Two quick notes:
1) November's newsletter was reprinted in edited form in the Boston Sunday Globe: https://www.bostonglobe.com/2021/12/11/opinion/metaverses-good/
2) In these stressful times, I hope the new year brings good news, both momentous and pedestrian, to each of you and your families.
Somewhere around 2005, we moved from the age of the website to the age of the platform. What Marshall Van Alstyne at Boston University (among others) calls demand-side economies of scale began to kick in for Internet businesses. The more people who join Facebook, or shop at Amazon, or watch YouTube videos, the stronger the incentive for a) other people to access the platform and b) advertisers to spend to get access to the growing, engaged population. For Uber and Airbnb, the flywheel effect is even stronger as the more riders/renters, the more drivers/property-owners, which in turn draws more riders/renters, and so on.
After 15 years of being both lightly regulated and extremely profitable, Internet platforms are being reined in through multiple regulatory efforts across the world. Facebook is running print advertising asking for regulation, and the EU has established privacy standards stronger than most anywhere else, at least on paper and for the moment. The evolving notion of “platform” from the days of Windows, through the browser wars of the 1990s, and into “the web as platform” thinking in Tim O’Reilly’s Web 2.0 framework has been heavily studied by business scholars including Annabelle Gawer (now at University of Surry) and her MIT doctoral advisor Michael Cusumano along with Van Alstyne and his co-authors. Another angle comes from communications and cultural studies, where Tarleton Gillespie of Microsoft Research and Jean-Christophe Plantin of the London School of Economics have published strong work.
My focus today is on the physical world, where platform thinking takes on a slightly different shape. Here, people including Erich Joachimsthaler of the Vivaldi consultancy have differentiated between products and platforms, with the latter providing tangible benefits to both investors and customers through network effects, lock-in, and brand differentiation. At a workshop where we both presented, Joachimsthaler outlined John Deere’s now-controversial strategy of infusing heavy equipment with software and sensors. So-called precision agriculture connects tractor-mounted GPS with soil tests and other diagnostics to plant, spray, water, and otherwise treat crops more specifically than at the field level.
For a time, Deere tried arguing that farmers did not own the $500,000 implement they bought — until they purchased annual software license renewals. The right-to-repair movement arose among farmers who wanted to be able to keep their equipment running during critical windows of planting and harvest. To do so, some downloaded 3rd-party software, illegally, to hack their tractors, combines, and other implements. Interestingly, Apple, AT&T, and other tech companies showed up in Midwestern states to argue against the farmers, fearing a precedent that would spread right-to-repair to electronic devices. Late this year, perhaps to pre-empt regulatory pressure, Apple now allows customers to buy repair parts and try their luck at fixing broken iPhones (https://www.apple.com/newsroom/2021/11/apple-announces-self-service-repair/). For their part, Deere responded to President Biden’s executive order mandating right to repair by stating they supported most instances of the practice.
Platforms occur elsewhere in the physical world. Automakers are aggressively pursuing a practice called “communization,” often using platform nomenclature. The objective is to share parts and assemblies across vehicles that might look different in the showroom but enjoy economies from simpler supply chains, easier new-product development, and faster time to market. Starting in about 2005, GM’s Lambda platform (itself a variant on the previous-generation Epsilon) was shared by 3-row SUVs from Saturn, Buick, GMC, and Chevrolet. Beginning in the 2017 model year, Subaru began migrating _all_ of its models to the Subaru Global Platform (SGP). Straight-line stability, NVH (noise/vibration/harshness), and ride comfort are all said to improve. At Honda, the goal is to reduce model variation across its five global vehicles by 66% between 2020 and 2025. In some ways, this movement anticipates an electric future: a typical Tesla’s part count runs on the order of 10,000, with the drivetrain being substantially simpler than a typical internal combustion vehicle, whose part count approximates 30,000. Traditional manufacturers have begun reinventing themselves, and platformization is part of the plan.
On a smaller scale, a few months ago I saw a Twitter post on the topic of tools, but it’s now lost and no academic sources showed up, so I’ll start from scratch by talking about . . . batteries. All of the major power tool companies — DeWalt, Makita, Hitachi, Bosch, Ryobi, Milwaukee, — and many house brands sell cordless power tool systems. The batteries are interchangeable among tools, but not brands. (Interestingly, DeWalt, Craftsman, and Black & Decker are all owned by the same parent company.) Thus, when shopping for a cordless drill, which is typically the “gateway drug” to a given platform, one might reasonably compare torque, battery life, weight, or other objective statistics. As one finds other needs — flashlights, garden tools, many kinds of saws, grease guns, and so on — the original purchase has created path dependencies. The Hitachi drills I may have bought last year on sale work perfectly well, but Hitachi doesn’t sell a cordless grease gun. If I buy a second 18-volt tool from a different vendor (Makita, let’s say), now I have 2 chargers, 2 or 4 batteries, and lower option value than if everything were fully interchangeable.
Each power tool brand has a range of product decisions to make in the platform context:
- What price range? Ryobi sells an 18-volt drill-driver kit for $50 at Home Depot; Hilti offers an 18-volt kit for $359.
- Given a price point, do we build in features, durability, or innovations (such as job site RFID tracking)? How does our brand convey our decision to the market?
- How wide is the application coverage? Carpentry, automotive, marine, lighting, gardening/landscaping only begin the list.
- How many SKUs are too many? Especially as supply chains are being strained, and most tools are made offshore, is there value in focusing on known strong sellers rather than chasing endless numbers of niches? Or does completeness of portfolio confer its own competitive advantage?
- What is our channel? With a large SKU count, physical retailers will be challenged to stock the right number and quantity of available products. At the same time, Stihl (makers of outdoor power equipment, much of which is also moving to battery power) has built a formidable network of small dealers, eschewing the big boxes entirely to move a small product line. Amazon, meanwhile, has launched between 50 and 100 private-label brands. Access to that audience comes with real risks, but as far as I could tell, no major brand has stayed off the site, um, platform. Hilti (from Liechtenstein), Festool (Germany), and Metabo (founded in Germany, now owned by Hitachi parent Koki Holdings now that Hitachi brand tools are owned by the US private equity giant KKR) are all available on the Seattle supersite — alongside CP CHANTPOWER, TODOCOPE, and Eastvolt, the Amazon brands of cordless tools.
- What else can we turn into a platform? Milwaukee launched its Packout modular tool storage line in 2017 and keeps adding both small parts totes and larger static and rolling tool chests to the line. Interestingly, all the pieces I have seen are made in Israel: the engineering tolerances and/or polymer formulations may preclude Chinese sourcing. Black & Decker’s DeWalt brand, meanwhile launched two lines — TStak and ToughSystem — that each demonstrated growing pains. Since 2020, each is on a 2.0 nomenclature, with more robust locking hardware, backward compatibility, and other enhancements, but the two systems cannot interoperate, and the discussion forums are full of disgruntled contractors who gave up on the brand after the poorly-thought-out initial effort. The system approach clearly commands a price premium: the Milwaukee Packout 16-quart modular cooler sells for $109, while a comparable Igloo rotomolded (think Yeti hard side) 20-quart is the same price and 25% larger.
Zooming back out to the bigger platform picture, given that Bosch makes both cordless power tools and batteries for e-bikes, what lessons are being applied to the younger market? Will bike batteries be at all interoperable, or will the power tool model carry over? Is my allegiance to the battery manufacturer (Phylion vs Bosch), the motor company (Yamaha vs Shimano), or the frame-maker (Specialized vs Giant)? Might my Bosch bike battery be able to swap with my cordless jackhammer battery, both of which come from the same company and run at 36 volts? At a larger level, electric cars, electronics, cordless tools, and e-bikes all rely on lithium-ion technology, and supplies of the required minerals — lithium, cobalt, and nickel — are all tight. Which markets have priority today and might tomorrow? Modern military hardware also requires significant quantities of cobalt as well — how will national security be invoked in the pursuit of rare-earth metals?
The decade of the 2020s could well be defined by the extension of platform thinking to residential construction, pharmaceuticals (mRNA models could point in that direction if you squint your eyes), entertainment (what’s the difference between a franchise/“universe” and a platform?), or education (check out Clay Shirky’s Revue newsletter). At the same time, just as the late 19th and mid 20th centuries saw backlash against the excesses of industrialization, so too will we see more scrutiny of platform models’ externalities. In any event, the race between innovation and new constraints, many long overdue (are too many Airbnbs a good thing for a residential neighborhood?), will bear watching.