Philippa Kaur

When evan spiegel, boss of Snap, wrote in a leaked memo that the social-media company had been “punched in the face hard by 2022’s new economic reality”, he might as well have been describing America’s digital darlings as a whole. After a multi-year bull run, the sector is suffering a sharp correction. The NASDAQ index, home to many consumer-internet firms, has fallen by nearly 30% in the past 12 months; the Dow Jones Industrial Average, made up of less techie firms, is down by less than 10%. Crunchbase, a data provider, estimates that American tech firms have already shed more than 45,000 jobs this year.Macroeconomics is partly to blame. Soaring inflation and rising mortgage repayments are leading consumers to cut back on discretionary spending—and most digital offerings are discretionary. Even the industry’s trillion-dollar giants have not been spared, despite continuing to rake in handsome profits. Alphabet, Amazon, Apple and Microsoft have collectively lost $2trn in market value over the past 12 months. If you think big tech has it bad, spare a thought for the not-so-big tech. In particular, three business models embraced by firms born after the dotcom crash of 2001—and subsequently by investors—are losing steam: the movers (which shuttle people or things around cities), the streamers (which offer music and tv online) and the creepers (which make money by watching their users and selling eerily well-targeted ads). Over the past year, the firms that epitomise these business models—Uber and DoorDash; Netflix and Spotify; and Snap and Meta (which has tumbled spectacularly out of the trillion-dollar club)—have shed two-thirds of their market capitalisation on average (see chart). And things could get worse. Despite being the global leader in ride-hailing, Uber is expected to report yet another quarter of negative free cashflow (the money companies generate after subtracting capital investments). In its 13-year life it has torched a cumulative $25bn of cash, equivalent to roughly half its current market value. DoorDash, the leader in food delivery, also remains lossmaking. So do Spotify (despite decent revenue growth) and Snap (in addition to sharply slowing sales). Netflix—a child of the 1990s but a streamer only since 2007—turns a profit but its revenue growth was down to 6% year on year in the third quarter, compared with a historical average of more than 20%. Meta’s revenues have now shrunk for two consecutive quarters. On the surface, the movers, streamers and creepers—and thus their problems—look distinct. On closer inspection, though, their businesses all turn out to face the same main pitfalls: a misplaced faith in network effects, low barriers to entry and a dependence on someone else’s platform. Start with network effects, or “flywheels” in Silicon Valley speak—the idea that a product’s value to a user rises with the number of users. Once the user base passes a certain threshold, the argument goes, the flywheel powers a self-perpetuating cycle of growth. It also explains why so many startups seek growth at all cost, spending millions acquiring ever more customers to get the flywheel spinning. Network effects are real. But they also have their limits. Uber believed that its headstart in ride-hailing gave it a ticket to riches, as more riders and drivers would mean less idle time for both, drawing ever more users into an unstoppable vortex. Instead, it encountered diminishing returns to scale: reducing average wait times from two minutes to one would require twice as many drivers, even though most riders would barely notice the difference. DoorDash’s hungry consumers likewise only require so many alternative Indian restaurants to choose from. And what network effects the movers enjoy are local; a user in New York cares little about the popularity of the app in Los Angeles.Spotify and Netflix also tried to capitalise on network effects, as oodles of data on the listening and viewing habits of similar users promised to deliver an unbeatable product. Belief that Netflix’s trove of user information would give it a winning edge in creating content has been shattered by flops like “True Memoirs of an International Assassin”, which scored a rare 0% audience rating on Rotten Tomatoes, a review website. For the creepers—whose social networks are a network-effects business par excellence—the worry is what happens if the flywheels start spinning in reverse. Meta had a scare in the fourth quarter of 2021, when it lost 1m users. That loss did not turn into a stampede; the company has added users since. Next time it may not be so lucky.The second problem—low barriers to entry—also looks like a supposed boon that turned into a bane. Advances in technology, from smartphones to cloud computing, allowed all manner of startups, including the movers, streamers and creepers, to build consumer software cheaply and quickly. But that also meant that copycats soon emerged, and easy money allowed them to offer generous discounts to quickly build the minimum necessary scale. Although Uber faces only one real ride-hailing rival, Lyft, in its home market, its global expansion almost immediately ran up against local rivals such as Didi in China or Grab and Gojek in South-East Asia. The combination of relatively simple products and free-of-charge user experience means a new twist on social media can be enough for a new challenger to gain momentum: just try to pry a teenager from TikTok. The barriers to entry for the streamers are higher—Netflix and Spotify spend a lot of money making or licensing content. But they are not insurmountable, especially for deep-pocketed rivals. To fend off the challenge from Disney, which is spending a total of $30bn a year on content, Netflix has to keep splurging, too, to the tune of around $17bn a year. Like customer-acquisition costs for the movers, content costs eat into streamers’ profits. Disney’s streaming services lost $1.1bn in the second quarter of this year and the company has said that its Disney+ platform expects to lose money until 2024. Heavy investment explains why Netflix’s free cashflow is equal to only 6% of revenue.The third flaw common to the three wobbly business models is their reliance on distribution platforms that are not their own. Uber and DoorDash pay a handsome fee to advertise on the iPhone and Alphabet’s Android app stores. Spotify forks over a 15% commission on subscriptions purchased on iPhones—a tax so annoying that it has filed a complaint against Apple over it. Netflix avoids the commission by forcing users to subscribe through their web browser, shifting the irritation to the customer—and quite possibly missing out on subscriptions.Worst affected by the lack of their own rails are the creepers. Their dependence on the iPhone-Android duopoly is an existential threat. Apple’s newish requirement that users give iPhone apps permission to track their activity across other apps and websites, a move since replicated by Alphabet, may this year cost Meta an estimated $10bn in forgone revenue. Parler, a creeper favoured by the far right for its liberal attitude to speech norms, was temporarily suspended by both Apple and Android. If American national-security hawks worried about TikTok’s Chinese ownership get their way and force Apple and Alphabet to expel it from their app stores, the rising star of social media could find itself similarly thwacked.The different business models do not face an equal balance of challenges. The movers would be in better nick if the industry had meaningful barriers to entry. The streamers may have been able to bat away new entrants if network effects had been stronger. And the creepers were in reasonable shape until Apple and Alphabet spoiled their party. One shaky pillar is problematic enough. Three of them is a disaster waiting to happen. ■ More

“The bird is freed,” tweeted Elon Musk late on October 27th, after at last completing his acquisition of Twitter. The world’s richest man (and third-most followed tweeter, fast closing in on Justin Bieber) now owns arguably the world’s most influential news platform. He has already reportedly sacked Twitter’s chief executive and has changed his own Twitter profile to “Chief Twit”.Mr Musk spent most of the past six months trying unsuccessfully to wriggle out of the deal. In April he agreed to pay $44bn for the company, just as tech stocks started to slide. By July Twitter’s market value had fallen below $25bn. Since then the climate has only soured. This week Alphabet, Amazon and Meta all saw double-digit percentage drops in their share price. Twitter’s much-criticised board has in the end extracted what looks like a sweet deal for shareholders.Is it a good deal for Twitter’s 240m daily users? Mr Musk has promised a more relaxed approach to content moderation on the platform, describing himself earlier this year as a “free-speech absolutist” and suggesting that only tweets that violate the law should be taken down. Like most social-media platforms, Twitter currently bans some posts that are undesirable but legal: it recently suspended Kanye West, a singer, for a string of anti-Semitic remarks, for instance.Yet Mr Musk seems to be cooling on this idea. On the day the deal was closed, he tweeted a message addressed to Twitter advertisers promising that “Twitter obviously cannot become a free-for-all hellscape, where anything can be said with no consequences!” Other social-media bosses have watered down their free speech absolutism in recent years, following Donald Trump’s presidency and the covid-19 pandemic, both of which sparked online waves of misinformation. Mark Zuckerberg, who had previously defended the principle of “everyone having a voice” banned once-permitted content including anti-vaccination material, Holocaust denial and QAnon conspiracies from Facebook in 2020.The other niggle is digital ads, which is currently how Twitter makes nearly all its money. Mr Musk has said that he “hates advertising”. There has been speculation that he might try to turn Twitter into a subscription product instead. Making this pay would be difficult. Twitter has a modest subscription option called Twitter Blue, costing $4.99 a month. But Twitter’s accounts suggest that the average American user brings in over $6 a month in ad revenue. Would people pay? Some might, but Twitter needs plenty of tweeters to keep its content coming. Mr Musk seems to be backpedalling here, too. He proclaimed on October 27th that “I also very much believe that advertisng, when done right, can delight, entertain and inform you…low-relevancy ads are spam, but highly relevant ads are actually content!”Any meaningful changes will be made harder by the immediate need to contain costs. Twitter is probably overstaffed: last year it had 1.5 employees for every $1m in revenue, compared with 0.6 at Meta. At the same time, if reports are true that the company is losing 75% of its workforce—either because they get the boot or are repelled by Mr Musk—getting anything done, let alone anything big, may prove harder. Mr Musk may not be in it for the money. But the private backers he brings along, including a few fellow billionaires and a Qatari sovereign-wealth fund, probably fancy a return on their investment. Twitter may be freed, but its owner may find himself in a $44bn cage. ■ More

The corporate jets descended on Riyadh this week, ferrying chief executives to the Future Investment Initiative, a talkfest nicknamed “Davos in the Desert”. A feud between the American and Saudi governments over an oil-production cut by the opec+ cartel, a move which benefits fellow member Russia, was not enough to keep away the bosses of giant American banks like JPMorgan Chase and Goldman Sachs. Nor was the kingdom’s record of human-rights abuses. Listen to this story. Enjoy more audio and podcasts on More

The office is where colleagues meet, work and bond. But it is also a time capsule, a place where the imprint of historic patterns of working are visible everywhere. The pandemic has heightened this sense of the office as a dig site for corporate archaeologists. It isn’t just that covid-19 has left its own trace in the fossil record, from hand sanitisers to social-distancing stickers. It is also that items which were useful in the pre-covid world make less sense now; and that things which were already looking quaint seem positively antiquated. Listen to this story. Enjoy more audio and podcasts on More

Huawei once looked unstoppable. Having began in 1987 selling phone switches from a flat in the southern city of Shenzhen, in 2012 the Chinese technology firm overtook Ericsson, a Swedish rival, to become the world’s biggest maker of telecoms gear. By 2020 its market share in the business exceeded 30%, roughly as much as Ericsson and Nokia of Finland, its two main competitors, combined. The same year it surpassed Samsung as the largest maker of smartphones. Its fast-growing software and cloud-computing businesses were beginning to compete with America’s ibm and Oracle.The American government had other plans. Successive administrations have regarded Huawei as a national-security risk, claiming that it had deep links with the People’s Liberation Army and that its gear could be used for spying (allegations that have not been proven and that Huawei denies). Washington has banned Huawei’s wares at home and urged allies to ditch them from their 5g mobile networks. Most cripplingly, it used export controls to starve the company of American technology and products, including computer chips, the manufacturing of which relies on such tech, wherever these come from. In the latest blow, on October 24th the Justice Department said it had indicted two Chinese spies for attempting to obtain inside information about a federal investigation into Huawei.All this has turned a company on track to be one of the world’s biggest into its most controversial. The results have been devastating. After years of uninterrupted growth Huawei’s revenues collapsed by nearly 30% in 2021, from a peak of almost $140bn the year before (see chart 1). As countries across the globe roll out 5g, Huawei’s market share for telecoms networks—its main business—looks set to decline. Its mobile-phone business is dead, insiders say. The company’s 78-year-old founder and boss, Ren Zhengfei, recently told employees in a leaked memo that the company was in a fight for survival. To prevail in that fight, Mr Ren is transforming the company from one laser-focused on a few core telecoms products to a provider of tech and services to a variety of industries, from automakers to agribusiness. Whether this transformation can succeed matters not just for Huawei. America’s campaign to forestall China’s rise as a technological superpower is intensifying. This month Joe Biden’s administration announced new restrictions, covering more Chinese firms and more areas where Washington and Beijing are vying for dominance, such as artificial intelligence (ai) and supercomputing. Huawei is thus a case study in how effective American sanctions really are, how Chinese firms can adapt to the new world order and, ultimately, whether China has a shot at winning the tech race.Immobile networkFirst, consider the American effort to block Huawei from the global 5g roll-out. Geographically, the results have been mixed. America’s strategy is working in the rich Western markets of its allies. Australia, Canada, New Zealand and Sweden have followed America in banning Huawei gear outright. New rules in Britain force carriers to remove all Huawei technology from public 5g systems by 2027. The French government has asked operators to rip out Huawei gear from many parts of their networks. Other countries, such as Japan, have not barred Huawei but signalled that the company is not welcome. The constant risk of fresh restrictions has led many customers in places without bans to steer clear of Huawei. This has happened in Italy and Portugal. The developing world still seems open to Huawei’s cheap equipment. The company is furnishing 5g networks in Indonesia, Saudi Arabia, South Africa and Turkey. Brazil, another potentially large market, has flip-flopped but does not appear poised to issue a ban. Huawei executives boast of more than 5,000 commercial 5g contracts globally, ranging from full deployment of 5g networks for national carriers to upgrading networks at ports. How many more such agreements it can ink depends in part on the effectiveness of American export controls, the second anti-Huawei weapon. The restrictions, which since 2019 have limited the sale to the company of high-end chips and Google’s Android mobile operating system, have already obliterated the firm’s once-thriving smartphone business. Huawei’s own operating system, Harmony, is unattractive to consumers since it offers few apps, and it offers few apps because it lacks the consumer numbers that would make it worth developers’ while. The chip ban, meanwhile, means that even though the company has built China’s 5g network its phones lack 5g because the required radio chips rely on American tech. This forced Huawei to spin off its Honor smartphone brand in 2020. Revenues from Huawei’s remaining devices business fell by 25% in the first half of 2022, compared with a year earlier.The impact of the chip ban on the carrier business is a closely guarded secret. The processors used in network gear are less advanced than those used in smartphones and some of them could be produced locally by chipmakers such as smic, a state-controlled firm. But probably not all, at least in the near future. The Tiangang processor, designed by Huawei’s HiSilicon chip division for use in 5g networks, was fabricated by tsmc, a giant Taiwanese contract manufacturer that can no longer supply Huawei as a result of the American rules. Publicly, Huawei claims to be shipping units as normal, thanks to a stockpile. But that “will start to run out very shortly”, expects Bill Ray of Gartner, a consultancy. Behaviour in tenders for carrier contracts suggests as much. In the past 18 months Huawei has routinely bid that the highest allowed price. This implies that it is trying to maximise profits while conserving its component inventories rather than seek market share, says Edison Lee of Jefferies, an investment bank. According to disclosures on a large tender for China Mobile analysed by Jefferies, Huawei kit accounts for 47% of China Mobile’s locally manufactured servers, down from 61% last year.Globally, the company’s share of telecom-gear revenues has so far declined by less than two percentage points from its peak of more than 30% in 2020, according to Dell’Oro, a research firm (see chart 2). But Huawei’s global sales of such equipment fell by 7% last year. Much of its remaining revenue comes from China and, abroad, from less lucrative 4g networks, which are still being built in poorer countries. As investment in China’s 5g roll-out winds down, Huawei’s global market share may be eroded, says Stefan Pongratz or Dell’Oro. The idea of saving the foreign 5g business by selling it to a Western owner, which Mr Ren entertained in an interview with The Economist in 2019, appears to have been shelved. Mr Ren is undaunted, however. His leaked memo in late August, in which he asks staff to “feel the chill” brought on by gloomy economic conditions in China and abroad, should be read not as an act of despair but as his way of rallying the troops, say some executives. And there is plenty for them to rally around. Mr Ren wants Huawei to become a purveyor of technology to a wide spectrum of industries. It has already sold 300m devices running on Harmony, including laptops, wearables such as smart watches and app-controlled home appliances. This month the Financial Times reported that it may attempt to relaunch the production of 5g phones using less advanced chips. And it is venturing beyond consumer goods and telecoms. It is making sensors to monitor soil conditions to help farmers fine-tune irrigation systems and cut back on fertiliser. The company is building a business in systems for clean-power generation. It has also become a big supplier of software and electronics for carmakers, with which it has teamed up to develop various bespoke systems, such as energy management for electric vehicles (evs). Huawei says that in July alone it had sold more than 7,200 aito m5s, a model of car jointly developed with Seres, a Chinese-owned electric-vehicle maker based in California.Huawei is also beefing up its enterprise division. The unit is building data centres and cloud-computing businesses around the world. Its prospects look strong in China, where the biggest source of demand over the next decade will come from the government (including at provincial and city level, where authorities are upgrading their systems with a view to offering more public services online) and state-owned companies (which are frantically digitising and installing the industrial “internet of things”). Huawei does not enjoy a technical advantage in such “infrastructure as a service” (IaaS) over giant local competitors such as Alibaba and Tencent. But it has the government connections needed to win the most important contracts over the next decade, says Yi Zhang of Canalys, a research group. In just a few years this has helped Huawei become the second-largest cloud provider in China, behind Alibaba. Many Chinese firms are tossing out Oracle databases and asking Chinese companies to build local ones. Huawei is scooping up this business. As revenues from devices tumbled in the first half of 2022, its overall sales from the enterprise division surged by 28% to 55bn yuan ($7.6bn), or about 18% of total revenues. Gartner reckons that Huawei has become the world’s fifth-largest IaaS provider (see chart 3). Maintaining a presence in foreign markets poses a bigger challenge. Mr Ren has long understood the importance of grabbing global market share. In the late 1990s he began deploying staff to far-flung places in Africa and South America in the hope of making local connections. The strategy helped make Huawei China’s first genuinely multinational corporation. Huawei’s new businesses are not expected to make headway in America. But the company thinks much of the rest of the world is fair game. Its energy-management products are growing fast in Europe. One insider points out that over the past three years Huawei has been building up its foreign iaas engineering capabilities in Africa, Latin America, the Middle East and South-East Asia. Barriers to entry in such businesses are high even in places that welcome Huawei. Much of the world’s computer technology runs on programmes designed by Microsoft, an American tech company. Huawei’s databases use Linux, an open-source operating system. The technical difficulty of hiring Huawei to replace American systems that run on Oracle and ibm systems, which are much more compatible with Microsoft, is high, says Boris Van of Bernstein, a broker. Edging out the American firms in China is one thing; doing so abroad is quite another. And although Mr Ren has amassed heaps of chips needed for its enterprise products, the new American rules will make these harder to replenish.Most important, taken together these changes amount to a revolution in how Huawei functions as a business. In the past its sprawling research-and-development (r&d) operation dreamed up new technologies, its engineers developed them into a few core products and its sales team sold those to customers in two main sectors: telecoms and consumer electronics. This one-way end-to-end system is being replaced by a more open, two-way model, where Huawei develops new products in partnership with its growing array of client industries. People close to the group say it now resembles a vast web of startups with deep r&d coffers. The company often spends 20% of annual revenues on r&d, as much as Meta and nearly twice as much as Alphabet. That More

By a dusty stretch of the deafening road from Chennai to Bengaluru lie three colossal, anonymous buildings. Inside, away from the din of traffic, is a high-tech facility operated by Foxconn, a Taiwanese manufacturer. A short drive away Pegatron, another Taiwanese tech firm, has erected a vast new factory of its own. Salcomp, a Finnish gadget-maker, has set one up not far away. Farther west is a 500-acre campus run by Tata, an Indian conglomerate. What these closely guarded facilities have in common is their client: a demanding and secretive American firm known locally as “the fruit company”.The mushrooming of factories in southern India marks a new chapter for the world’s biggest technology company. Apple’s extraordinarily successful past two decades—revenue up 70-fold, share price up 600-fold, a market value of $2.4trn—is partly the result of a big bet on China. Apple banked on China-based factories, which now churn out more than 90% of its products, and wooed Chinese consumers, who in some years contributed up to a quarter of Apple’s revenue. Yet economic and geopolitical shifts are forcing the company to begin a hurried decoupling. Its turn away from China marks a big shift for Apple, and is emblematic of an even bigger one for the world economy.Apple’s packaging proclaims “Designed by Apple in California”, but its gadgets are assembled along a supply chain that stretches from Amazonas to Zhejiang. At the centre is China, where 150 of Apple’s biggest suppliers operate production facilities. Tim Cook, who was Apple’s head of operations before he became chief executive in 2011, pioneered the company’s approach to contract manufacturing. A regular visitor to China, Mr Cook has maintained good relations with the Chinese government, obeying its requirements to remove apps and to hold Chinese users’ data locally, where it is available to the authorities.Now a change is under way. Mr Cook, who has not been seen in China since 2019, is wooing new partners. In May he entertained Vietnam’s prime minister, Pham Minh Chinh, at Apple’s futuristic Cupertino headquarters. Next year Apple is expected to open its first physical store in India (whose prime minister, Narendra Modi, is a fan of gold iPhones). The two countries are the main beneficiaries of Apple’s strategic shift. In 2017 Apple listed 18 large suppliers in India and Vietnam; last year it had 37. In September, to much local fanfare, Apple started making its new iPhone 14 in India, where it had previously made only older models. The previous month it was reported that Apple would soon start making its MacBook laptops in Vietnam. Some of Apple’s newer gadgets show the way things are going. Almost half its AirPod earphones are made in Vietnam and by 2025 two-thirds will be, forecasts JPMorgan Chase. The bank reckons that, whereas today less than 5% of Apple’s products are made outside China, by 2025 the figure will be 25% (see chart 1). As Apple’s production system is shifting, its suppliers are diversifying away from China, too. One crude measure of this is the share of long-term assets that Taiwanese tech-hardware and electronics firms have located in China. In 2017 the average figure was 43%. Last year that had fallen to 31%, according to our estimates using company and Bloomberg data. The most pressing reason for the scramble is the need to spread operational risk. Two decades ago the garment industry beefed up its operations outside China following the sars epidemic, which paralysed supply chains. “sars made it very clear to everyone operating in China that you needed a ‘China+1’ strategy,” says Dominic Scriven of Dragon Capital, an investment firm based in Vietnam. Covid taught tech firms the same lesson. Lockdowns in Shanghai in the first half of this year temporarily shut a factory operated by Quanta, a Taiwanese firm, which was believed to be making most of Apple’s MacBooks. Customers had to wait months. Avoiding this kind of chaos is the “primary driving force” for Apple’s supply-chain moves, according to Gokul Hariharan of JPMorgan Chase.Another motive is containing costs. Average wages in China have doubled in the past decade. By 2020 a Chinese manufacturing worker typically earned $530 a month, about twice as much as one in India or Vietnam, according to a survey by JETRO, a Japanese industry body. India’s ropey infrastructure, with bad roads and an unreliable electrical grid, held it back. But it has improved, and the Indian government has sweetened the deal with subsidies. Vietnam offers tax rebates and holidays, too, as well as free-trade deals, including one recently signed with the European Union. Bureaucracy around visas and customs remains a pain. But the work ethic is similar to that in China: “Confucius still gets them out of bed in the morning,” says one foreign executive in Vietnam.Apple also increasingly sees locals as potential customers, particularly in India, the world’s second-largest market for smartphones. Apple’s gadgets are too pricey for most Indians, but that is changing. In July Apple reported that its revenues in India had nearly doubled in the past quarter, year on year, driven by the “engine” of iPhone sales. This is diminishing China’s relative importance as a consumer market. At its high point in 2015, China accounted for 25% of Apple’s annual revenues, more than all of Europe. Since then its share has steadily shrunk, to 19% so far this financial year (see chart 2). By the sounds of it Xi Jinping, China’s president, would like it to fall further. At a Communist Party shindig on October 16th he urged “self-reliance and strength in science and technology”, suggesting that foreign importers may face stiffer competition from Chinese national champions. He repeated the phrase five times.This points to the last, but potentially most significant reason for Apple’s shift: geopolitics. Rising tensions between China and America have made China an increasingly awkward place to do business. Heightened Chinese political sensitivity has added friction on many fronts. This summer, for instance, Apple reportedly had to ask Taiwanese manufacturers to label their products “Made in Chinese Taipei” to appease newly finicky Chinese customs officials (at the risk of angering Taiwanese ones). America, for its part, has become more aggressive in its competition with China’s domestic tech industry. On October 7th America announced a ban on “us persons” working for some Chinese chipmakers. On the same day it added 30 Chinese companies to a list of “unverified” firms its officials had been unable to inspect. Apple had reportedly been about to sign a deal to buy iPhone memory chips from one such company, ymtc, which can offer low prices thanks in part to a Chinese government subsidy. Following America’s export controls that deal was put on ice, according to Nikkei, a Japanese newspaper.The question is whether moving production physically out of China will be enough to avoid future crackdowns. Even as Apple makes more of its gadgets outside China, it is no less reliant on Chinese-owned companies to build them. Chinese manufacturers such as Luxshare, Goertek and Wingtech are taking an increasing share of Apple’s business beyond China’s borders. Luxshare and Goertek are reported to be making AirPods in Vietnam, helped by the fact that some Taiwanese rivals, such as Inventec, have scaled back their work for Apple in recent years. Indian media reported in September that the Indian government might allow some Chinese companies to set up production facilities in India. Chinese companies’ share of iPhone electronics production will rise from 7% this year to 24% by 2025, believes JPMorgan Chase, which predicts that in the next three years Chinese companies will increase their share of production across Apple’s range of products.Could Chinese manufacturers outside China be targeted by American sanctions? For now this is unlikely, believes Nana Li of Impax, an asset manager. “There are no handy alternative [suppliers] available with the same level of experience, efficiency and cost-effectiveness,” so cutting them off would hurt American firms, she points out. In time, that could change. Countries like India and Vietnam are keen to build up their own suppliers. Tata is reportedly in talks with Wistron, a Taiwanese manufacturer, about making iPhones in India. Indian manufacturers report that “the fruit company” is discreetly on the hunt for local suppliers. Given the direction of relations between America and China, it is surely sensible for Apple to place some side-bets, before restrictions go any further. Chinese firms outside China are safe for now, says one Western investor in Asia. But “the noose is tightening”. ■ More

Camden, a small town in the backwoods of southern Arkansas, is having an unusual brush with the outside world. It is a quiet place. At this time of year there are more Halloween dolls tied to its lampposts than there are people in the streets. It also has a reason to keep its head down. The nearby Highland Industrial Park, which has a few manicured lawns amid thousands of acres of thick forestry, is home to the factories of some of America’s biggest weapons manufacturers, such as Lockheed Martin and Raytheon Technologies. “It’s been kind of a hidden secret,” says Michael Preston, Arkansas’s secretary of commerce. Or as a local businessman whispers, “it’s a fear thing: ‘shhhh’.” Listen to this story. Enjoy more audio and podcasts on More

Any manager worth their salt knows the value of spending time “walking in their customers’ shoes”. There are many ways to do it. You can observe customers in their natural habitat. Pernod Ricard’s boss recently told Bloomberg, a news service, about his habit of bar-hopping in order to see what people want to drink. Such research is a lot less fun if your company makes soap dispensers for public toilets but the same principle applies. Listen to this story. Enjoy more audio and podcasts on More