 |
|
|
|||||
Youd think it would smell like coffee. Instead, theres a faint whiff of livestock in the air. For the past hour and a half, some 14,000 Starbucks store managers have been streaming into the Thomas & Mack Center on the campus of the University of Nevada, Las Vegas, filling the arena. Even though its only noon on this June day, the music is pumping and people are dancing under purple lights. Some start a wave, which ripples through the crowd. In just a few hours, theyll sample the companys new 1971 Roast from their seats, in what has to be one of the worlds largest coffee tastings ever staged. But even the scent of those dark-roast beans cant erase the barnyard aroma from the years of rodeos that have been held in the arena. This three-day Las Vegas extravaganza, known as the Starbucks Leadership Experience, marks the companys first such event since 2019, and the excitement is high, particularly since Brian Niccol is about to give his first major address since becoming CEO last September. Lil Jons Get Low rumbles through the arena as Niccol makes his way past the crowd, shaking hands and posing for selfies like hes running for office. Soon, he takes the stage to loud applause and quickly begins laying out his vision for returning the company to its 2010s heyday, when Starbuckss stores were known as comforting coffeehouses and its stock was on a steady climb. Niccol has been giving a version of this talkpitching what he calls his Back to Starbucks planto investors and analysts for the past several months. Now, hes facing his employees. Look, were here for a simple reason, he says, surveying the audience. Okay, a really simple reason. Our future success depends on all of you in this room. That may be true, but as Niccol well knows, much of it rests on his shoulders. When Starbucks recruited Niccol from Chipotle, it appointed him both chairman and CEO and offered him nearly $100 million, including stock awards and a cash bonus, for the first four months alone on the job. (Going forward, hes eligible for at least $24 million annually in equity and pay, a third more than McDonalds CEO Chris Kempczinski received in 2024.) It also offered him access to the corporate jet for commuting between his home in Newport Beach, California, and Starbuckss headquarters in Seattle. (He has since acquired a local residence.) Its an extraordinary pay package that bets heavily on Niccols ability to pull the iconic coffee chain out of an unprecedented slump. After decades of stamping its siren across the globe, Starbuckss reign as the worlds leading coffee company is faltering. Its global dominanceroughly 41,000 stores worldwideis being threatened by the Chinese chain Luckin, which has some 24,000 outposts, while domestic challengers such as Blank Street, Scooters, and Dutch Bros have found footholds in being anything but another Starbucks. Starbuckss transactions peaked around 2019, according to Niccol, just before the pandemic, and its revenue is stagnating. Starbucks managed to bring in $36 billion in fiscal year 2024nearly flat with 2023by offsetting dwindling customer numbers with higher prices. Changes in leadership havent helped. Niccols predecessor, Laxman Narasimhan, lasted just 17 months as CEO, despite his endorsement by the companys legendary leader, Howard Schultz. Narasimhan was pushed out last August amid falling sales and activist campaigns by Elliott Investment Management and Starboard Value. Niccols recruitment was the last major act of now-retired board member Mellody Hobson, who was banking on his reputation as the worlds leading fast-food strategist and a digital tactician. But its also a bit of a gamble. For this brand turnaround, the master of operational efficiency needs to focus on something radically different: improving customer experience. The operational process is whats going to free us up to give a great experience again, Niccol tells me. During his six years at Chipotle, Niccol revived the chain through heretical acts within the lo-fi company: adding a second make line to service online orders, giving cooks screens to smooth out logistics, and using digital rewards to entice customers. Before that, at Yum! BrandsNiccol was chief marketing officer of Pizza Hut and then Taco Bell, eventually becoming CEO of the latterhe launched apps for both brands, bolstering their sales. At Chipotle and Yum! Brands, Niccol accelerated the shift to digital orders. But Starbucks doesnt need to digitize its operations. Today, 31% of its orders are already placed within its rewards-laden app, which is so effective that its become something of a North Star for the fast-food industry. If anything, digital orders are working too well, undermining the ecosystem of a Starbucks café. Over the past decade, Starbucks has become known for chaotic stores staffed by baristas who struggle to keep pace with orders; unpredictable waits that have made mobile ordering a defensive strategy to get your coffee on time; and cafés that offer neither the space nor seats for customers to hang out or work at their laptops. For many people, the daily ritual of a Starbucks latte no longer feels worth the (considerable) price or trouble. When the opportunity at Starbucks popped up, I spent a little bit of time looking at, ‘Okay, what makes Starbucks Starbucks?’ Niccol says. It was clear that everybody really had a connection to the experience. And unfortunately, a lot of them talked about it in the past tense. The Starbucks machine, Niccol argues, had gone too far, replacing hand-scrawled Sharpie names on cups with printed labels, overwhelming baristas so they didnt have time for human connection, and either ripping out café seats entirely during the COVID-19 pandemic or replacing them with the sort of hard, 15-minute chairs that, legend has it, Ray Kroc pioneered at McDonalds. In Starbuckss early years, Schultz borrowed from urban theorists to describe his cafés as the third place between work and home. Niccol aims to not just revive that ideabut to reconcile it with the iPhone era. Nearly a year into the job, his plans are still solidifying. He brought back the Sharpie and updated the dress codes for baristas (despite protests). Hes offering free coffee refills, and hes reintroduced the condiment bar so that customers can add cream to their coffee themselves and stopped upcharging for alt milks. (He also no longer offers free water to non-paying visitors, to discourage loitering.) Bigger changes are coming, including the redesign of 1,000 of Starbuckss 11,000 company-operated cafés in North America by the end of next year, and adding 30,000 seats back to stores over the next few years. Niccol is also planning to sunset the companys 90 or so mobile-order and pickup-only stores, which he has described as overly transactional. We were creating soulless vessels to just hand people coffee. That isnt right, says Niccol. Im going to win more drive-through and mobile orders because of the coffee experience that you have every once in a while when you sit in our café. In the restaurant industry, store redesigns are part of a proven playbook, says Bank of America senior research analyst Sara Senatore. People want to go where [other] people are, she explains. That free refll isnt just for people who get the refills, just like the seating isnt just for people who use the seating. That visual is important for everybody. Whether Niccols changes are coming quickly enough for Wall Street is another matter. For the third quarter of fiscal year 2025, Starbucks grew revenue by just 4%, while same-store sales declined by 2% in the U.S. and 2% globallymarking six quarters in a row of same-store sales declines. There are early signs of success: An increasing number of stores are seeing growth in transaction volumes; non-rewards customers are starting to return. But even Senatore, a Starbucks bull, noted on the companys April earnings call that its a little slower in terms of earnings recovery than maybe some of the previous turnarounds weve seen from you, Brian. Chipotles same-store sales had improved the quarter Niccol arrived. But with Starbucks, the challenge is more complex: No one ever cherished Chipotle for its plush seats and warm atmosphere. Mike Grams, COO of Starbucks, walks into a Starbucks bathroom and inhales, filling his nostrils like hes breathing the Alpine air of a Ricola commercial. The bathroom, thank god, smells like air freshener. Were at one of Starbuckss first redesigned stores in Bridgehampton, New York, and, alongside Meredith Sandland, Starbuckss EVP and chief coffeehouse developer, Grams is taking me on a coffeehouse walk, where he surveys the customer experience of a Starbucks. The duo, who are both just 100 days into their new roles, last worked with Niccol at Taco Bell. Niccol coaxed Grams, who had risen from store manager to COO at Taco Bell, out of retirement to help resuscitate Starbucks. (Niccol describes Grams, who brims with football coach energy, as a human quad shot.) Meanwhile, Sandland, the former chief development officer at Taco Bell, spent the past three years building ghost kitchens software at the startup Empower Delivery. When she joined Starbucks, the coffee chain licensed Empowers software and hired six of its engineers. Gramss coffeehouse walk will soon be practiced by Starbuckss 1,000 North American district managers, who visit different cafés a few days a week. On this afternoon, he begins outside the store, inspecting the Starbucks sign for bird nests. Then he walks through the space with all his senses on alert, ensuring sounds, sights, and smells layer harmoniously. Sandland notices the stores music is a bit too loud for conversation, but shes pleased to hear the coffee steam behind the counter. Grams catches a thin line of dirt piling up at the threshold of the exit door. Its easy to walk past very basic things, Grams says. Getting the store design right is another challenge. Today, most Starbucks locations are a study in stark white, black, and gray; every surface has been hardened for easy cleaning, down to the chairs. The result is a space thats fine to look at but miserable to linger in, built less to be sat in than hosed off. [Our] coffeehouse needs to be better than any living room youre in, Niccol says, citing affordable luxury as the companys new guiding principle. Its easy to see why the company chose to show off this café to me. Its cozy, with rich green walls and louvered shutters on the windows. The worn wood floor has been refinished instead of covered over with Starbuckss usual tile. Books sit on shelves. Large plants fill corners. The counter, with its drab display of sample food items, will also get an update. Grams points to the pastry case, where flat frozen croissants are presented as if for autopsy on sterile white plates. Those plates will be replaced with baskets and fluffier croissants, freshly baked in-house. The Espresso Lane: Meredith Sandland, EVP and chief coffeehouse developer,is helping Starbucks streamline operations with new software. [Photo: Jessica Chou] Under Niccol, the company is also reviving multi-zone seating, pockets that are meant to serve different purposes. Here that includes a large wooden dining room table where remote workers can use their laptops, as well as low, plush chairs for reading. Café tables, still set with those hard 15-minute seats, allow pairs of people to catch up. But in other areas, those chairs are topped with a one-inch leather cushion, making them more inviting. The biggest update may be the new padded leather bench seating. When I settle into it with my full weight, I find its both comfortable and supportive of my back. Its also great for adding seating capacity to smaller stores. Niccol has commissioned some additional furnishings. They include glasses and mugs that will nod at the siren while remaining elegant, he says, and what he calls an iconic chairsomething so comfortable and expensive it might only appear in half or two-thirds of stores, a throwback to the quirky purple Starbucks chair of the 90s. Its got to be the seat that when you walk in, youre like, Man, I cant wait for him to get up. Im hopping in that chair the second he does, Niccol says. Analysts had estimated that these redesignswhich are now debuting across New York and Los Angeles, and spreading to 1,000 stores over the next yearcould cost as much as $1 million apiece. Niccol insists thats an overestimation of what he calls a store uplift. He characterizes what hes rolling out now as cosmetic renovations, and says theyll cost around $150,000 apiece and can be accomplished over the course of a single night. (He also puts the price tag for a true store renovation at about $450,000 or less.) The effect can be contagious: Historically, even having just one fancier Starbucks café in an area has driven repeat business to half a dozen lower-touch stores in the vicinity. And improving a store experience has a rapid effect on a brands reputation. Thats why, back in 1999, Starbucks hired Ted Jacobs as an early global head of design. Hed helped reinvent retail with the launch of flagship NikeTown stores in cities around the world. But even the most enticingly redesigned stores will feel superficial if Starbucks doesnt find a solution for its biggest problem: the pool of customers hanging by the counter for their drink order, ruining any feeling of coziness. Its an issue that has plagued the company for a decade. Bill Sleeth, who was the VP of design at Starbucks under Schultz from 2010 to 2017, remembers exactly when things went awry. After the companys rapid expansion in the early 2000s, when it adopted a somewhat cookie-cutter approach to store design, Sleeth worked to infuse the cafés with more comfortable and local elements. He also introduced the high-touch Reserve and Roastery stores to bring a halo effect back to the brand. With Sleeths initial global store redesign strategy working (the companys same-store sales grew between 6% and 9% annually between 2010 and 2015), he began testing noel store layouts to help reduce the line of customers waiting to place orders. But the companys tech team had other plans. In September 2015, Starbucks launched Mobile Order and Pay nationwide. The whole thing took us by surprise, Sleeth recalls. If it had been just a few ordersjust a convenience for customersit wouldnt have been a big deal. But it completely transformed the business. Previously, every Starbucks had been designed to accommodate a winding queue in front, right before the cash register. But mobile-order pickups essentially repositioned the crowd to an ad hoc holding pen after the register. Cafés simply didnt have the space for this. So drinks piled up on counters, and customers pooled into the middle of the store. Sleeths team built little risers to hold all the drinks ready for pickup, and more recently, Starbucks updated stores to move the drink area closer to the front doors. Even so, the vibe for anyone sticking around is grim. Our stores blew up in our face. Our success became that handicap, Sleeth says. It broke us. Michelle Eisen was a Starbucks barista in Buffalo when, in 2015, she and her colleagues began affixing receipt-like stickers onto all cups instead of hand-writing names. They loathed the automation. We were resistant, she says. It seemed contradictory to the experience we were supposed to provide. In her eyes, the paradigmatic Starbucks coffeehouse disappeared along with those Sharpie-signed cups. But nearly 10 years later, Niccols mandated use of the Sharpie was, to Eisen, even worse. Head counts had been quietly cut over the yearsthe average Starbucks store had 24 employees five years ago; today, it has 18even as the drink menu became increasingly complicated. The job had gotten harder. One of Niccols first orders of business was paring back that menu, eliminating 30% of options, things like melon syrup and freeze-dried pineapple. (Though he is introducing a protein cold foam.) But Eisen says that customers still came in and asked for their old standbys, and she found it impossible to turn them down. And then, suddenly, she was supposed to pick up a Sharpie and start writing on dozens of cups per hour. Im an experienced barista, and [even] I struggled, especially during peak. If I could get a smiley on every cup, it was a huge feat, she recalls. Its awful to feel like youre doing everything in your power and youre just drowning. In May, Eisen quit her job to join the Starbucks Workers United union as a full-time organizer, helping baristas around the country join the ranks of the 12,000 employees across 620 North American stores who have voted to unionize as of time of publishing. (In addition to its 11,000 company-operated stores in the region, Starbucks has 7,000 licensed ones; the company runs just over 50% of its stores globally.) Eisen offers a counter perspective on Niccols Back to Starbucks plan, outlining how something as simple as updating the dress codefrom tops of any color with faint patterns to clean black T-shirtshad massive repercussions for employees without washing machines, who can be sent home without pay if they fail to comply (many baristas spill on themselves at least once a shift). She also points out that refusing a visitor’s request for a cup of water always ends in a fight, making a case that Niccols operational updates have largely come at the expense of the already beleaguered workforce. Indeed, after Niccol announced the dress code updates in April, the union responded with coordinated strikes across 175 stores, creating a terrible public relations moment for a company attempting a turnaround. (According to analyst firm Paragon Intel, Niccols lack of experience negotiating with unions is one of his few weak spots as a CEO: The National Labor Relations Board found Chipotle violated labor laws under his leadership by closing a store in Maine that was attempting to unionize.) Starbucks and the union are currently in stalled contract negotiations, with pay increases a sticking point. Grounds Keeper: Mike Grams joined Starbucks as COO in February to help oversee its store redesign. [Photo: Jessica Chou] Niccol seems to acknowledge his early missteps. In our conversations, he floats the idea of charging a quarter for Starbuckss triple-filtered water. He also wonders if he should have first tested the dress code at the companys new Starting Five storesfive cafés where Starbucks corporate now tries out products and ideas before releasing them more widely. But overall, he defends the moves. Ninety percent of the peoples feedback was hugely positive. The 10% that decided to be very public about their dissatisfaction got a lot of pickup, and I was surprised by that, Niccol says. But I think we made the right choice. He sees Starbuckss labor challenges as rooted in operational problems. Back-of-house updates, he argues, will buy back time and sanity for the staff and create a warmer atmosphere for customersensuring baristas have the wherewithal to provide the hospitality of remembering customers names and handing them their drinks. The companys biggest target is Sleeths former nemesis: mobile ordering, which produces a sea of tickets that can overwhelm staff. (Niccol likens this dynamic to a memorable scene from The Bear, when the kitchen at the Chicago sandwich shop is so flooded by online orders that the staff just gives up.) This September, Starbucks is rolling out new logic for Mobile Order and Pay called Smart Queue. Instead of cueing baristas to make drinks (ordered in the store and online) on a first come, first served basis, it will more intelligently prioritize orders. In-store drinks will be ready within four minutes and online orders within 12 minutes, while food and drinks will be grouped together. People ordering via the app will be able to select a pickup time for 15 minutes or further into the future, the sort of simple upgrade that will allow practical prioritization. Niccol reports that 80% of in-café orders are handed off within four minutes in stores that are already using Smart Queue. Sandland, as head of store development, is leading the project. There might be the same number of people coming in to get their drink, but if you can better match the right drink to the right person, theres going to be fewer drinks sitting on the bar and fewer people hanging around waiting, she says. Other updates include a point-of-sale terminal that lists 82% of orders as one-tap buttons on the home screen for cashiers, making it easier for them to keep things humming along. Theres even a tab for viral fan-made drinks, which Starbucks promotes through a new Secret Menu in its app. (The company scours social media for these drink ideas, tweaking recipes for scalability.) Starbucks is developing an espresso machine that pulls four shots at a time instead of three, saving 27 seconds when someone orders a quad shot, which it hopes to start rolling out next year. And its upgrading software to make it easier for managers and employees to swap shifts. Then theres the question of staffing. When I was in Las Vegas, I found myself in an elevator with two store managers. I mentioned that I was going to meet Niccol and asked if they had one design request for him. After a few moments of contemplation, one of them simply said, More staff. Niccol gets it. The fundamental fact is were going to have more people in the stores servicing customers. Theres no other solution. Its a human-to-human business. Were going to put in the additional one, two, three people. Lets just say the roster of 18 becomes 20. These hirings should take place over the next year and will cost the company roughly $500 million. To pay for them, Bank of America analyst Senatore estimates that prices will need to go up by 3% or sales will need to grow as much as 5%. Starbucks plans to offset these investments with cost savings across the company. Starbucks has also announced that it will add an assistant manager role to every store, someone promoted from within, to take pressure off managers who often spend their days off handling tasks like scheduling. When this policy was announced on stage in Vegas, the crowd of managers roared. Niccol is so confident in his vision that hes already thinking ahead to when Starbucks could more than double its global footprint, reaching 100,000 stores. He details a plan based on three store sizeswhich he laughingly calls tall, grande, and venti. (Hes especially bullish on the tall store, a cozy, cost-effective espresso bar format with approximately 10 seats; the first will open in New York City.) Most will have drive-throughs. All will have seats. Niccol also denies that Starbucks is planning to sell off its China business, saying hes keeping the stores and would like to expand to 20,000 in the country with the right partner. This expansion seems a touch quixotic when Niccol concedes that only a handful of new stores can open in North America in 2026 with dozens following in 2027, given the lead time involved with obtaining permits. And he still needs to validate his model. Hes been pulling new levers to create urgency inside the company. In July, he offered his three top executives $6 million in stock grants if they hit cost-savings targets by the end of the yearand double that if they implement his in-store policies more quickly. He also mandated that corporate employees return to office, in Seattle or Toronto, for at least four days a week by October. (Following criticism of his initial plan to work remotely from Newport Beach, Niccol now works out of the Seattle office.) Still, theres one major question looming over Niccols Back to Starbucks plan. Even if Wall Street gives him the time to revive the third place, do customers still want it? Starbucks is just one of many quick-service restaurants that have ripped out seats and opened smaller stores over the past several years. The third place didnt just die at Starbucks; its disappeared across much of our world. In the era of DoorDash and, yes, mobile ordering, customers increasingly view fast-food restaurants as vehicles for food, not ambience. To Niccol, that presents an opportunity. Thats where the value of Starbucks comes to light, because its so much more than just a cup of coffee, he says. There arent many places like this anymore. Bringing hospitality back to the hospitality industry shouldnt be a novel idea. In 2025, thats all the more reason its essential.
Category:
E-Commerce
Years ago, not long after I had struck out on my own to start my own PR agency, I attended a party where I met a prominent journalist at a top-tier news outlet and we had a nice chat and exchanged contact details. For years we didnt cross paths professionally. But one day, she reached out. A source for a very sensitive story needed advice on how to handle the press scrutiny that she would soon be experiencing. There was no contract, no fee, no askjust someone in need. I offered my counsel freely. As a result, over the years, the reporter has continued to refer clients to us and has introduced us to other colleagues who could be helpful to our clients. And while those referrals have certainly been valuable, what they really represent is something more powerful: trust. Beyond the numbers, theres a hidden currency in business that yields richer returns than any deal you could close. It is the power of favors, freely given and authentically received. In a world where AI can automate emails, suggest connections, and streamline transactions, the human gestures that cannot be automated become even more valuable. A well-built favor bank can unlock long-term relationships, opportunities, and goodwill that no amount of money can buy. This may sound obvious but I was recently reminded of this idea when I was asked for a finder’s fee when a former colleague made introductions for possible new business. I was struck by how transactional that felt but also surprised because I have given and received introductions freely, keeping my favor bank philosophy in mind. In 2015, researchers at the University of South Californias Brain and Creativity Institute found that even the smallest acts of kindness (holding a door open with eye contact and a smile) result in thank you, or even an offer of help in return. This is the foundation of favor banking: That small, thoughtful gestures can influence how others respond, yielding larger returns over time, both personally and professionally. But with AI handling so much of our day-to-day efficiency, favor banking is what ensures those relationships dont get reduced to transactions. How to build a favor bank Lead with authenticity. Every year over the holidays, we send highly personalized gifts to our clients. For a client who couldnt stop raving about the wool products she bought while on vacation in Scotland, I sent a locally made wool throw blanket with a note explaining why I thought shed love it. For a journalist my team worked with whod just had a baby, I made some home-cooked Indian food, offered with no strings attached. Since then, weve gone from casual acquaintances to trusted collaborators. AI might remind you of a milestone or surface a clients preference, but only you can decide to turn that information into a gesture that feels real. Sometimes authenticity means going beyond small gestures to helping people through major life changes. As media jobs grow more precarious, Ive had more and more journalists turn to me when they lose onebecause the relationships weve built are rooted in trust. In one case, a former on-air anchor at a major network reached out, and I was able to connect her with a think-tank client where she became a fellowgiving her space to figure out next steps. Others have gone on to land roles at magazines or media outlets abroad. What mattered most wasnt the placement itself, but that they trusted me enough to reach out during a vulnerable moment. Share your expertiseOne of the most effective ways to build a favor bank is to offer your expertise freely, without viewing every interaction as a business opportunity. In a recent example, a scrappy young researcher availed us of his expertise with a complex project for a client in the defense-tech space, free of charge. By doing smart, efficient work for us, he knew hed begun cultivating an ecosystem of goodwill that will pay dividends. The same principle applies to founders using AI tools: scale what you share publicly with technology, but never stop cultivating one-to-one generosity privately. Be strategicbut never transactional Of course, offering free services too often can be unsustainable. But when done selectively, it builds trust and goodwill. The ability to distinguish between when to charge for your services and when to offer them for free is essential. Heres one example: Seeking technical advice, the board member of an under-resourced nonprofit approaches the CEO of a cutting-edge AI company. Typically, the CEO might charge for the servicebut theres often much greater long-term value in doing it for free. That board member is likely connected to other organizations and companies. By offering support without a transaction, the CEO is building goodwill that could lead to opportunities when the time comes to tap into that favor bank. People remember how you helped them, and will keep you top of mind. When (and how) to call in a favor When do I call in my favor? In fact, its less a question of when than of how: With directness, confidence, and a respectful acknowledgment of the relationship youve already cultivated. Skip the preamble. Dont hedge. Dont bury the ask in paragraphs of context. Just say what you needclearly, confidently, and with respect for the relationship youve built. Be specific: Can you introduce me to X? Would you be open to a 15-minute call to weigh in on Y? Would you consider reviewing this deck? It makes it easier for them to say yes, and it reinforces the value of the relationship. Youre not asking out of nowhereyoure tapping into a foundation of goodwill that youve already earned. Why founders cant afford not to do this Strapped for time and resources, start-up founders may think they cant afford to dole out favors for free. But thats the danger of short-term thinking: Amid the chaos of starting a business from scratch, you can easily lose sight of what a robust ecosystem of those you can trust and rely on when you need it the most. In other words, start-up founders cant afford not to practice favor banking. In the early days, its tempting to chase every dollar. But the strongest foundations are built on trust, not transactions. A founder who leans only on transactional efficiency risks missing the very relationships that make scaling possible. In a world thats increasingly driven by algorithms, AI, and transactional thinking, the human element is more important than ever. In fact, borrowing a phrase from a philosopher client, in the age of AI, I am bullish on humans. A favor bank isnt built overnight, but its returns are invaluable. Whether its helping a friend with a newborn or connecting a client with a valuable introduction, its those thoughtful gestures that ultimately make the biggest difference in lifeand business.
Category:
E-Commerce
Wheneven Sam Altman thinks there is an AI bubble, then there most likely is an AI bubble. But its even worse than that. There isnt just one AI bubble: there are three. First, AI is almost certainly in what economists call an asset bubble or a speculative bubble. As the name suggests, this is when asset prices soar well above their fundamental value. A classic example of this kind of bubble is the Dutch tulip mania of the 17th century, when speculators drove up the price of tulip bulbs to astronomical heights, convinced that there would always be someone willing to pay more than they had. As I write, Nvidia is trading at 50 times earnings, Tesla at an astounding 200 times, despite falling revenues, while the rest of the Magnificent 7 (Google, Amazon, Apple, Microsoft, and Meta) are enjoying significant boosts thanks to the bets they are taking on an AI-led future. The chances of this not being a bubble are between slim and noneand while Slim hasnt quite left town, hes booked his ticket and is packing his bags. Second, AI is also arguably in what we might call an infrastructure bubble, with huge amounts being invested in infrastructure without any certainty that it will be used at full capacity in the future. This happened multiple times in the later 1800s, as railroad investors built thousands of miles of unneeded track to serve future demand that never materialized. More recently, it happened in the late ’90s with the rollout of huge amount of fiber optic cable in anticipation of internet traffic demand that didnt turn up until decades later. Companies are pouring billions into GPUs, power systems, and cooling infrastructure, betting that demand will eventually justify the capacity. McKinsey analysts talk of a $7 trillion race to scale data centers for AI, and just eight projects in 2025 already represent commitments of over $1 trillion in AI infrastructure investment. Will this be like the railroad booms and busts of the late 1800s? It is impossible to say with any kind of certainty, but it is not unreasonable to think so. Third, AI is certainly in a hype bubble, which is where the promise claimed for a new technology exceeds reality, and the discussion around that technology becomes increasingly detached from likely future outcomes. Remember the hype around NFTs? That was a classic hype bubble. And AI has been in a similar moment for a while. All kinds of mediasocial, print, and webare filled with AI-related content, while AI boosterism has been the mood music of the corporate world for the last few years. Meanwhile, a recent MIT study reported that 95% of AI pilot projects fail to generate any returns at all. But what does all this mean for your organization? What should you be doing to respond? Do the Bubbles Matter? Bubbles are fine things in bottles of champagne, but in business contexts they are generally viewed as something to be avoided. So, when we see that AI is likely in three bubbles simultaneously, the immediate instinct may be to swerve urgently away from AI. I recommend resisting that instinct. Two of the three bubbles are largely irrelevant for most organizations and should simply be ignored. The speculative bubble is the result of a modern version of tulip maniainvestors bidding up the price of equities on the hope of future performance. The overheated valuations and crazy multiples are only problems for organizations that are involved in or directly exposed to the financial speculationand most organizations are not. A market crash may cause broader pain for the economy, but that is a potential environmental issue that all businesses will need to navigate. It should have little direct effect on a carefully planned AI implementation strategy. And as for the infrastructure bubble, well, if it turns out we really are building too much, the problem will be one of overcapacity, not overvaluation. For most organizations this is not only irrelevant, it may also lead to positive outcomes, because overcapacity will mean falling prices for those who want to use that infrastructure. This leaves the hype bubble, and this is where things get interesting. The hype bubble does have an important lesson for most organizations, but it isnt the one we might thinkeven if 95% of AI pilot projects fail, the issue here isnt that AI cant deliver value, but that many companies are approaching the technology in the wrong way. Dotcom Deja Vu Ive seen this before. I was in the dotcom boom (and bust) of the late ’90s and early 2000s. I saw Pets.com burn through $300 million before imploding, I saw the NASDAQ crashing by 78% and I read the articles by pundits who authoritatively declared that the internet was a fad. Yet during that same meltdown, Amazon was methodically building fulfillment centers and refining its recommendation algorithms. Google was quietly perfecting search. PayPal was solving payment friction. And thousands of companies were developing their first e-commerce capabilities, with greater or lesser degrees of success. The point is simple: a thing can be hyped beyond its actual capabilities while still being important (and to be fair to Sam Altman, he also makes this point in the piece quoted above). Just because AI is in a hype bubble does not mean that AI is fake news or that there isnt huge value to be extracted from it. The hype bubble simply means that some people are overexcited about AIit doesnt mean that there isnt something to be legitimately excited about. I made this argument in the dotcom era and I make it again now. What happened then will happen with AI. When valuations correctand they willthe same pattern will emerge: companies that focus on solving real problems with available technology will extract value before, during, and after the crash. In sum, companies with systematic approaches to extracting value from the technology will thrive. What becomes crucial, then, is your approach to capturing that value. So, how do you actually achieve that goal? The Value Creation Playbook The companies capturing real value follow three pillars of systematic implementation: Problem-First Architecture starts by mapping organizational friction points. Where do humans waste time on repetitive work? Where do information bottlenecks slow decisions? What processes consistently produce errors? Only after identifying these problems do successful companies consider AI solutions. Portfolio Balance means mixing time horizons and risk levels. Quick wins (one to three months) might include offthe-shelf tools for document processing. Strategic bets (3 to 12 months) could involve custom solutions for core business processes. Moonshots (more than 12 months) explore new business models. A retailer might implement an inventory chatbot this quarter while developing predictive analytics for next quarter and testing autonomous purchasing agents for next year. Holistic Integration connects AI initiatives to each other and to business strategy. Successful companies break down silos between IT, operations, and business units. They create feedback loops between projects. A manufacturing companys quality control AI feeds data to predictive maintenance AI, which informs supply chain AI. Each system makes the others smarter, creating compound value that isolated pilots never achieve. This is how you build value regardless of bubbles: systematically, purposefully, and starting today. Benefitting from Bubbles Far from being a threat, the AI bubble might be the best thing that could happen to pragmatic adopters. Consider what speculative excess delivers: billions in venture capital funding R&D youd never justify to your board; the worlds brightest minds abandoning stable careers to join AI startups, working on tools that youll eventually be able to use; infrastructure being built at a scale no rational actor would attempt, driving down future costs through overcapacity. While investors bet on which companies will dominate AI, you can cherry-pick proven tools at competitive prices. While speculators debate valuations, you will be implementing solutions with clear ROI. When the correction comes, youll also be able to benefit from fire-sale prices on enterprise tools, seasoned talent seeking stability, and battle-tested technologies that survived the shakeout. The dotcom bubble gave us broadband infrastructure and trained web developers. The AI bubble will leave behind GPU clusters and ML engineers. The smartest response isnt to avoid the bubble or try to time investments in it perfectly. It is to let others take the capital risk while you harvest the operational benefits. The bubble isnt your enemy. If you play your cards strategically, it can be a major benefactor. A valuable distraction Perhaps the greatest gift of the bubble discourse is the distraction it provides. While commentators debate whether Nvidia is overvalued and conferences overflow with AI bubble panels, something interesting happens: the noise creates perfect cover for serious operators to build lasting value. This psychological dynamic creates genuine competitive advantage. The bubble debate gives skeptics intellectual permission to waitafter all, why pursue that interesting AI project if the whole AI thing will inevitably crash? Meanwhile, companies quietly pursuing systematic AI implementation face less competition for talent, less pressure on timelines, and less scrutiny of their initiatives. The louder the bubble talk, the more space opens for those willing to take a methodical approach to building value.
Category:
E-Commerce
All news |
||||||||||||||||||
|
||||||||||||||||||