexisting cash cows. Think in terms of Google disrupting Microsoft by organizing the chaotic web — as opposed to Firefox just building a better browser than Internet Explorer. Don’t build a better mouse-trap; build a waterwheel the mouse can spin to produce hydroelectricity.

In contrast to Walmart’s greatest failure, the occupation of Amazon’s blindspot was perhaps Shopify’s greatest success, which actually serves to ratify parts of Disruption Theory.

At the same time, the divergent fortunes of Shopify and its half-sister, Etsy, provide an interesting empirical case study that ultimately exposes weaknesses in today’s generally-accepted Aggregation Theory. Furthermore, this Shopify/Etsy bifurcation demands specific refinements to the emergent Moat Map doctrine…


Aggregation Theory and the Moat Map

To grok all of this, you have to understand Aggregation Theory and the Moat Map themselves. The first is a pillar of tech’s academic business strategy, as developed by Stratechery’s Ben Thompson throughout this decade and officially codified by him in 2015. The second is a more nascent framework, proposed in 2018 by the same Ben Thompson, along with the Map’s cartographer (of sorts), James Allworth.

In an effort to make sure everyone’s speaking the same language, here’s a great description of Aggregation Theory, which articulates the difference between modern tech businesses and traditional analog ones, from a mashup between Ben’s own Stratechery newsletter and Adventures in Consumer Technology:

Big Tech’s markets all trend toward monopoly, because they operate multi-sided networks with zero barriers-to-entry. Thus, network liquidity is the basis of competition for many of them: Who has the most buyers and sellers; the most producers and consumers; the most supply and demand; etc. That liquidity sets-off the virtuous cycle of network effects, wherein scale improves user experience improves scale and so on. If you add software’s zero marginal costs to that virtuous cycle, you get [Ben Thompson’s original] Aggregation Theory:

“[A]n aggregator can secure a critical mass of consumers that gives them power over suppliers; suppliers are then incentivized to deliver their product to the aggregator’s specifications which improves the overall experience, allowing the aggregator to further increase their consumer base, which further strengthens their bargaining position relative to suppliers. To be fair, this is another way of describing a two-sided market, but what makes aggregators unique is the zero marginal distribution costs enabled by the Internet and the zero marginal transaction costs enabled by computers, which means such companies can scale to basically the entire world.”

Pursuant to that, here’s a description of The Moat Map, from Stratechery’s eponymous article, which tried to articulate the duality of aggregators (like Facebook and Google) vs platforms (like Microsoft and Apple):

The Moat Map

This relationship between the differentiation of the supplier base and the degree of externalization of the network effect forms a map of effective moats [for example]:

• Facebook has completely internalized its network and commoditized its content supplier base, and has no motivation to, for example, share its advertising proceeds. Google similarly has internalized its network effects and commoditized its supplier base; however, given that its supply is from 3rd parties, the company does have more of a motivation to sustain those third parties (this helps explain, for example, why Google’s off-site advertising products have always been far superior to Facebook’s)…

• Apple and Microsoft, meanwhile, have the most differentiated suppliers on their platforms, which makes sense given that both depend on largely externalized network effects. “Must-have” apps ultimately accrue to the platform’s benefit.

With those definitions now out of the way, let’s turn back to the issue at hand. Enough with the jargon; on to the meat-on-this-bone…


The Shopify/Amazon crossover episode

Given Shopify’s success, there’s a real question as to when it will attack Amazon on Amazon’s turf. In other words, when will Shopify cross-over from being a mere platform for distributed 3rd party merchants to aggregating its own centralized ecommerce destination?

That’s definitely a question of when — not if — as I discussed upon the advent of the Moat Map back in 2018:

I don’t see the empirical evidence to suggest that a strict proprietary (aggregation) approach — nor a strict 3rd party (platform) approach — is sustainable or desirable ad infinitum… it’s worth adding a time element to this: If a company starts in one end of the moat, it seems like they inevitably have to expand toward the middle. Furthermore, I don’t think they get pulled to the middle by a Siren’s Song; I think they just have to go there eventually — in pursuit of growth, defensibility, and vertical/horizontal integration…

Without incorporating dynamism, the Moat Map framework would lead to misdiagnoses like Amazon remaining a proprietary ecommerce play (no Fulfillment or AWS); Netflix procuring only 3rd party content (no Originals studio); Apple foreclosing on its developer ecosystem (no App Store).

Moat Map 2.0: Flips the axes to put suppliers/supply as the explanatory variable

No matter how secure Shopify’s platform moat, they’ll eventually be compelled to migrate into the aggregation business. Indeed, Amazon already made its own migration (in reverse). Bezos’ team started by aggregating products that it procured from upstream suppliers — merchandise held in Amazon’s inventory; marketed via Amazon’s digital storefront; then sold by Amazon to downstream consumers. Eventually, Bezos & Co launched a pseudo-platform, Fulfilled by Amazon (FBA), which let its suppliers outsource inventory, shipping, and customer service to Amazon. The same suppliers as always were able to continue providing the same upstream product to Amazon, and the same consumers as always were able to continue purchasing the same downstream product from Amazon, but Bezos’ crew now gave the former an option to outsource all of their midstream needs to Amazon, which cast a wider net around the supply-side — more gross merchandise volume (GMV) for Amazon and more selection for its end-users:

Amazon’s expansion from FBM aggregation into the FBA platform resulted in domination of the entire hemisphere — or, at least, the most desirable territories therein

From the first, Shopify did much the same for its upstream producers as did FBA for Amazon’s — except Shopify merchants have always been the face of their own supply chains: Consumers interact entirely with the merchant in the merchant’s own storefront, which is merely supported by Shopify’s invisible integration into the midstream logistics. That’s a stark contrast to those merchants in Amazon’s orbit, in which all products are enveloped in Amazon branding front-and-center.

While migration toward the center of the Moat Map is inevitable (and advisable), it’s also worth noting that nobody should start in the middle. To paraphrase “The Four Winds of Modern Media”:

Given today’s [abundance], survivors start at extremes of some combination between [platform/aggregation] and scale/niche… You cannot start in the middle of either spectrum and grow out‬, because your competitive disadvantages let your rivals squash you from the outside-in.

Amazon’s migration through the Moat from its initial extreme to a more centered disposition began with its FBA launch. Whereas Amazon exclusively benefited from its original aggregation model (Ben’s “internalized network effects” with the economies of scale accruing to Amazon’s own bottom line independent of its suppliers’), FBA started to let Amazon redistribute some of those gains back into the supplier ecosystem (“externalized network effects” with Amazon’s economies of scale getting leased-back to its suppliers in a more co-dependent model). Accordingly, Shopify started in the opposite pole, with its benefit largely externalized unto merchants in a mutually-dependent arrangement.

This duality makes sense when you consider the difference between the supplier cohorts: Per simple supply and demand economics, Amazon’s commoditized suppliers warrant less excess profit for themselves, whereas Shopify’s differentiated suppliers command more. That’s an important point about the nature of supply. In fact, back in 2018, I postulated that this was not only one of the most important explanatory variables for influencing business strategy, but also a variable entirely mishandled by the legacy Moat Map:

Regardless, arguing that Twitter should be a platform — as opposed to an aggregator — strikes me as backward reasoning. In other words, you started with the conclusion that Twitter should be/should have been a platform, therefore Twitter’s content must be differentiated in order to fit it into the advisable “moat”. In reality, I think the explanatory variable here is the commodity supply [user-generated content “UGC”], and therefore what follows is an aggregator/internalized/proprietary strategy — at least per the “Moat Map” framework’s prescription. (FWIW, the nature of supply seems to be the explanatory/independent variable, so I’d also suggest making that the X-axis on Ben’s 2x2.)

The arrow of causation should point the opposite direction than the Moat Map’s design implies. And if you think the presentation of the Moat Map’s axes was just cosmetic, the Map’s cofounder, Ben Thompson himself, recently seems to have suggested that the explanatory variable is deliberately “network effects”, not ‘the nature of suppliers’:

The Moat Map discussed the relationship between network effects and supplier differentiation: the more that network effects were internalized the more suppliers were commoditized, and the more that network effects were externalized the more suppliers were differentiated.

Therein, Ben seems to clearly state that a startup’s flavor of network effects (internalized/externalized) dictates the nature of the supply it should target (commoditized/differentiated) — but, again, it’s obviously the other way around.

Shopify began with the insight that there was an opportunity among highly differentiated merchants — the kind who are so niche/unsubstitutable that they can successfully operate destination sites. This was orthogonal to Amazon, not to mention a pocket of turf that had been relatively neglected/underserved in the ecommerce space. This insight is the essence of the aforementioned Janus Rule.

Starting with that independent variable (“differentiated supply”), the logic of choosing whether to be a platform or aggregator — fulfillment or storefront — is simple, as follows, from “Ditching Uber’s Playbook”:

With that explained, the upshot is that a digital business wants to strategically occupy the moat highlighted above, because:

• If suppliers are highly commoditized, then a new entrant should optimize for a proprietary solution (e.g. supply is easy to procure/too abundant, thus aggregation is the major attraction for demand);

• If suppliers are highly differentiated, then a new entrant should optimize for 3rd party solutions (e.g. supply is hard to procure/too scarce, thus a platform is the major attraction for suppliers)

Accordingly, occupying the other two quadrants outside of that moat just doesn’t make sense:

• If suppliers are highly commoditized, then 3rd party solutions either exacerbate abundance by adding another layer of redundant fragmentation (increasing deadweight loss) or create products inferior to that of an aggregator;

• If suppliers are highly differentiated, then a proprietary solution either exacerbates scarcity by adding another layer of redundant aggregation (invoking the agency problem) or creates a product inferior to those of specialized 3rd parties).

Even though the R&D required to build an aggregation storefront is a relatively low-bar for a company of this size, it’s still tough for Shopify to compete in that new surface area, because nobody has a problem that needs this particular solution. Fulfillment has always been the biggest pain-point for Shopify’s current clientele, who, again, are highly differentiated merchants with direct-to-consumer (DTC) destination sites and relatively low unit sales (at least compared to the average FBA merchant). Hence, these suppliers all need access to some sort of economies-of-scale — without which the logistics required to run all of these global ecommerce businesses independently are uneconomical at such low individual output rates. Like Amazon, Shopify amortizes large fixed costs and spreads them out across 820k fragmented merchants, who each pay a small marginal cost every time they tap these combine resources. As such, Shopify has converted all of what would have been big chunks of uneconomical upfront capex (borne duplicatively by every individual merchant) into a tiny bit of opex (borne marginally by every sale).

That’s how Shopify is endowed today — and whom it’s endowed with. Supplementing that franchise with a storefront under the Shopify banner would be a tough second-act, not due to development or tech or anything that trivial, but because of what the decision tree excerpted above says:

• If suppliers are highly differentiated, then a proprietary solution either exacerbates scarcity by adding another layer of redundant aggregation (invoking the agency problem) or creates a product inferior to those of specialized 3rd parties).

…among other factors (e.g. differentiated supply is “hard to procure” or aggregate for all the Economics 101 reasons I’ve discussed before).


The Etsy case study

What’s changed between 2018 — when I originally published my hypothesis about the Moat Map’s direction of causation — and today is the body of evidence. We now have a control group with which we can evaluate the null hypothesis. To wit, Etsy is an interesting company to discuss within the framework of my proposal from above:

Starting with that independent variable (“differentiated supply”), the logic of choosing whether to be a platform or aggregator — fulfillment or storefront — is simple[.]

It strikes me that Etsy pursued the same original insight as Shopify, specifically neglected/underserved merchants in a niche orthogonal to Amazon’s services. We can substitute “Etsy” for “Shopify” in the following, which I’ve copied from the excerpt above:

[Etsy] began with the insight that there was an opportunity among highly differentiated merchants — the kind who are so niche/unsubstitutable that they can successfully operate destination sites. This was orthogonal to Amazon, not to mention a pocket of turf that had been relatively neglected/underserved in the ecommerce space.

…but the problem is that Etsy chose an aggregation strategy over the platform one. Etsy is a wild success, no doubt, but its chosen path also seems to have led to a smaller market opportunity than Shopify’s: Etsy has a market cap of $7.8B vs $30.6B for Shopify; revenues of $790M vs $2B; and 5yr EBITDA CAGR of 47% vs 71%. As such, Etsy is a cautionary counterfactual that’s rather susceptible to Shopify expanding into its territory and subsuming its business — like Amazon did to everyone else in its own hemisphere, including Walmart and Ebay.

Point is, the Moat Map framework (with the arrow of causation pointing in the right direction) is a testament to Shopify’s path-of-least-resistance and Etsy’s path-of-most-resistance. Here is the route Etsy chose (and Shopify avoided):

[Addressing this target market] with a storefront kind of business under the [Etsy] banner would be [tough], not due to development or tech or anything that trivial, but because… If suppliers are highly differentiated, then a proprietary solution either exacerbates scarcity by adding another layer of redundant aggregation (invoking the agency problem) or creates a product inferior to those of specialized 3rd parties)… among other factors (e.g. differentiated supply is “hard to procure” or aggregate for all the economics 101 reasons I’ve discussed before).

The Judgment of Solomon (except they’ll actually cut this baby in half!)

Perhaps Etsy started with the arrow of causation pointing the wrong direction (i.e. ‘Our thesis here at Etsy is aggregation, so therefore we will aggregate underserved differentiated suppliers because nobody else is’). Or, maybe Etsy never considered independent/dependent variables and correlation/causation. Regardless, Etsy and its own mirror reflection, Ebay, made what we’ll call “The Misjudgment of Solomon”, representing an upstart’s erroneous choice to settle for only half the mandate befitting of its target market’s needs…

The Misjudgment of Solomon: When a startup occupies an undesirable locus outside of the Moat by attempting to be either an aggregator of differentiated suppliers (rather than serving as the platform its market demands) or a platform for commoditized suppliers (rather than the aggregator needed).

Such misjudgments fall outside of the Moat occupied by the truly righteous heirs in the proper duality — in this case, outside of the quadrants occupied by Amazon and Shopify.

Etsy and Ebay mired in the Misjudgement of Solomon and Walmart stuck in the middle

To be sure, we can probably debate how differentiated/unsubstitutable/etc Etsy’s suppliers really are (particularly compared to Shopify’s). For example, if Etsy’s goods have higher demand elasticity compared to those of Shopify merchants, then maybe aggregation wasn’t as bad of a strategic choice as it would have been for perfectly inelastic SKUs. But, even then, we could still question why Etsy chose the relatively smaller TAM associated with aggregating differentiated supply, given the discretion to have chosen platform for the same suppliers.

Regardless, Etsy vs Shopify is an interesting empirical study, especially since the companies were founded around the same time: 2005 vs 2006, respectively. Of course, I’m looking at this from today’s perspective, which is a mere snapshot in time. As the errors of Spotify’s ways will attest (i.e. Ek’s Parlay and podcasts), anything can happen in the future, which begs the question: What’s the strategically sound way for Shopify to expand its own empire?


A Faceless Man

Shopify should expand from its original platform (“OG”) into a faceless aggregator (“FL”), which should result in its domination of its own hemisphere (a la Amazon FBM+FBA)

In the same way that Amazon took its aggregator franchise under its own banner then ported it to a new platform that was kept under the Amazon banner (FBA), Shopify should do the mirror opposite to expand in its own right, per the Janus Rule: Shopify should take its faceless platform and port it to a new faceless aggregator. This approach would involve Shopify making bulk purchase orders with select merchants — as does Amazon-the-aggregator — then taking title to hold that merchandise in Shopify’s own inventory, but having it marketed via the merchant’s digital storefront and sold by the merchant to downstream consumers in the same was it is today.

Compare that to the arrangement of Amazon’s original and prevailing aggregator, Fulfilled by Merchant (FBM), as already discussed above:

Bezos’ team started by aggregating products that it procured from upstream suppliers — merchandise held in Amazon’s inventory; marketed via Amazon’s digital storefront; then sold by Amazon to downstream consumers.

The difference between Amazon’s FBM and Shopify’s hypothetical equivalent would be Shopify’s facelessness: All customer-facing activity operates under the merchant’s banner — not the aggregator’s.

Remember, this starts from the assumption that Shopify’s merchants are highly differentiated destinations with DTC relationships. They have significant brand goodwill, so they don’t need Shopify’s (unknown) brand to unlock incremental sales. What they need is opportunistic capacity — the ability to throttle production to meet regular and irregular demand spikes like seasonality or hypergrowth.

A Faceless Man

For example, a Shopify merchant who’s already growing 80% year-over-year probably won’t have the capacity to meet an influx of holiday demand because he/she doesn’t have the spare production capacity or cash flow available. Were Shopify to track this merchant’s growth trajectory and forecast the holiday spike, they could place a big bulk order in July that would provide the merchant with enough cash and calendar to ramp-up in time for the year-end rush. (This could occur via reverse enquiry too, with the merchant asking Shopify for the assistance rather than Shopify having to bid for all business on an unsolicited basis.) Since an aggregator can carry inventory at a much lower cost of capital than a relatively diminutive merchant, Shopify would take this freshly-financed product off of the merchant’s hands, but the merchant’s end-buyers would still interface with him/her directly since his/her brand is already like a supernova with its own gravitational pull. Ultimately, Shopify would make its money on the retail markup above its wholesale-esque costs. (Shopify could potentially have advertising opportunities in the future too, but only well after it’s expanded beyond the most perfectly unsubstituable merchants in its universe.)


The Amazon and Netflix foils

Etsy’s aggregation headwind (relative to Shopify’s platform tack) demands another hard look at Aggregation Theory and the Moat Map. For instance, both frameworks count Amazon and Netflix as archetypal examples of aggregators — but neither fits the definition.

Let’s start with Amazon. No part of Amazon’s business can be characterized by “zero marginal costs”, a charge levied by the definition of Aggregation Theory above:

[T]his is another way of describing a two-sided market, but what makes aggregators unique is the zero marginal distribution costs enabled by the Internet and the zero marginal transaction costs enabled by computers, which means such companies can scale to basically the entire world.

To be fair, Ben Thompson has since clarified, changing that from a “zero marginal cost” hard-line to a much softer “effectively zero marginal costs” qualifier. Yet, even with that lower-bar, Amazon’s marginal costs are by no means negligible — or even low — and they never have been.

Ben’s model considers companies, like Amazon, who invest in enormous amounts of fixed cost infrastructure, such that incremental transactions (e.g. pageviews) have effectively zero marginal cost. In reality, not only is there a non-zero marginal cost to serve pageviews (i.e. compute albeit small for a company of Amazon’s size), but there is also significant marginal cost associated with actual revenue-generating transactions — whether fulfillment be for its ecommerce or cloud storage/computing business lines.

So, Ben’s model is purely theoretical — an idealized extreme that cannot exist in reality. In practice, infrastructure is actually a capital expenditure for Amazon that increases in a step-like function (as opposed to a linear function): Amazon builds a data center and fills it with servers/sensors/gear/people; then, when demand increases to fill that data center’s capacity utilization, Amazon needs to provide more slack by either building another data center or adding another chunk of servers/sensors/gear/people to the preexisting one. These are formidable marginal costs — to say nothing of the transaction and distribution costs to fill an actual customer order.

Furthermore, to flip this on its head, consider that most businesses in today’s economy run on clouds like Microsoft Azure/Google Cloud/Amazon Web Services (AWS), so the costs borne by such tech-infused businesses are almost purely marginal. Sure, from startups to SMBs to Fortune 500 giants, modern businesses get to forgo the expensive fixed costs associated with setting up their own infrastructure up-front, but that savings just gets deferred to the back-end, increasing their variable costs.

When Annotote adds more users to its network, it pays an incremental cost to AWS for the storage and compute associated with its new users’ throughput. Granted, that’s been a decreasing marginal cost as Annotote has scaled — a la exponential decay — but it’s not “effectively zero” nor even “low”. (It’s not even low relative to gross margins or unit economics!)

Were Ben just talking about decreasing marginal costs, then he’d be describing mere economies of scale; but he specifically underscores “zero” as “what makes aggregators unique”.

Now let’s talk about Netflix. I’ve been building the case against Netflix’s aggregator status for years, because it violates Ben’s own criteria, including:

  1. Aggregators experience “declining customer acquisition costs” (CAC) as they scale
    ^ Netflix’s CAC has increased
  2. Aggregators are unique due to their application of network effects
    ^ Netflix doesn’t feature network effects
  3. Aggregators can scale to the entire world because of “zero distribution costs and zero transaction costs”
    ^ Netflix features substantial distribution and transaction costs
  4. Aggregators trend toward winner-take-all monopolies by virtue of these first three items
    ^ Netflix’s market has not trended toward winner-take-all nor monopoly; in fact, competition has increased
  5. Aggregators own the transaction “where money actually changes hands”
    ^ Netflix doesn’t control its point-of-sale (e.g. app stores and cable bundles)
  6. Aggregators commoditize, modularize, and atomize supply
    ^ Netflix hasn’t managed to accomplish any of this; in fact, content has grown more valuable

Ben has tried to accommodate these caveats by classifying Netflix as a “Level 1 Aggregator”, the description of which is wholly indistinguishable from a traditional category killer, according to Stratechery’s “Defining Aggregators”:

Level 1 Aggregators acquire their supply; their market power springs from their relationship with users, but is primarily manifested through superior buying power. That means these aggregators take longer to build and are more precarious in the short-term.

The best example of a Level 1 Aggregator is Netflix. Netflix owns the user relationship and bears no marginal costs in terms of COGS, distribution costs¹, or transaction costs²… the more content Netflix acquires, the more its value grows to potential users. And, the more users Netflix gains, the more it can spend on acquiring content in a virtuous cycle.

Level 1 aggregators typically operate in industries where supply is highly differentiated, and are susceptible to competitors with deeper pockets or orthogonal business models.

¹ Obviously bandwidth in the aggregate is a particularly large cost of Netflix

² In all cases, credit card fees excepted

Beyond the casual footnotes regarding non-zero marginal costs 😏, note that these observations apply just as well to yesteryear’s industrial leaders — whether Carnegie’s US Steel, Rockefeller’s Standard Oil, or Bell’s AT&T — who all increased the value of the end-user experience via vertical and horizontal expansions. These aren’t novel concepts. Even “effectively zero marginal costs” are not markedly different than those featured by early TV and radio, which spread across geographies with low marginal distribution costs and effectively zero marginal cost to end-users.

In sum, Aggregation Theory appears to be more of an observation about economies of scale, increasing customer selection, and decreasing consumer price than anything else. That’s a thread I want to pull on…


Aggregation Theory and the Moat Map, disrupted

Shopify isn’t even dealing with the network effects required to subject the business to Aggregation Theory or the Moat Map. Shopify’s benefits are more economies of scale than network effects, per the definition of network effects in “The Special Internet Standard”:

[T]he effect… that an additional user of a good or service has on the value of that product to others. When a network effect is present, the value of a product or service increases according to the number of others using it…

Network effects are commonly mistaken for economies of scale, which result from business size rather than interoperability… Interoperability has the effect of making the network bigger and thus increases the external value of the network to consumers… primarily by increasing potential connections and secondarily by attracting new participants to the network.

That excerpt goes on to give the following example, in which you can more closely relate Shopify to The New York Times than Facebook:

Whether a blog, BuzzFeed, or The New York Times, an additional reader doesn’t increase the value of the product for other readers. In contrast, the value of Facebook, Twitter, and Google increases with every incremental new user — a positive externality with a positive feedback loop.

So, in addition to my reversing the Moat Map’s aforementioned “arrow of causation”, I also flagged its problematic use of “network effects” as a scale on its other axis. At the very beginning I proposed changing the scale to a spectrum between “proprietary” and “3rd party” rather than “internalized” and “externalized”. Beyond making the Map more descriptive and intuitive, this modification would also rectify its misuse of “network effects”.

My takeaway from this whole exercise is that the lower the marginal costs, the higher the propensity for aggregation, which is more of a synonym for consolidation than anything else. Nothing is truly zero marginal cost, but a lot of these tech businesses and markets are materially lower marginal costs than their predecessors — due to some sort of innovation that predicated a supply or demand shock — and thus, in most cases, the plane of competition has been predicated on scale, which has ratcheted-up to a degree that predecessors could never have imagined: e.g. Radically lower marginal costs enabled regional scale to become global scale and thousands to become billions in exponentially shorter order than ever before. That scale is central to Ben’s “user experience” criteria, which is the gateway to entering the slipstream of what I’ve called the “virtuous cycle” — scaling with the tailwind of both network effects and virality — albeit rare in its purest form (e.g. Facebook not Netflix).

That clarification is particularly helpful in articulating the difference between the two ways everyone talks about aggregation…

  1. Aggregation (as in Aggregation Theory):
    We lump together a handful of modern tech companies who are characterized by this kind of low marginal cost markets in which aggregation wins (i.e. rapidly and rapaciously scaling thanks to a massive reduction in barriers-to-entry and frictions-to-scale);
  2. Aggregators (as in the Moat Map):
    Separately, we delineate between Aggregators and Platforms as the dualism for a product’s orientation to its supply chain (i.e. either an Aggregator with proprietary/internalized network effects or a Platform with 3rd party/externalized network effects per the Moat Map)

In regards to #1, we all intuitively think these companies have a lot in common and Aggregation Theory has become the catch-all to explain that commonality with increasing precision. That precision in the definition of Aggregation Theory is where it’s gotten mucky — so fraught with unacknowledged/unexplained exceptions like those mentioned above. It’s just too prescriptive when you have, for example, a Level 1 Aggregator (with expensive/scarce/differentiated supply) and a Super Aggregator (with free/abundant/substitutable supply), which coincidentally starts to characterize the difference between Platforms and Aggregators in #2. Those are just two radically different businesses — far too different to be conflated by the same mental model.

What I’m saying is that Aggregation Theory’s definition, as it stands today, is one step too far — too rigid for such a loose confederation of businesses and too clever by half. Instead, my stance is that Aggregation Theory can simply explain the rules of engagement (“plane of competition”) every time a market’s marginal costs are materially lowered — ratcheted-down along with friction, barriers-to-entry, etc. Then, combining those insights with my refined Moat Map would provide both new entrants and incumbents alike with a powerful framework for navigating competition/disruption.


Implications

Ben Thompson, Aggregation Theory, and the Moat Map are really groundbreaking treasures that will (continue to) be discussed for decades in academic, business, and social settings. Far from demolishing any of these institutions, the refinements herein are mere iterations designed to reinforce their preexisting edifice.

For example, were we to apply the refined versions of the doctrines proposed herein, the one-two punch of Aggregation Theory 2.0 and Moat Map 2.0 would have been far more prescriptive for Walmart, who needed to mount a counteroffensive against Amazon’s disruptive threat two decades ago, then needed to fall-back into a more defensive scheme after Amazon had already scaled — as discussed at the top. Unfortunately, Walmart lacked a playbook back in the naughts; then the roadmap provided by the legacy Aggregation Theory this decade would have led it on a fool’s errand.

As discussed above, this has pretty major implications for strategic planning — not just for a startup establishing its moat (e.g. Etsy’s cautionary tale), but also for an incumbent figuring-out expansion (e.g. Shopify’s faceless man tack).

This has additional implications for competition and regulation. For instance, to evaluate the health of competition in a given market, regulators may want to drop market constituents into two separate buckets — one for each of the Moat’s endpoints. In antitrust matters, it’s hard to imagine problematic monopoly powers being manifest outside of that Moat — given the Misjudgment of Solomon’s dictum and the Janus Rule’s single path-of-least-resistance for each given supply chain (differentiated vs commoditized). At the same time, the market leaders at each end of the Moat cannot use rival(s) who occupy other quadrants as evidence of competition.

Finally, any time an innovation materially lowers the marginal cost structure of a business, regulators should now know what to look-out for with regards to imperialist strategies designed to consolidate problematic market power.


Applying the Janus Rule and the Misjudgment of Solomon

Annotote hails from a pole in its own market’s moat. You don’t want to waste knowledge worth keeping, so now there’s a network that gives you highlights of everything you need to read and lets you annotate anything you want to save or share for yourself. One person’s annotation is another person’s summarization. Get straight to the point without wasting anyone’s time or attention, with Annotote, your antidote to the information overload:

All signal. No noise. Annotote. #leaveyourmark

Adventures in Consumer Technology

No IT Dept: You're On Your Own

Anthony Bardaro

Written by

“Perfection is achieved not when there is nothing more to add, but when there is nothing left to take away...” 👉 http://annotote.launchrock.com

Adventures in Consumer Technology

No IT Dept: You're On Your Own

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade