Early Indications February 2021: Foundations and Scaffolding

First of all this month, I ask a favor. Last week, the website for a professional doctorate in information studies at Syracuse University went live. I direct that program, and we are looking for 10-12 mid-career leaders in information industries to join us in writing a thesis in information issues, broadly defined: enterprise architectures, misinformation/disinformation, the future of work, privacy/security, and data and analytics are all in play. The program teaches no content courses from the existing catalog: the members of the cohort will be sufficiently diverse and accomplished that finding common ground would be impossible. Rather, we focus on the writing of a 5-chapter thesis, with each chapter a common semester deliverable in semesters 4-8. Methods courses and topic selection precede those semesters, and semester 9 is devoted to thesis defenses. The course is mostly synchronous on-line, with 12 of 51 credits being earned in 2 1-week residencies per year. These residencies will offer topical seminars of broad interest: those of you who attended my Center for Digital Transformation meetings at Penn State will have had a taste of these. Cost is on par with an executive MBA or many DBAs, about $100k; no financial aid is available through the program.

The ask: because we had to wait nearly a year for state of New York approval of the program design, we are on a very short runway to fill this year’s class that will begin with a May 24 residency. Applications are due March 15 for preferred consideration, but we will look at later candidates if the class doesn’t fill. We have 5 applications already submitted, from fascinating people of substantial accomplishment: the cohort will be 3 years of the best conversation you’ve ever had, I predict. 40 more people have already registered for information sessions this week: Wed March 3 for the general population and Thursday the 4th for military and veterans. Here are the links: please forward them to anyone you know who might be interested, or let them know to email me directly to set up a call. Thank you for the assist.

**********
Let me start by enumerating a few themes, which I’ll then try to tie together.

As I watched the World Wide Web emerge early in my professional career, I shared the hopeful positivism (or, more accurately, positivist hopefulness) of Tim Berners-Lee and his intellectual kin. Having recently finished a doctoral dissertation in which I learned to construct Boolean queries inside expensive CD-ROM data sets, I was ecstatic to find in Google plus the Web the biggest research library ever conceived, much less built. Some readers may know that in ancient libraries, including Alexandria, books were chained to the shelves: print knowledge was that valuable. Now (circa 1996), I didn’t know if information actually wanted to be free, but a whole lot suddenly became so. From access to such wealth, it was a short hop to the belief that people could operate under less uncertainty and make better, fact-based choices and decisions.

People wiser than I knew better. As I discovered while researching my new book on YouTube and TikTok (coming next year from MIT Press, it looks like!), James Katz at Rutgers saw in 1998 the cost of removing gatekeepers to content dissemination. If everybody could publish an opinion, the Yeatsean center could not hold:

The Internet and the Web allow for the quick dissemination of information, both false and true; unlike newspapers and other media outlets, there are often no quality control  mechanisms on Web sites that would permit users to know what information is generally recognized fact and what is spurious

Years later, when DARPA ran the geospatial intel challenge of having ad hoc teams coordinate via social media to find 10 red weather balloons, I failed to grasp the importance of the counterintelligence efforts that slowed the winning team (out of MIT’s Media Lab) by spoofing IP addresses and GPS coordinates. Rewatching the 2010 video a couple weeks ago was another realization that the information universe does not only, or probably primarily, operate under logical assumptions. Rather, PT Barnum, Joseph Goebbels, and George Orwell seem to set much of the tone.

A second theme relates to the contention (by Scott Galloway among others) that COVID-19 was an accelerant more than a disruptor. The pandemic, and our responses to it, made socio-technical developments happen far faster than predicted: telemedicine visits, remote work, grocery/meal delivery, and Zoom schooling are all a permanent part of the cultural landscape only months after the initial lockdown. All of these new practices stress the existing infrastructure, whether it’s laptops for 5-year-olds, rural broadband, a sustainable economic model for car-share drivers, privacy practices for connected video cameras, or the simple but often impossible task of checking up with vulnerable neighbors and family members. This need to invest in 21st-century infrastructure — cultural, economic, and physical alike — was highlighted by The Economist in a recent article on mass transit. If buses and subways are not perceived to be safe from pandemic spread, people will turn to cars, with dire consequences for densely populated urban areas. New York, Tokyo, Hong Kong, and Mexico City cannot absorb all the migration headed their way with a car-based transport model. Gridlock, of every type, is not sustainable. At the same time, infrastructure is slow to rebuild, tough to expand, and expensive to maintain. Tax revenues are down everywhere, making the barriers to mass transit investment that much higher. And subways are far from the only economic priority.

Finally, it’s depressing to look at what the US tech sector has become. I’ve touched on this before, so won’t belabor the point. With the emergence of ad-revenue-powered software development, it’s hard to see tech innovation on the scale of the web browser, the search engine, the original online retail model, or the smartphone. Silicon Valley + Seattle have turned a lot of energy to scaling the aforementioned innovations, and many important developments have emerged, to be sure: mega-scale cloud computing, computational photography, and chipset design each exhibit true breakthroughs. For all of these upbeat notes, we must confront the failures of privacy protection, behavioral manipulation, energy consumption (with Bitcoin’s massive inefficiency exhibit A), and the industry’s contributions to economic inequality. Many business continue to run on Excel and email, again, tools that do not scale.

These three themes — access to knowledge fueled misinformation rather than enlightenment, the need to rejuvenate infrastructure, and Silicon Valley’s turn toward meal delivery and social media rather that worthier challenges — were floating around in my thinking as I read a short blog post by the tech journalist Om Malik (with a hat tip to Jan Chipchase for the pointer).

Malik posits that a precondition for getting anything done in a group is sharing an understanding of reality. He says it better than I could, and quotes the likes of psychologist Jerome Bruner:

"Our culturally adapted way of life depends upon shared meanings and shared concepts and depends as well upon shared modes of discourse for negotiating differences in meaning and interpretation,” the late psychologist Jerome Bruner wrote in The Acts of Meaning. “By following a set of rules governing interpersonal communication, people inadvertently modify their private, idiosyncratic conception of a state of affairs and reach a common understanding of that situation. As noted, these shared representations constitute the contents of a culture.” 

Given that Facebook is actively trying to build as many systems of meaning as there are ad market segments, and Google is imposing its own clustering algorithms on our ads, our emails, and our YouTube viewing, what counts as foundational bedrock in US (or any other) culture? Concepts (that carry critically important commitments) as essential as voting, epidemiology, and tolerance are no longer assumed. Millions of people are convinced that the US election outcome was fake, that the coronavirus is fake and/or a foreign plot, and that white supremacy is the core tenet of major political parties. Taken together, these beliefs undermine the foundations of western democracy. When people of different colors are successfully demonized as cover to extreme rent-seeking by white wealthy one percenters, and the blockage of economic mobility by the latter is blamed on the former, the consequences are truly life altering. As Anne Case and Angus Deaton make the case in their book Deaths of Despair, life expectancy, economic mobility, and personal wellbeing are all casualties of the past 40 years of health care, tax, and environmental legislation being written by industry lobbyists, many of them former legislators or aides, for the benefit of the already-wealthy.

A foundation holds a building up from the bottom. Good ones last centuries or millennia. Scaffolding is far more visible, but it is temporary and not typically structural. The belief that all people are created equal and are endowed with inalienable rights to life, liberty, and the pursuit of happiness is foundational and served the United States well for 200+ years. The frequent fights of outsiders to be let in to that promise attest both to the power of the promise and to the benefits to some groups of denying it to others. Now, as Malik laments, there appear to be fewer and fewer shared realities. Reality TV, the utter antithesis of its name, made the Kardashian family a multi-billion-dollar enterprise. Fake news has done the same for Rupert Murdoch and Mark Zuckerberg. Meanwhile, political parties of all persuasions are torn by internal fights over the shared reality in which players on the same team operate. Religions from Anglicans to Catholics are seeing disillusioned and emboldened members either defect or stay and sow dissent within the ranks: there are currently 19 different Baptist subdivisions, according to Baylor University’s research center devoted to the denomination. Epidemiologists and public health official acknowledge that they have done a poor job navigating nuance and clarity at a time when faith in science has been tested by both external political headwinds and internal narrowing of disciplinary foci.

The core question is both simple and troubling: without something common to believe in, groups of people splinter. Religion, science, government, and economics are all torn by divisions over both what is real and what matters. Sport can bring groups together, often temporarily, and even here, ESPN had to shut down comments on the website because flame wars were instant and usually vicious. The moderator of a historical outreach website -- not what one would expect to be a snakepit -- had to quit as the site's comments became a “cesspool.” 

The tools that Berners-Lee and others invented have spread both knowledge and, maybe primarily if we are honest, divisiveness. One possibility is that people have been this divided for decades: white supremicists are as old as this nation. Now that they and everyone else have a public voice, maybe we are seeing a long-term lack of commonality that newspapers, TV stations, and book publishers covered over with their gatekeeping function. Alternatively, the collapse of Enlightenment epistemology took far longer than the history of the Internet, and it’s possible social media hastened the rise of pluralistic voices that could appeal to personal prejudice rather than shared norms of moral clarity, evidence-based argument, or peer review (none of which were as robust as its adherents claimed ).

In any event, as the world enters post-COVID-19 reality, which voices will gain the most adherents? Which institutions -- the press, government, academia, civil or spiritual religions -- can be either born or reborn as adequate to the moment? Which core beliefs can be reinvigorated — the US Declaration of Independence is acknowledged in foundation documents of roughly half the 192 countries at the United Nations — and which ones need to be invented in a post-newspaper, post-industrial, post-fact world? What ephemeral trends, scaffolding if you will, can draw adherents temporarily yet visibly and effectively? (The US Democratic Party struggled to counter the lies and insults Donald Trump consistent employed in 2016 and again in 2020: he successfully rewrote the rules of engagement, with help from the aforementioned Fox News and Facebook.) Without a foundation, preferably one in empirical reality, humanity is going to endure social chaos. My hunch is that we are seeing a race to master one of these new media channels with a compelling narrative: form and content will synch better than they do now, with substantial rewards for the entities that get there first.

Early Indications January 2021: Re-envisioning Shopping

According to a recent cover story in The Economist, 3 trends in Chinese e-commerce suggest a variety of paths that U.S. companies might follow. It’s a fun exercise to use these developments as a prism through which to rethink which people buy things online, how different demographics buy different things, and how a variety of adjacencies might be explored in various geographic, technical, and economic terrains.

However it is configured, Chinese e-commerce dwarfs any global online market. In addition to being big, it’s different in that Chinese commerce is much more likely to be conducted via a mobile device: 90% vs 43% stateside, for example. Finally, and most interesting for our purposes, Chinese e-commerce rarely uses storefront metaphors borrowed from physical retail. Social commerce (think neo-QVC live-streaming meets TikTok meets Facebook), serious omnichannel, and mobility affordances including gamification, geofencing, and precision coordination (“have the pizza arrive 10 minutes after I get home”) all can contribute to new ways to socialize, restock, entertain ourselves, and discover new hobbies, art, and cuisines.

There’s a lot going on there, so let’s back up and parse a few things. Buying stuff already performs a variety of functions. We need to procure vitamin water, salsa, ramen, and similar staples with some regularity; there’s not a lot of entertainment to be had here, except maybe when a group of roommates hosts a dinner party or something similar. Other purchases — engagement rings, headstones, kitchen renovations — are both more substantial and less frequent. Buying some things — cars for example — can be so onerous and frustrating that entirely new models such as Uber’s and Tesla’s can gain traction in part because people hated the old practices so much. In the US particularly, there exists a massive amount of physical retail space, much of it built on the idea that people went to stores (read malls) to engage in “retail therapy,” socialize, exercise, or just pass the time. Finally, buying things related to one’s hobby — fabric, model airplanes, kitchen tools, books — is an entirely different kind of retail, potentially as impulse-driven as a Netflix rental or as researched and rationalized as buying a car or dream vacation. All of these factors mean that retail can vary tremendously by locale: the US, for example, has swaths of extremely low population density, the world’s largest (measured on a per capita basis) physical retail footprint, a state-run logistics provider that performs abysmally, countless tax jurisdictions to navigate, and historically high income inequality. All of these matter for what shopping might look like.

When we say “online shopping” or “e-commerce,” then, it’s important to distinguish what flavor of shopping we mean: resupply, fashion, whimsy, gifting, information goods, and domestic infrastructure (think home appliances) each are conducted through different channels, at different life stages, in different human configurations, with varying amounts of forethought, entertainment value, and financial apparatus. That said, the Chinese megatrends (social commerce especially) present U.S. retail with intriguing starting points rather than blueprints. What might this kind of reconsideration of shopping lead to outside the particularities of the Chinese context?

1) Shopping can be social

Whether it’s picking out school outfits, outdoor gear, or food for the big picnic, shopping in packs can be fun. As those packs migrate from the mall to the smartphone, it feels like there are many potential models yet to be explored: Facebook is neither the first nor the last word here. Whether it’s impromptu alumni reunions among hall mates or military families keeping in touch across the world, how might we assemble ad hoc groups around some variety of commerce experience? Zoom-powered watch parties are already a way to share a binge or movie across distance — what if Shopify (which grew faster than Amazon last year) built an app inside Zoom to facilitate “shop parties,” whether at the sporting goods store, the boutique, or the appliance store as new homeowners attempt to navigate the refrigerators. Who’d “own” that app? How “sticky” would it be across shopping experiences (CVS to Lowe’s? Ulta to Gap?), demographics, or geographies?

2) Showing beats telling

In some ways, online video is catching up to the decades-old home shopping TV model, the in-store makeup counter, and the high school shop class. Before the Internet, people still had to figure out which cookware they liked, how to apply mascara, or how to drill a pocket hole. Teaching and demonstrating are a huge business — more people in the US watch celebrity chefs than cook gourmet meals. Both live-streaming and online video repositories are in their early stages, I believe, of being utilized within much more seamless and comprehensive commerce experiences. Again, what will be the principles of speciation? Will clustering happen around the demographic, the product category, the spending at stake (will a table saw video resemble a lipstick tutorial?), or the mega platform hosting the experience? Will Facebook spread itself too thin trying to be too many things to too many people? Speaking of Facebook, where might VR/AR fit into this model?

3) Shopping games

11 years on, game designer and professor Jesse Schell’s conference talk about gamification of real life remains compelling viewing. He noted how the physical and the virtual were coalescing, way before Pokemon Go; how point systems, dating back to Weight Watchers and supermarket loyalty programs (Green Stamps), continued to proliferate; and how the next wave of innovation would be less technical and more behavioral. He saw early on the power of the Facebook dopamine pump, as evidenced in Farmville (RIP), and helped inspire a wave of Internet-connected everyday items with his toothbrush example. Where might we head next? What mashups of YouTube, Twitter, Fortnite, Untappd, and Walmart will emerge? Among which demographics? In what product categories? What role might payment platforms like Visa or Venmo play? What about logistics providers like FedEx or UberEats?

4) What is the buying unit?

For a time in the 1990s, B2B demand aggregators were the hottest startup category. Labs that needed a specific reagent could pool their demand, the thinking went, getting bulk pricing for small-quantity buyers. Ariba, Chemdex, and many (!) others sought to become a new kind of market intermediary. Later, the same thinking came to retail: if I need a printer, the lower coordination costs afforded by the Internet mean that I and 249 strangers also needing printers should be able to present HP or Canon with a bulk order just for us. Sites such as MobOffer sought to operationalize this idea. We’re used to supply aggregators: it’s what retail is premised on. But what if TikTok, AAA Motor club (which already does this for cruise line bookings), the Western Carolinas Organic Cheesemakers Association, or other groups could mobilize buying power for more goods and services that matter to their members? Then what happens if the buying groups are algorithmically constituted, much as Waze forms ad hoc anonymous social networks of proximate drivers?

5) Where is the application logic?

Not to sound too technical, but this is an important point. In the US, online commerce began with websites. The physical retail experience was very much front of mind; competitors fought over the shopping cart metaphor in court. Fast forward 25 years, and expand to ~150 countries, and 1995 US brick-and-mortar retail is no longer a universal frame of reference.

After about a decade of desktop web commerce, largely in the west, Apple built on Japan's early DoCoMo learnings and introduced the mobile app as a new programming model. Much of the website experience carried over, but there were glimpses of new retail experiences here and there (in-app purchases are one example). Now, with most of the planet connected via mobile devices and with billions of people who don’t remember driving to Walmart or Sears as their mental anchor, Chinese and other online commerce sites are putting commerce into messaging. 

This model makes sense: texting somebody at the store to remember the eggs, or asking someone at home what kind of beer Emily likes, has become habitual. This extreme drop in coordination costs (making grocery lists isn’t nearly as important as it was 20 years ago) suggests commerce could logically be proximate to messaging. In addition, group-texting provides another use case for embedding commerce. If a half-dozen people are convening on the impromptu picnic, figuring out who’s bringing hotdog buns could easily trigger a buying moment. In a permission-based scenario, if I let Krogers monitor certain of my text exchanges, the store could drop coupons, inventory availability, or price comparisons into my queue. If the word “seltzer” is used, and I’ve given opt-in approval, a prompt — “what flavor?”, “how many?”, or “brand preference?” — could get me to click-and-collect transaction with only a few voice commands, no shopping carts (metaphorical or physical) necessary. 

Speaking of physical vs virtual, embedded QR codes serve as hyperlinks between the messaging and brick-and-mortar domains. I can discover something on my phone and walk into a physical retailer to find the exact item and possibly get some form of discount or loyalty points. Here is another potential application of the gamification theme noted above, and another highly convenient omnichannel implementation.

******

Some of these scenarios are already playing out in North America; others are still years off, if they ever take hold. Who’s poised to win, and who might need to scramble?

-Facebook has the reach across platforms, which they are more tightly integrating as we speak, and the behavioral know-how to learn how to trigger the desired consumer actions. Talk of breaking the platforms apart could slow Zuck down, at least initially.

-Google has failed to build social connections, dating back to Orkut and as recently as (I’m betting) Meet. In addition, search crawls don’t work across apps, but this is less of an issue when Android devices provide Google with lots of user data. If the messaging layer turns out to be a US variant of WeChat, hypothetically, that invisibility to web crawls could be a major impediment.

-Amazon hasn’t missed many steps over its 25 years, and I don’t see the Seattle Godzilla getting outflanked by a startup, by Walmart, by Alibaba, or by Facebook on social/mobile commerce. Amazon already leads the way in many aspects of omnichannel, for instance. Furthermore, as Scott Galloway notes, Amazon is unique in its ability to turn “core competency” logic on its head: rather than sticking to what differentiates it and outsourcing the hard back-office stuff, Amazon takes those hard back-office tasks — order management, data centers, and logistics so far — and gets so good at them that it can sell those services at a healthy profit. Finally, Amazon is also good at driving wedges into non-adjacent markets: Audible, Kiva, Twitch, Annapurna, and Blink were brilliant acquisitions. Now watch what Amazon will do with PillPack and Zoox.

Given the changes in London, Brussels, and Washington, along with the change in trajectory of the coronavirus, it’s hard to see how cross-border cultural transmission will fare in the coming year. Whatever happens to TikTok post-Trump, the future of Chinese tech companies in the west is uncertain for many reasons, air travel only one of them. As much as I think WeChat, AliPay, and Pinduoduo have much to teach us about what shopping is and can become, my gut says it will be firms that translate the Chinese (and Indonesian, and Japanese, and Korean) practices — rather than export them — that will gain traction in the US; Europe is sufficiently different that I won’t make any claims for what happens across the Atlantic. 

Early Indications December 2020: The bounce-back myth

Maybe it was all the Zoom toasts over the holiday pledging a return to in-person celebrations next year, but I’ve heard too much talk about the timetable for a return to “normal.” That is, there seems to be an expectation that after enough immunizations, life can somehow resume the rhythms and patterns that defined life in 2019. I don’t share this expectation: we have seen consequential political transitions in the EU, UK, and US, for starters. There has been a global financial recession. 2020 saw extreme climate events, including wildfires and hurricanes, that presage more such disasters. Nobody knows how many more businesses will close, permanently, in 2021. For one example, many people assembled Airbnb empires of multiple rental properties that now present impossible mortgage payments. It’s one thing to prevent evictions in the pandemic, but I doubt courts will stop foreclosures of such speculative properties whose impact on residential neighborhoods is not wholly positive. That’s just an initial accounting of the impact.

Here are a few questions identifying sectors, markets, and institutions that will emerge from 2020-21 forever changed.

1) What is the new media mix?

The breakout of TikTok into mass consciousness has certainly been noteworthy over the past year, but it’s not the only important story. The game platform Roblox has been around for about a decade, but its hold on the under-10 demographic is stronger now than ever. TikTok is a key player in 12-24, and podcasts consumed voraciously by middle-agers are seeing huge investment by Amazon, Spotify, and others. YouTube has gained traction among every segment from 5 to 95. Two questions: who wins 21-34, especially if Facebook must divest Instagram? Also, what happens to legacy media: radio is in trouble for sure, and it’s not clear that network television can adapt to the streaming model. ABC, given both its Disney parent and its ESPN sister network, should be in better shape than CBS, for example, but how will we know?

2) What is the future of “date-night” entertainment?

Music venues are getting some bailout money, iirc, but I’m guessing Live Nation and its kin will see money sooner than small-footprint venues. Musicians are but one class of artists hit hard by the Covid recession, and many are doing clever podcasts, instructional videos, and the like to keep the electric bill paid, but this industry will bear watching. Similarly, what is the future of movie theaters? What, in particular, is the future of movie theaters in shopping malls, which have their own existential changes to confront? Going to the mall for a meal, some strolling, and a movie will not be as appealing in 2022 as it was in 2002, for lots of reasons. Restaurants, meanwhile, are closing by the thousands: how will new entrepreneurs address the mix of meal delivery, carry-out, catering, and interior table service? Will new ethnic cuisines fill some of the empty storefronts? Will the meal-delivery apps’ attractiveness persist post-Covid? A larger question informs many economic ones: how much will our cocooning get thrown off in a collective binge of sociality, and how much will staying in (from dining, from work, from shopping) define the new normal?

3) What’s ahead for the tech sector?

Will there be a new category of device or application that rearranges the landscape as the PC, Internet browser, search engine, or smartphone did? Will so many revenue streams continue to be ad-based? How will computer/app/smartphone companies address the mobility market (see #6 below)? How many Internet platforms will survive? Is Pinterest, for example, a feature or a product? Will health tech be a major new consumer market? Will Silicon Valley maintain its primacy? HP, Oracle, and Palantir are leaving, Google is distributing its operations, and California’s many drawbacks — regulatory, climatic, and cost-of-living — provide other areas with ample opportunity to demonstrate their attractiveness, particularly in a work from home world.

4) How will governments address budget shortfalls?

Covid-related expenses coupled with drops in tax receipts are presenting many cities and states with a grim scenario for the next few years. Education, social safety nets, policing, and other functions will shrink: many states have balanced budget provisions that preclude any form of borrowing. States with large tourism sectors (Nevada and Hawaii for starters) are especially hard hit. How will private philanthropy co-evolve with shrinking state and local programs? How much seed corn will be destroyed? That is, how many qualified job applicants will lack key skills in 2 or 10 years? How many foregone investments, in whatever form, will deprive future years of returns on industrial attractiveness, after-school programs, or mass transit maintenance? How many “deaths of despair” and “diseases of despair” will plague our towns and states going forward? 

5) What happens to education?

Apart from the budget crunch noted above, schools will cope with Covid-related fallout for years. How many teachers will get sick and/or burn out after this trying year? How will pedagogy evolve to include remote learning as a routine component? At the college and university level, schools with large endowments have enjoyed robust investment returns; all institutions with seen exceptional (one-time) costs soar, particularly for investments with no long-term payoff. Plexiglas shields come to mind, but so do (in the instance of Syracuse University) hundreds of thousands of saliva tests. Academic hiring freezes are prudent, given the uncertainties involved, but are creating logjams in PhD programs in which advanced students are staying on rather than graduate, see student loans come due, and face bleak job prospects. Maybe the biggest question: how many of the critically valuable (for both their skills and their tuition dollars) international students a) will be allowed and b) will choose to attend college or graduate school in the US vs Australia, Canada, France, Germany, or elsewhere?

6) How will people get around?

Electric cars are obviously big news, but bicycles are even more interesting, especially at the global level. Paris is changing its infrastructure to be more bike-friendly, driven by 2020's modal shift from subways (where the fear of Covid dropped ridership) to cycling at a mass level. Like bikes, standing desks are enjoying a work from home sales surge, but cycling is far healthier still. As the electric conversion proceeds, charging will continue to be an issue, for urban street-parkers in particular. As work from home changes the frequency and nature of commuting, what happens to urban office space and the associated parking infrastructure? 

Many eyebrows were raised by the recent news story that Apple’s Project Titan is still targeting a consumer-facing automotive product for 2024 or so. More interesting, to me, is Amazon, which has made no announcement. Consider:

-Amazon is learning a lot about drones, including path planning, machine vision, and vehicular control: three key competences of autonomous driving.

-Amazon employs some of the best machine learning talent on the planet.

-Amazon is building a fleet of more than 100,000 delivery vehicles (for comparison, the USPS fleet numbers about 200,000). Using autonomy internally or in a B2B scenario could be a smarter play than targeting individuals: Google Glass would have fared much better had it targeted industrial and commercial use cases rather than consumer behavior, a fact not lost on Jeff Bezos I’m sure.

-Amazon has effectively infinite computing capacity (and vast programming capacity) on tap.

That’s four reasons, constituting a unique footprint (UPS lacks AI server farms; Google lacks a massive fleet of vehicles), why Amazon could shake up mobility. Oh, and they just bought these guys.

That’s a cursory list: work, dining, mobility, education, entertainment, and media are all in play. I’m similarly interested in travel (who’s buying airline or ski resort stock right now?), supply chains/manufacturing (can Boeing reclaim its mojo?), and most important, health care. Ohio State just found heart damage in 15% of athletes who had tested Covid-positive (a scary-big percentage), but there’s no diagnostic as of yet to inspect the virus’s long-term impact on the brain regardless of whether senses of smell/taste rebound.

Yes Covid-19 has a relatively low death rate, so far, but in the nightmare scenario, millions of people worldwide could be cardio-pulmonary-neurological time bombs. It’s not improbable that we’ll see lung scarring, heart malfunction, and/or cognitive/emotional damage on a massive scale, threatening long-term debilitation (possibly including accelerated onset of Alzheimer’s) of huge numbers of people, some of whom were silently infected in 2020. (I’m not free-form speculating here: see here and here for starters.) That sure doesn’t sound like “the good old days” to me.

Early Indications October 2020: Why this time will be different

Nearly 30 years ago, businesses across the world began a surge of investment in information technologies. Spurred by desktop computers that were fast enough to run graphical user interfaces, by the adoption of the Internet starting with email, and by the combination of management consultants and enterprise software vendors, companies began “reengineering the corporation,” as one best-selling book termed it. Chief Information Officers were named and, with great rapidity, often replaced. SAP, Siebel, Oracle, Sun Microsystems, and Microsoft all minted millionaires out of programmers and architects, while the big systems integrators — IBM, Accenture, and many others — hired, trained, and churned through thousands of young hires to staff the many and massive projects. As for results, opinions varied in both academic and managerial circles, but every CEO, no matter what industry, had to confront the decision as to how (not whether) to invest in IT.

There are reasons to wonder whether the time has come for a similar wave. Machine learning has dropped in price and proven its capabilities to help address new kinds of problems. Cloud services have changed the capital spending landscape sufficiently that IBM is spinning out everything else and focusing 100% on virtualized computing. Internets of things, industrial and consumer alike, create stunning possibilities for improvements in efficiency, safety, and innovation. New languages and architectures including Tensorflow, containers, adaptive processing platforms like FPGAs, and no-code development tools extend both the making and using of enterprise computing to new scenarios. And no longer are programmers limited to Microsoft, Oracle, and the usual developer toolsets: Zoom just announced it wants to be a platform, joining the likes of Slack, Dropbox, and of course Salesforce.

There’s a growing list, most recently driven by Covid-19, of consequential business and social problems that require attention. The good news is that the aforementioned technical developments make new solutions (and types of solutions) possible. Finally, businesses need the kinds of transformations that IT makes possible. Amazon’s scale, Airbnb’s flexibility, Google’s reinvention of the advertising industry: none of these competitive forces could run on 1990s computing. Whether it was the lengthening of global supply chains, the offshoring of millions of western jobs, or the complex financial instruments that made modern banking possible, modern business relies on more and more complex software applications, ideally accessible anywhere any time on any device.

Why will this time be different? I’ve spotted many reasons, and readers will no doubt have even more. Let’s look at a few.

1) The global pandemic will make capital spending extremely challenging

Yes interest rates look like they will remain low, facilitating borrowing. But the uncertainty of the pandemic’s initial duration along with the virus’s long-term effects, susceptibility to vaccines and/or therapies, and potential mutation will not clarify for years. Hard-hit industries continue to multiply: travel and tourism, hospitality, oil, retail, and the public sector only begin a list. Corporate spending on anything investigative is likely to remain constrained for 3 years, probably more.

2) The IT vendor landscape is changing

In the old days, “nobody ever got fired for buying IBM.” Later, duopolies or oligopolies emerged in sector after sector: databases (IBM, Oracle, and Sybase or another flavor of the month), enterprise operating systems (Unix variants, Windows NT), networking (Cisco and Juniper), and so on. Now, several factors represent a changed landscape from 15 years ago.

A) The installed base is a complex mess of new and old; on-premise and cloud; startups, warhorses, and unsupported failed companies; mobile and desktop; top-down and bottom-up; free and licensed; and backbones from accumulated merged entities. “Greenfield” is a ridiculous notion in this day and age.

B) Virtualization has destroyed the old ISO “layer cake” model of who did what. Is a storage area network storage or networking? Given that a social graph is a database, can our DBAs run one? Is Google Maps data or application? VMWare, a software company 80% owned by Dell, a hardware company, sells virtualization that runs on Amazon's AWS. Who's the lead dog on that sled? Figuring out which vendor can, should, and will do what, especially in the aforementioned complex legacy environments, is nontrivial.

C) The IS shop can no longer play gatekeeper. Beginning with Salesforce, business units and even departments began bypassing central IT and expensing SAAS seats that solved real problems, ramped up in days not months, and avoided the types of IT staffers who were either too intrusive (“you need to do it this way”) or unresponsive (“fill out a trouble ticket and we’ll sort them by priority”). In some companies, central IT tried to block Facebook at the firewall, smartphones notwithstanding.

D) While there are still big vendors selling IT, the buying environment is more complex. “One-stop shops” are harder to find. SAP stock fell 20% in one day a week ago. IBM is reorganizing, again. Amazon does AWS extremely well, but it’s not a holistic solution, especially absent pretty capable developers to deploy it on the client side of the relationship. The trend toward process outsourcing means that Accenture might run a big chunk of a company’s supply chain, but the outsourcer’s handoff points to both IS and internal SCM are far from standard or fixed. And what the contract says might not be how things actually work on the ground. Thus the days of Oracle, Deloitte, and Compaq teaming up on a big bid look very different: recall that a recent government cloud contract came down to AWS and Microsoft after Oracle and IBM were eliminated.

3) Today’s business problems are different than 1995’s

Whether it’s computational biology at a pharma company, chatbots in customer service, detecting synthetic media (deepfakes) in video on social media, or administering fair tests via online instruction, the enterprise software of the past decades won’t have, or be able to accommodate, the machine-learning and other computational resources to address many of these evolving business needs. The jokes about ERP and poured concrete resonated for a reason, and in a world where international trade agreements can change weekly, terrorism and sabotage take new forms, and markets can evaporate overnight (hello air travel), agility has become a watchword that big software packages typically don’t speak. 

4) Today’s organization is different from 1995’s

Remote workers and the new childcare realities that they bring, contractors with evolving legal status, offshore factories migrating closer to (maybe not “back”) home, pressure from multiple sides to address racial and gender equity issues, people at retirement age staying on given insufficient 401(k) savings, and the need to recruit millennials with different skills, norms, and values from the core staff — it’s hard to see much business as usual, and the upheaval will increase, not decrease, in the foreseeable future. The call to be “data-driven” echoes from the C-suite outward, but people with the skills to do and to manage such processes are not yet available in sufficient numbers. (If anything, scientific and quantitative literacy in the U.S. are declining.) Deciding to do the right things is a key part of management, but doing those things right is tough if the requisite skills are just not available. 

Finally, this time will be different in part because some things stay the same. People still resist change, personally and especially collectively. Most organizations, ERP investments notwithstanding, still run on Excel + email. Work/life don’t balance. Risk-taking has been suppressed by “risk management” departments to the point where the status quo becomes organizational law. 

If it’s going to be different, what should we look for this time around?

Privacy, algorithmic transparency, and naïveté all need to be addressed by substantive debate and laws with teeth: people have proven incredibly easy to game and the massive behavioral experiments being run by Google, Facebook, and Amazon have concrete consequences. In enterprise IT, meanwhile, when cash is king, customers can extract real change from vendors. Finally, maybe some organizations won’t waste a good crisis and instead begin the hard work of reinventing the cost structure rather than nibbling at it, the value proposition by truly engaging with customers, and the organization by taking a hard, fresh look at what’s possible and sustainable. (The Nordic countries are one place to start.)

Jeanne Ross is a longtime fixture at MIT’s Center for Information Systems research, having been director and now Principal Research Scientist in nearly 30 years there. In her new book Designed for Digital, she and two co-authors look at the IS organization that is emerging in the post-ERP/CRM era: these backbones are necessary but not sufficient, and IoT, machine learning, analytics, blockchain and many other emerging technologies need to be assessed and where applicable utilized. Her exemplars of “big companies that get it,” drawn from the CISR’s global roster, might be surprising but I can attest, after seeing hundreds of mid-career masters students, that Schneider Electric and Philips are in fact making positive moves. 

There is a lot to learn from in the book, but three core lessons from the leading companies apply in the context of this newsletter. 

1.  They experiment repeatedly.

2.  They co-create with customers.

3.  They assemble cross-functional development teams.

In Ross’s words, “These three challenges attack your habits because they tend to be different from what you have done.” In other words, it’s the behaviors that matter, not the change management consultants, or the CIO who has “a seat at the table” (most still don’t), or the project’s projected ROI.  

In learning from startups, Ross notes that big companies don’t really get the notion of the pivot, of starting down a road then changing course after new information becomes available. Yes the established company has channels to market, capital, and brand equity the startup lacks, but, in the end, the startup poses a threat to the incumbent for this single reason: small, open-minded organizations are better at changing their mind and shifting direction in light of experimental evidence.

Thus the short answer to the opening question — why will things be different this time? — is that size matters, and small is beautiful because agility often matters more than mass. For many reasons, look not for 5-year $50 million transformations, but rather quick-hit, small-scale pilots with the freedom to evolve and ideally pivot.

Early Indications November 2020: Intel Outside

It’s been a little over 9 years since the Silicon Valley venture capitalist Marc Andreesen proclaimed that “software is eating the world.” At the time, his credentials included playing a key role in the invention and commercialization of the web browser before becoming an investor whose portfolio companies included Facebook, Zynga, LinkedIn, Foursquare, Skype, Groupon, and Twitter. In the almost-decade since he staked out his intellectual position, software has created billions of dollars of wealth at the aforementioned Facebook, at Google, and Netflix. As Andreesen’s own list of ostensible world-eaters illustrates, however, it’s unclear (judging from Microsoft acquisitions Skype and LinkedIn, Groupon, or Foursquare) how much his thesis came true. Think back to 1994-2000, and recall the software companies that indeed turned the world upside-down: Akamai, Netscape, Google, PayPal, MySQL, VMWare, and Salesforce.

As those last two companies foretold (as did Andreesen’s second startup, Loudcloud — later called Opsware), the “where” of software was changing from running on a device or customer premise to connecting via the Internet to some vast data center at an undisclosed location. As we look at 2000 to 2020, many of the companies “eating the world” may in fact build software, but their competitive differentiation includes hefty portions of hardware and infrastructure. It’s easy to see that Amazon writes clever software at huge scale, but with dozens of data centers and more than a million employees, many of whom have nothing to do with code, it’s not meaningfully called a software company. Similarly, ByteDance writes clever algorithms to power its TikTok and other similarly addicting services around the world, but absent server farms, there’s nothing happening.

When scanning US-based innovation in the past 9+ years since Andreesen wrote, there’s really not much to see apart from Airbnb and Uber/Lyft: important companies, for sure. Using software to arbitrage everyday people’s capital (at Airbnb, this was true at the outset but less so now) is a game changer, but it remains to be seen if the three companies can be consistently profitable: right now the ride-sharing companies keep losing prodigious sums of money while compressing drivers’ wages and flexibility further each year. As I argued in a journal article a couple years ago, it remains unclear how much Uber’s notoriously bad behavior derived from a crappy culture fostered by the co-founder and how much it resulted from the realization that the entire model could never work. The whole point of taxi medallions was to limit supply to maintain profitable pricing; absent limits on supply, price will race to the bottom of a market.

All of this is a somewhat circuitous way of suggesting that while software is incredibly important, hardware not only matters, but is actually where more interesting things are currently happening. Intel is (to use this year’s most overused word, albeit correctly) an iconic company, on par with AT&T, GM, and Wal-Mart in that it helped define an entire epoch in business history. Former CEO Andy Grove’s paranoia about being disrupted in Christensonian fashion eventually came true: Arm chips, dismissed so easily at their launch given low benchmark performance alongside their power efficiency, were the final blow that broke the boulder of Intel’s industry dominance.

2020 has been a fascinating year on the microprocessor front. Softbank sold Arm to Nvidia to raise cash to help right the sinking Vision Fund ship. In October, meanwhile, AMD bought Xylinx, the dominant player in the broadly-defined system-on-a-chip market. Taiwan Semiconductor Manufacturing Company, meanwhile, is fab-to-the-world, most visibly to Apple, and its stock has more than doubled off its Covid-low in March. Like Apple, Google designs its own silicon, and outsources manufacturing to either TSMC or GlobalFoundaries, AMD’s spun-out semiconductor operation now owned by the Abu Dhabi sovereign wealth fund. Amazon, meanwhile, bought an Israeli chip-design firm in 2015 that now designs the Arm-based custom silicon (reportedly manufactured by TSMC) that powers big chunks of AWS. Facebook appears to rely heavily on Intel, with AI chips rumored to be in development.

Given how few of us work in close proximity to Google TPUs or Amazon Gravitons, Apple’s recent launch of its M1 chip is the first experience everyday people can have of this chip revolution. The M1 is in broad outlines a close relative of the A14 Bionic chip that powers the most recent iPhones. It is available in the Mac Mini and some laptop computers (that, remarkably, are on sale for Cyber Monday). Given that both chipsets are Arm-based, the recent announcement (in July) and release (in November) of Apple products that migrate off Intel processors is extremely significant. The performance gains are simply staggering: my favorite is that the M1 can emulate an Intel chip — relatively efficiently, but not as fast as Apple’s Arm-native code — faster than an Intel can run at full throttle. This performance comes with battery life measured in days rather than hours, without cooling fans, and at low price points. The fact that TSMC can produce 5-nanometer traces while Intel is announcing delays in its 7-nm products, out into 2022, helps explain the disparity. 

It’s unclear what markets Intel will have left: Apple is phasing out Intel chips, desktop computing is in the midst of a steep decline, and Intel captures little of the tablet market. The Wintel duopoly, meanwhile, is no longer an oligopoly, and Microsoft is focusing its energies on the cloud rather than the desktop. Lots of laptops were sold to support Covid-driven work-from-home initiatives, but that blip in sales will most likely not last given general economic softness and the size of this sudden (and one-time) refresh cycle.

I’ll spare my readers the fanboy-like praises of the many reviewers (here’s one summary) but it did warm my heart to remember seeing technology breakthroughs on a regular basis: the Mosaic browser, Altavista then Google search, Keyhole EarthViewer, Gapminder Trendalyzer, YouTube, the iPhone, then . . . what? The iPad, EarPods, Apple Watch, endless meal-delivery apps that will bankrupt already-drowning restaurants — none of these really count as anything that engages me the way the M1 reviewers report feeling. Just to take one example, opening the lid of an M1 Mac laptop has become a kind of sport, given the instant screen readiness. Even browser windows and simple operations are said to snap in a way that redefines the computing experience, reminding a certain kind of user why we went into this business in the first place. 

Early indications September 2020: YouTube and Drill Music

A month ago I had no idea what drill rap was. Then I read a review of a new book, Ballad of the Bullet, in The Economist. Thanks to the wonders of Covid librarianship, the school’s copy was shipped to me a few days later, and I then read a thoroughly engrossing and impressive piece of scholarship.

Eight years out of his UCLA PhD, Forrest Stuart is now a sociology professor at Stanford, but in the interim he taught at the University of Chicago and ran an after-school program aimed at helping community members cope with the violence of their surroundings. Once he was exposed to the rappers from the neighborhood who were trying to follow the path blazed by Chief Keef (Keith Cozart), Stuart embedded himself with them and saw firsthand the intersection of a sliver of opportunity amidst crushing poverty, social media popularity (driven by taunts of bravado), and street violence resulting from that bravado being challenged or usurped. The artifice of created personas, distributed via free online channels, fueled both multiple trappings of success (a new variant on sex, drugs, and rock and roll) and physical constraints on movement outside one’s turf.

Much as “reveal codes” taught a generation of people HTML 20+ years ago, so too was the YouTube/music playbook open for all to read. Cozart unleashed a fierce style of rapping, shaped by the brutality of his surroundings, that stood out from most other types. The videos reinforced the harshness of the sonics and were not the product of hundreds of thousands of dollars in production expenses. Authenticity and a Darwinian epistemology were paramount. As Stuart summarizes the movement, “If there is a dominant message running through virtually every drill song, video, and related content, it’s an appeal to superior authenticity: I really do these violent deeds. I really use these guns. I really sell these drugs. My rivals, however, do none of this.” (p. 6) Copying Cozart’s visual style, production techniques, distribution channels, and lyrical subject matter was straightforward, and Chicago became the home of a musical subgenre that has spread to London, Los Angeles, and elsewhere. As of 2016, 31 of 45 gangs in a six-square-block area had uploaded YouTube content. (p. viii)

Paradoxes abound. In contrast to the one Laptop Per Child school of thought, the teens Stuart observed were extremely adept at social media from a smartphone orientation; laptops for tasks such as video editing were hard to come by. Rather than learn conventional school subjects, the “drillers,” as Stuart calls them, were focused on social media. This focus took several forms. Primarily, one broadcast one’s persona via YouTube, Facebook, and Instagram. The fact that these were personas sometimes escaped police and prosecutors, who took social media posts literally, then used them as evidence of activity that may or may not have actually transpired. (Stuart notes the difference between black teens posing with firearms and white counterparts who were tolerated, celebrated, or applauded by police.) In addition, social media was also used as operational intelligence, predicting opposing gang members’ whereabouts, ideally unsuspecting and/or unaccompanied. Drive-by shootings could follow.

Second, for all the national fame (at different times in the book Stuart’s drillers travel to Indianapolis, Atlanta, and Los Angeles), being recognized even a block or two outside one’s home turf could be extremely dangerous. One consequence was to be “found wanting,” as one was forced on camera to renounce one’s gang superiority, walk back one’s prior claims (a certain kind of poser was known as a “computer gangster”), and commit other emasculations. 

Third, the benefits of fame tended to be more social than financial. Much as most blues musicians were never paid royalties by record labels, drillers often uploaded their videos to sites owned by videographers, producers, or other people more expert in managing Google’s revenue-sharing. Cash payouts could come when rappers were “featured” on another aspiring artist’s videos, but cash also was expended on said videographers and recording studios. In addition, other entities cashed in on the drillers, ranging from bloggers who highlighted social-media “beefs” that could in these quarters escalate to violence to Google and Facebook themselves. Instead, the benefits of local fame — “clout” in the nomenclature — could be as simple as getting respect from one’s family (one rapper had been kicked out of his mother’s house before finding fame and being let back in) or getting attention from females in the court of teen public opinion.

The book’s insights are many.  While there may be posturing that suggests drug dealing, for many teens climbing the hierarchy of the “corporate” gangs of the ‘80s is no longer an option: established dealers distance themselves from the social media frenzy and no longer stake out a newcomer to the operation with product and a market to prospect. The teens depicted in the book actually lost money in their brief experiment with dealing, through friends-and-family discounts, stolen stashes, and too many in-house samples of the product.

Rather than make money on drugs, the drillers quickly learn algorithmic scaffolding: unknown rappers can capitalize on better-known acts by posting “diss tracks” that use big viewership numbers to pull the newcomer along, albeit at real risk: you can only insult a national name so many times before retaliation comes. Elsewhere, the stereotype of the “digital divide” is tested, found wanting, and replaced by a more nuanced view of “digital disadvantage” among the urban poor. Micro-celebrity for these YouTubers is not of the same variety as that of travel “influencers” or fashion bloggers: they can’t quit if the rewards are insufficient or the blogger gets bored. This all-in pursuit of fame can be two-edged: as teens grow into adulthood and perhaps seek to leave gang life behind, erasing one’s social media presence, linked as it is to the gang, can be much more difficult than getting the tattoos removed.

There is much more to applaud. Stuart reflects carefully and usefully on the dangers of ethnography as voyeurism or self-aggrandizement. Ballad of the Bullet has much to teach about poverty and its potential remedies, about race and racism, about the atypical adoption (and consequences) of digital technologies. Most centrally, the highly consequential intersection of online fame with street-level mortality fuels new insights in the larger inquiry into online video’s many externalities. Forrest Stuart brings stand-up credibility, clear prose, and reflective insight to a corner of the Internet few adults will have encountered. His book broadened my perspective markedly, and I recommend it enthusiastically.

Early Indications August 2020: Book Review of Thomas Gryta and Ted Mann, Lights Out: Pride, Delusion, and the Fall of General Electric

As with so many aspects of economics and finance, an entity’s name often says little about what it actually does or contains. The Dow Jones Industrial average, a basket of (currently) 30 stocks, has never really been an accurate reflection of the U.S. manufacturing sector. Founded in 1896 by Charles Dow (co-founder of current-day Dow Jones), the original index was comprised of 12 companies, many of which represented what today might be called basic materials: cotton oil, sugar, tobacco, rubber, lead, and coal. The “Distilling & Cattle Feeding Co” is still with us as Jim Beam, now owned by the Japanese Suntory group. Laclede Gas still provides natural gas to communities in Missouri. Any pretext of only tracking companies that made things was quickly lost: AT&T and Western Union joined in 1916; Sears Roebuck and F. W. Woolworth were added in 1924. More industrially, General Electric was one of the original 12 companies on the Dow, dropped off two years later, and readded in 1899. 

GE’s long presence on the Dow belies the evolution of the company, which by 2000 was an advantageous blend of an investment bank and a heavy/advanced manufacturing conglomerate: in terms of naming accuracy, it was very “general” but not particularly “electric.” The business practices and assets of the factory-based side allowed the bank to borrow at the lowest possible interest rates at the same time that equity markets favorably valued the company more like a metal-bender than a bank. The double standard eventually disintegrated, and under CEO Jeff Immelt GE shed many of its financial services divisions in the midst of 370 divestitures; at the same time it acquired 380 companies. Bankers and lawyers were delighted — M&A fees ran an estimated $1.7 billion to fund the shuffle — while shareholders had less reason to cheer. This evolution eventually failed, and today GE faces fundamental reinvention. As it promises, Lights Out credibly tells the story of the fateful 20 years in which the company came to its moment of reckoning.

I had many questions that the book did not answer. For many years I taught whole sections of GE employees in my online supply chain masters classes at Penn State. Their devotion to Six Sigma/Lean Manufacturing was durable bordering on enthusiastic, but Lean is not usually a good tool for fostering radical innovation. I hope someday someone else is able to tell that story. Second, GE was a leader in adapting 3D printing to industrial production (not just prototyping) and while additive manufacturing wasn’t going to save the Titanic that was GE as of 2018, I wonder what is being salvaged from those pioneering efforts. The big question, which outsiders will never be able to answer, concerned the culture in which bad decisions went unchallenged by either the board or by middle managers with sufficient data to see pitfalls: the Alstom acquisition was built on shaky logic at the outset, and the EU’s pressure for concessions made GE leadership’s business case at closing completely fanciful.


Much of the book is devoted to the portrayal of CEO Immelt as global potentate, and rightly so. He flew with a spare business jet following him “just in case,” which cost shareholders many millions of dollars over his period in the job. Meetings with kings and presidents were part of the job; GE was one of those companies not only too big to fail but too big for political leaders to ignore. For all his skill in these settings, the job as Immelt defined it removed him still further from customers and their market realities.

Immelt made his name as a salesman, and the personality trait of not taking no for an answer (and the unshakable confidence a master closer must possess) led to multiple deals in which GE either overpaid or bought an asset for the wrong reasons: “synergies” that were so often promised rarely materialized on the income statement. More crucially, Immelt focused his attention far more on the share price than on his customers, and for all his desire to be valued like a tech company, GE appeared to have little of Apple’s human-centric design sense, Amazon’s customer-obsessiveness, or Google’s user-facing performance improvements. Knowing the complex financial structures of the business through the eyes of trusted lieutenants appeared to be far more important than step functions in the customer value proposition. Buying market share may work for a few quarters, but rarely did a Baker Hughes, an Alstom, or a Vivendi drive innovation and customer value.


One of Immelt’s highly visible missteps in his latter days was an apparently under-informed faith in the “industrial Internet.” This set of extremely expensive initiatives, designed more to goose the stock price than to deliver customer value, set out to establish GE as the Google of connected MRI machines, drilling platforms, and locomotives. It had all the signals of what a colleague in consulting once called “management by magazine,” the practice of a high-ranking executive reading some oversimplified account of computational magic and saying “make it so” in his or her company. Immelt showed no evidence of understanding cloud vs edge computing, networking and storage at massive scale, data curation, or the limits of data interoperability (standards and protocols). In and of themselves, these technical shortcomings shouldn’t be fatal: the Internet of Things is still in its early stages, and the challenge of instrumenting heavy machines and digesting the data they produce is non-trivial.

Immelt’s failing was more fundamental: he never seemed to have asked his business and technical experts for even back-of-the-envelope financial projections. Let’s assume sensors and analytics could deliver the promised 1% performance improvements — in jet engine fuel economy, in gas turbine reliability, in locomotive predictive maintenance accuracy. Who would pay for those sensors, that infrastructure, and the requisite brainpower? Companies that forked over multiple millions for a GE-made asset were (and are) not likely to pay for the privilege of letting GE siphon data from their assets only to sell it back to them in the form of more expensive service contracts: “here’s a way to run your generation facility longer between shutdowns. That’ll cost you $x million.” 


The story of GE’s embarrassing commercials featuring Owen the programmer who couldn’t lift his father’s sledgehammer and eschewed writing code for layering tropical fruits into photos of small furry animals turns out to be merely the tip of the Predix iceberg. Those commercials were intended for an ill-defined audience: 20-something data scientists weren’t turning down Facebook or Uber to work for GE, and the company’s buyers of capital equipment also did not react consistently favorably to the ads. Some contended that the campaign was intended for investors, who did not react as GE hoped, and the company’s retirees were largely furious at the tone as much as the content. Below the iceberg’s waterline, as it were, the combination of a naive (at best) business model with an insufficiently robust technical team meant that Predix was doubly doomed: making such a whiteboard vision work at scale still exceeds GE’s level of digital expertise years later, and even if the technology could have worked, there was never a realistic path to profitability in the market.

In the end, three aforementioned trends converged to bring down a mighty company: GE was dropped from the Dow in 2018, replaced by Walgreens Boots. First, the company relied more heavily on financial engineering (in the process, falling prey to Wall Street’s quarterly focus) than technical innovation. Second, GE lost focus on customer markets: Immelt bought and sold companies, not generators or drilling platforms. Finally, leadership got cute, trying to achieve with M&A what it could not with fundamental strategy and execution — while silencing voices of realism and dissent. Strategy, culture, and quality of execution all contributed to the fall of GE, which raises the question of lessons.


If we assume that companies age and decay faster than before — it’s hard to image another 120-year Dow tenure — companies like Microsoft, Google, and Apple are in mid-life. Many founders have disappeared: Apple lost its resurrected co-founder/messianic CEO almost 9 years ago, Google’s co-founders have retreated from operational responsibility, and Bill Gates is a hero of public health, not a monopolist-villain; Microsoft and Apple have both executed successful succession plans (the former after a 14-year mistake). The Economist recently profiled Google at mid-life (the jury is still out on its succession plan), and investors have to be very nervous about succession at Amazon. 


What can GE teach these companies? 1) Nothing can grow forever, and we see limits to scale: Dell peaked in 2006 and posted record net income in 2020 after six straight years of losses. IBM revenue peaked in 2011, arguably after its industry influence did. 2) Losing sight of customers and innovating to solve real issues can be fatal: Immelt and Steve Ballmer at Microsoft shared several characteristics, including being sales executives. 3) Managing to the share price is a real temptation that Jeff Bezos and Tim Cook (for very different reasons) have avoided. 4) Most important, you have to make the right technology bets and have the engineers who can make them work. GE had no signature customer-facing innovation for years, certainly nothing on the scale of Amazon’s cloud or Alexa platforms, Apple’s pivot to iPhone-linked services, or Google’s AI-powered operations in multiple markets. Call it revenge of the nerds, but the best CFO in the world can’t compensate for a shortfall of well-deployed and well-managed technical talent. Google has both a top-drawer CFO and terrific technical talent, but there are strains in the relationship between the two; Facebook too has cultural issues between programmers and management. It may turn out that keeping your engineers happy and well focused is the mark of a healthy company in the mid-21st century.