#216: Innovation Stacks, Gaming, Preventative Tech, Oncology, & Bitcoin

In his new book, Square’s Co-Founder Jim McKelvey explains how companies that tackle unsolved problems can beat the competition and become industry leaders by building Innovation Stacks. He describes Innovation Stacks as chains of products and features that start-ups have to build in order to survive.

Square’s original goal was to arm millions of small merchants with point-of-sale devices to accept credit card payments. Key to its success was a fast, simple online process enabling merchants to sign up. In contrast, banks put businesses through a merchant account application process that takes a week or longer, only to reject most. Using complex underwriting models, Square disintermediated the banks by approving merchant applications in short time, taking merchant risk on its own balance sheet by becoming the Merchant of Record for its small business customers. Because it designed point-of-sale devices to track merchant transactions and collect “big data” in real time, we believe Square’s underwriting models are much more accurate – especially for historically underserved small businesses – than are banks. Square was able to collect large amounts of data and train underwriting models because it gave away the “square” reader for free, as it cost 99.9% less to produce than comparable portable credit card readers.¹

When competitors attempt to copy features that disruptive companies introduce, they may miss the importance of the Innovation Stack upon which the feature is built. A single feature is almost impossible to copy, turning the Innovation Stack into a major competitive advantage. According to McKelvey, companies built on Innovation Stacks tend to navigate crises better than others because they are used to solving problems in the face of significant competitive pressure and considerable uncertainty.

[1] In 2009, the original Square reader cost 97¢ to produce while the cheapest portable credit card reader cost $950, according to Jim McKelvey.

Having launched videos and music, Amazon is continuing its expansion into media with the upcoming release of two original games: Crucible and New World. At $160 billion, the gaming market appears much larger than the music and movie markets, and is aligned strategically with Amazon’s Twitch game streaming and Amazon Web Services (AWS).

Twitch is the most popular live streaming service outside of China. Hours viewed on Twitch outweigh those of its competitors by almost 3 to 1. After flattening out in 2019, its viewership grew by 15% sequentially, or roughly 60% at an annual rate, in the first quarter as the coronavirus outbreak forced more people to stay at home.

At this time, Amazon’s Twitch is a niche business perhaps because, outside of chatting, streamers and viewers do not interact very much. With original games, Amazon will be able to create in-game tipping and other opportunities for Twitch to offer gamers entirely new experiences and economics. If it succeeds, Amazon will demonstrate the power of vertical integration enabling platform owners to innovate on product design and build even higher barriers to entry.

In 2015, researchers estimated that the economic burden of seasonal influenza in the US was $11.2 billion. Interestingly, paid sick leave (PSL), earnings lost from the death of productive workers, and other indirect costs contributed more to the economic burden than did the direct medical costs associated with influenza.

ARK estimates that if employees take 50% of their PSL when sick with a transmissible disease, the economic burden of influenza on employers could total roughly $40 billion annually. Even the $40 billion could be a significant underestimate because many businesses do not offer PSL, giving employees no choice but to go to work when they are sick.

ARK estimates that if employers were to spend approximately $300 a year per employee, preventative technology could reduce the economic burden of influenza on businesses significantly. Potentially, biometric devices like the Oura Ring that can detect fever and other symptoms could alert employees to stay home and save employers significant costs. 

Cancer can be treated more effectively when oncologists diagnose it early. As a result, the United States Preventive Services Task Force (USPSTF) recommends population-wide screening for several cancers including breast, colorectal, and cervical. Nevertheless, most diagnoses do not take place until the onset of late-stage symptoms because adherence to the guidelines has been mixed. Many individuals do not comply because they are afraid of biopsies and other invasive tests associated with the diagnostic process. In ARK’s view, a non-invasive test with a low false-positive rate should be able to detect cancer earlier and improve patient outcomes dramatically.

Liquid biopsies that detect cancerous DNA fragments in blood are gaining traction as evidence of their clinical utility mounts. GRAIL recently published data validating its liquid biopsy as a screen for fifty of the most common cancers. The test uses machine learning to detect a cancer’s tissue of origin (TOO) for patients who test positive. Based on data from more than 1,300 patients in its validation cohort, GRAIL’s liquid biopsy boasted a game-changing false positive rate of 0.7%, making it potentially eligible for population scale use. Its sensitivity, or false negative rate, was 56.1%. A negative result does not rule cancer out but a positive result likely rules cancer in. For patients with positive tests, the TOO was correct 93% of the time, even in Stage 1. Indeed, GRAIL’s ability to detect early stage pancreatic cancer, a highly lethal cancer diagnosed typically after it has metastasized, suggests the possibility of profound clinical utility.

Based on an ongoing study of roughly 8,000 patients, GRAIL has suggested that its test could have a positive predictive value (PPV) of 51% in a large, asymptomatic population, roughly 13X better than the current USPSTF screens for breast, colorectal, and cervical cancers.

In a recent piece, data analyst Matt Ahlborg outlines how a country wide electric outage in Venezuela has uncovered the nature and scale of Bitcoin usage across Latin America. On March 7th, Venezuela experienced one of the largest electrical blackouts in the nation’s history. According to Venezuelan outlets, the electricity outage affected 22 out of 23 total states. Naturally, Bitcoin transaction volumes on Venezuela’s largest bitcoin trading platform fell by more than 70%.

Perhaps more surprising is the impact the power outage had on bitcoin trading volumes in neighboring countries. Countries including Peru, Colombia, Mexico, Chile, and Argentina experienced a strikingly similar dip in volumes. The strong correlation in volumes across Latin American countries indicates that bitcoin is being used as an intermediary currency between fiat currencies from people transferring value to and from Venezuela.

Because the Venezuelan government imposes controls on its currency’s ability to freely exchange with other currencies and has sanctioned money transmission businesses, remitters wishing to transfer value to and from Venezuela are resorting to informal money transmitters with Bitcoin as a crucial intermediary. Matt Ahlborg’s full analysis can be found here.