Be­zos Un­bou

Forbes - - Bezos Unbou -

nlike Amer­ica’s other tech gi­ants, Ama­zon doesn’t have a tra­di­tional cam­pus. e 45,000 or so em­ploy­ees and ex­ec­u­tives in Seat­tle, out of 575,000 world­wide, fan across nu­mer­ous high-rises down­town and in the South Lake Union neigh­bor­hood. Ama­zon’s “head­quar­ters” de­faults to where Je Be­zos, the com­pany’s founder and CEO, hap­pens to be, cur­rently Day 1 Tower. Its name comes from a per­pet­ual Be­zos maxim: that, rel­a­tively, we’re still at “day one” of the in­ter­net—and, by ex­ten­sion, that Ama­zon is just get­ting started.

at’s get­ting harder to say with a straight face, with sales, pro ts and the stock price all soar­ing, the lat­ter up 270% over three years and 103% in the past 12 months. Ama­zon is clos­ing in on Ap­ple to be­come the world’s most valu­able com­pany, and Be­zos, whose per­sonal net worth ap­proaches

$160 bil­lion, has in the process be­come the planet’s rich­est per­son, by far.

UNev­er­the­less, Be­zos talks about Ama­zon like it’s a giddy startup that just closed its Se­ries A. “For all prac­ti­cal pur­poses, the mar­ket size is un­con­strained,” says Be­zos, his rolled-up sleeves show­ing o Pop­eye-like fore­arms, the prod­uct of midlife weight train­ing that has pro­duced buzzed-about re­sults for the 54-year-old. His growth ra­tio­nale comes from a “su­per-lucky” con uence: e re­tail mar­ket, Ama­zon’s orig­i­nal quarry, is “many tril­lions,” as is, he says, the cloud mar­ket that Ama­zon Web Ser vices (AWS) pi­o­neered. “ ere are di er­ent busi­nesses where the mar­ket is limited,” adds the man whose com­pany should hit $210 bil­lion in rev­enue this year. “But we just don’t have that is­sue.”

If Je Be­zos is al­ready the world’s most feared busi­nessper­son, the prospect of him “un­con­strained” should sober ev­ery cor­po­rate leader. Yes, he’s ruth­less and a master of the long game, but Be­zos’ great­est strength, borne out over the past few years, has been his abil­ity to shape-shi Ama­zon into ad­ja­cent busi­nesses—some of which were ad­ja­cent only in ret­ro­spect—on a mas­sive scale. It’s quanti able: Forbes has been rank­ing in­no­va­tive com­pa­nies for eight years, and we re­cently worked with a trio of man­age­ment pro­fes­sors to try to de­ter­mine the coun­try’s most in­no­va­tive busi­ness lead­ers ( see p. 84). e four-prong method­ol­ogy—in­cor­po­rat­ing pub­lic rep­u­ta­tion and in uence, value cre­ation and the pre­mium that in­vestors as­sign to the chief ex­ec­u­tive— places Be­zos squarely at the top.

“What Je Be­zos has done and is likely to do is per­haps the most re­mark­able achieve­ment I’ve seen,” War­ren Buf­fett told me last year, a er I asked him, open-ended, to name the most im­pres­sive busi­ness mind in his al­most eight decades of mar­ket-watch­ing. “Be­cause he’s taken two very ma­jor in­dus­tries, and si- mul­ta­ne­ously,mul­ta­ne­ously, and sort of un­der the nose of com­peti­tors, he’s be­come in e ect the leader and is rede ning them and suc­ceed­ing at re­ally big busi­nesses.”

ough Be­zos and Bu ett were re­fer­ring to re­tail and the cloud, Be­zos is ac­tu­ally un­con­strained in far more ways. First, thanks to AWS, the com­pany fa­mous for em­pha­siz­ing growth over prof­itabil­ity is nally spit­ting out bil­lions— and Be­zos has the mar­ket cred­i­bil­ity to rein­vest it in pretty much any way he wants. Sec­ond, the scale that Ama­zon needs for growth prac­ti­cally de­mands ag­gres­sive­ness. And, nally, by dom­i­nat­ing re­tail and dig­i­tal busi­ness services, both of which touch al­most ev­ery other in­dus­try, he’s now po­si­tioned to move ad­ja­cently into just about any busi­ness where he nds added value. He’s play­ing in the multi­bil­lions in at least four mar­kets—health­care, en­ter­tain­ment, con­sumer elec­tron­ics and ad­ver­tis­ing—that con­sti­tute many of the com­pa­nies not al­ready terri ed of Ama­zon. It’s no co­in­ci­dence that each of those four ei­ther hits or ap­proaches the “tril­lions” po­ten­tial Be­zos al­luded to.

While his pi­o­neer­ing peers of the

rst dot-com era em­braced and pop­u­lar­ized the “open ki­mono,” Be­zos has al­ways viewed stealth­i­ness as an as­set, mask­ing new ini­tia­tives in­side larger ex­pen­di­tures and feign­ing dis­in­ter­est in

“WHAT BE­ZOS HAS DONE IS PER­HAPS THE MOST RE­MARK­ABLE ACHIEVE­MENT

I’VE SEEN.”

bur­geon­ing fa­vorites. As Be­zos’ pub­lic pro le has ex­panded, pub­lic ut­ter­ances and in­ter­views (de­spite his own­er­ship of the Wash­ing­ton Post) have be­come in­creas­ingly rare. Be­zos re­fuses to dis­cuss Don­ald Trump, who has taken to beat­ing up on him and the Twit­pos­ton­twitPost on Twit­ter, but he clearly un­der­stands he has a tar­get on his back. When asked, as the head of an as­cen­dant ad­ver­tis­ing com­pany, whether he took any lessons from Face­book’s tra­vails last year, his an­swer was suc­cinct, po­lit­i­cal and in­con­ceiv­able. “No,” says this ad­vo­cate for cor­po­rate learn­ing, paus­ing for a few sec­onds to un­der­score that he wasn’t go­ing there. Ditto ques­tions about be­com­ing a data com­pany. “I’ve never re­ally thought of Ama­zon in that way,” says the man who runs as data-driven a com­pany as any, be­fore re­vert­ing to his stump speech. When it’s sug­gested that it’s at least a tool, Be­zos quickly in­ter­jects: “One of many tools.”

Nev­er­the­less, dur­ing the morn­ing he spent with Forbes out­lin­ing how he chan­nels in­no­va­tion and chooses where to ex­pand, a road map for Ama­zon’s fu­ture emerged. Given Ama­zon’s size, it moves both ver­ti­cally and hor­i­zon­tally, each di­rec­tion por­tend­ing a lot more dis­rup­tion. Even ve years ago, Be­zos seemed con­tent merely to try to sell ev­ery­thing to ev­ery­body, be­com­ing the bane mostly of re­tail­ers and whole­salers. But this master in­no­va­tion artist now has the ul­ti­mate pal­ette: any in­dus­try he chooses.

FOR THIS UN­CON­STRAINED era, the most important word at Ama­zon is yes. Be­zos ex­plains, cor­rectly, the tra­di­tional cor­po­rate hi­er­ar­chy: “Let’s say a ju­nior ex­ec­u­tive comes up with a new idea that they want to try. ey have to con­vince their boss, their boss’s boss, their boss’s boss’s boss and so on—any ‘no’ in that chain can kill the whole idea.”

at’s why nim­ble star­tups so eas­ily slaugh­ter hide­bound di­nosaurs: Even if 19 venture cap­i­tal­ists say no, it just takes a 20th to say yes to get a dis­rup­tive idea into busi­ness.

Ac­cord­ingly, Be­zos has struc­tured Ama­zon around what he calls “mul­ti­ple paths to yes,” par­tic­u­larly re­gard­ing “two-way doors”: de­ci­sions that are o en based on in­cre­men­tal im­prove­ments and can be re­versed if they prove un­wise. Hun­dreds of ex­ec­u­tives can green­light an idea, which em­ploy­ees can shop around in­ter­nally. “He knows and we know that you can’t in­vent or ex­per­i­ment with­out some fail­ure,” says Je Wilke, the long­time Be­zos lieu­tenant who runs Ama­zon’s con­sumer and re­tail op­er­a­tions. “ ose we sort of cel­e­brate. In fact, we want them to oc­cur all over the place. Je doesn’t need to re­view those. I don’t need to re­view those.”

But re­gard­ing the larger ideas and ver­ti­cals—a.k.a. “one-way doors”—that change the di­rec­tion of the com­pany, Be­zos prides him­self on play­ing “chief slow­down o cer.” He’s look­ing for three things. First, orig­i­nal­ity. “We have to have a di er­en­ti­ated idea. It can’t be a ‘me too’ o er­ing,” he says. Sec­ond, scale. “We’re gi ed with some very large busi­nesses we’ve built over time, and we can’t a ord to put our en­er­gies into some­thing that if it works, it’s still go­ing to be small.” And, nally, a Sil­i­con Val­ley-worthy ROI. “Even at sub­stan­tial scale, it has to have good re­turns on cap­i­tal.”

Ul­ti­mately, the ideas that hit that troika, Be­zos says, em­anate from one of two mod­els. Ei­ther by look­ing back­ward at cus­tomer needs—as in we’ve no­ticed peo­ple act a cer­tain way, so let’s try to serve them with a prod­uct. Or peer­ing for­ward—we know how to do some­thing valu­able,

so let’s nd cus­tomers.

e Ama­zon jug­ger­naut ul­ti­mately stems from the lat­ter. Orig­i­nally a niche player, Be­zos could have eas­ily fo­cused just on be­com­ing the world’s dig­i­tal book­store, the way cra s be­long to Etsy and shoes to Zap­pos (now owned by Ama­zon, nat­u­rally). But in mas­ter­ing sell­ing books, Be­zos saw that he could use those tools—from in­ven­tory man­age­ment to rec­om­men­da­tion en­gines— to move ad­ja­cently: rst into mu­sic and DVDS, then toys and elec­tron­ics, then pretty much any­thing that can be sold at re­tail. And he again lever­aged that knowl­edge au­da­ciously (and suc­cess­fully) in open­ing up Ama­zon as a plat­form for in­de­pen­dent sell­ers, who were for­merly com­peti­tors—ce­ment­ing his legacy as a trans­for­ma­tive re­tailer, in the pan­theon with the likes of Sam Wal­ton, Aaron Mont­gomery Ward and Sears Roe­buck’s Julius Rosen­wald.

is is where the story should have ended. But as the Western world’s dom­i­nant on­line re­tailer, Ama­zon was also solv­ing huge tech­no­log­i­cal and lo­gis­ti­cal prob­lems, and rather than view those skills merely as ac­cre­tive to the core busi­ness, Be­zos saw them as busi­nesses unto them­selves. An in­ter­nal need for part-time de­vel­op­ers led to Mechani- cal Turk, one of the world’s rst crowd­sourc­ing gig mar­ket­places. Build­ing a stag­ger­ingly e cient de­liv­ery in­fra­struc­ture led to the Ful ll­ment by Ama­zon ser­vice, and mas­ter­ing how to take money for each pur­chase led to Ama­zon Pay. Most important, as Ama­zon be­gan build­ing enor­mous ca­pa­bil­i­ties to store its data in the cloud, Be­zos gured out that other busi­nesses might want to store their data there as well. In 2017, AWS had $17.5 bil­lion in rev­enue.

Even those cus­tomer-driven con­cepts cre­ate skills-based div­i­dends. Take the Kin­dle reader, Ama­zon’s rst foray into hard­ware, back in 2007. Wilke, who came from Al­liedsig­nal, re­mem-

bers protesting to the board. “I spoke up and said, ‘I don’t agree. I think we’re likely to miss our planned de­liv­ery date. Our yields are gonna be too low. We’re gonna un­der­pro­duce. We’re gonna frus­trate cus­tomers. Hard­ware’s hard. We’re a so ware com­pany.’

“Je said, ‘Well, I’m will­ing to con­cede that all those things will happen, and I still think that the right vi­sion for our com­pany is to be re­ally good at build­ing hard­ware, so we need to get started learn­ing.” And so they did. Kin­dle is a ro­man­ti­cized prod­uct at Ama­zon, both be­cause of the hard­ware break­through and be­cause it harks back to the com­pany’s roots in books. It didn’t trans­form the compa- ny, though, and other hard­ware fail­ures, like the dis­as­trous Fire smart­phone, fol­lowed. But Be­zos’ de­ci­sion also even­tu­ally led to the Ama­zon Echo smart speaker, a true game-changer.

“We have a lot of hard­ware ex­pe­ri­ence to­day, but back then we didn’t have those skills,” says Be­zos, laugh­ing. “You have to be pa­tient. It’s not go­ing to be just time to learn a skill. It might take time for the thing to re­ally ower.” In other words, if you study the skills Ama­zon is cur­rently learn­ing, you’ll have a fair idea of what it will soon be sell­ing.

And right now Be­zos is learn­ing about health­care. It’s Amer­ica’s big­gest in­dus­try—18% of Gdp—and one of its most ine cient. Last year, Be­zos, along with Bu ett and Jpmor­gan Chase CEO Jamie Di­mon, an­nounced that their three com­pa­nies would pool their ef­forts, hir­ing high-pro le CEO Atul Gawande to lead a non­pro t ini­tia­tive to de­liver bet­ter care at a lower cost for their own em­ploy­ees, with the idea of cre­at­ing a scal­able, clon­able model. It’s no small thing: ose three com­pa­nies em­ploy 1.2 mil­lion. Add in their de­pen­dents, and it’s like run­ning a pilot pro­ject for ev­ery­one in Ore­gon or Con­necti­cut.

Be­zos is adamant about Ama­zon’s in­ten­tions here. “ is is a non­pro t ini­tia­tive, as you guys know. It’s very dif­fer­ent,” he in­ter­jects, be­fore even a full

ques­tion can be asked about it. Bu ett, for his part, con­curs, hav­ing ex­plained to me a few months ago: “We were del­uged by peo­ple a er the an­nounce­ment who said, ‘We want to join in.’ And we said, ‘You don’t have to join in. Steal ev­ery­thing we get, if we get any­thing.’ ” e “if we get any­thing” is key. “Like Colum­bus leav­ing, we don’t know where the hell we’re go­ing ex­actly,” Bu ett added. “But we hope there’s an­other con­ti­nent out there and we don’t go o a shelf at some point.”

But even if they do sail o the Earth, Be­zos wins, as Ama­zon bur­nishes its skills ac­count­ing for around one h of the do­mes­tic econ­omy. While any break­throughs de­vel­oped with Bu ett and Di­mon “would still be in­side that non­pro t en­tity,” Be­zos says, “each of the com­pa­nies can pur­sue their own ini­tia­tives.” Be­zos has al­ready started. In June, Ama­zon agreed to pay al­most $1 bil­lion for Pill­pack, a startup that de­liv­ers prepack­aged daily pre­scrip­tion en­velopes. It’s ev­ery­thing that Ama­zon is good at: ful ll­ment, cus­tomiza­tion and de­pend­abil­ity. And it’s an­other toe in the health­care pool.

Be­zos is also go­ing to school in ad­ver­tis­ing. Ama­zon’s most re­cent quarterly per­for­mance re­vealed a star­tling num­ber: Ama­zon is on pace to ex­ceed $8 bil­lion in ad­ver­tis­ing rev­enue this year— roughly dou­ble last year’s to­tal. And why not? Google might know what you’re in­ter­ested in buy­ing, Face­book might be able to de­duce what you’d be in­clined to buy, but Ama­zon knows what you’ve ac­tu­ally bought, or even whether you showed in­tent to buy.

at raises all sorts of is­sues. Be­zos talks ex­u­ber­antly about his con­sumer ob­ses­sion, but there are few cus­tomers who want to be tar­geted with more ads. Be­zos says the trust that Ama­zon has built with its cus­tomers will en­sure the com­pany doesn’t cross a line—and that cus­tomers, in turn, will give him the bene t of the doubt. “It’s very valu­able, and so you would never do any­thing to jeop­ar­dize it,” he says. “It’s what al­lows you to ex­pand the busi­ness.” If Be­zos can walk that line, it’s easy to en­vi­sion the Face­book-google ad­ver­tis­ing “du­op­oly” gain­ing a third ma­jor en­trant. WHEN YOU WALK AMONG THE Seat­tle high-rises, the most in­ter­est­ing build­ing in Ama­zon’s neigh­bor­hood—with apolo­gies to a stun­ning new bio­sphere de­signed to serve as some­thing of a com­pany com­mons—is a food store that pokes out from the bot­tom of Day 1 Tower. Ama­zon’s $13 bil­lion Whole Foods pur­chase last year is only the sec­ond-most-in­ter­est­ing gro­cery ini­tia­tive at Ama­zon; the rst is Ama­zon Go, an 1,800-square-foot ex­per­i­ment in fric­tion­less brick-and-mor­tar pur­chases that opened in Jan­uary.

Go, a food store, might be the most Ama­zo­nian thing at Ama­zon. Be­zos has al­ways stressed fru­gal­ity, and here are his em­ploy­ees (and a few Seat­tle res­i­dents and tourists) shop­ping for their lunches, buy­ing them from Ama­zon and in doing so also pro­vid­ing the com­pany reams of data to hone its skills. More important, Go demon­strates what’s pos­si­ble when di er­ent as­pects of Ama­zon’s myr­iad op­er­a­tions come to­gether. Go en­com­passes knowl­edge gained from pur­chases at Whole Foods; from Ama­zon’s in­creas­ingly so­phis­ti­cated al­go­rith­mic and hard­ware ca­pa­bil­i­ties, in the form of AI, camera and sen­sor tech­nolo­gies, which com­bine to gure out what’s taken o the shelf (and put back) and who did the tak­ing; and from Ama­zon Pay, which seam­lessly han­dles the trans­ac­tion once the com­pany’s app regis­ters the in­puts. Grab­bing a cho­co­late milk here on the way out feels like shopli ing—no check­out lines, no scan­ning, no swip­ing. If health­care and ad­ver­tis­ing rep­re­sent huge ex­pan­sion ver­ti­cals on Ama­zon’s de facto ver­ti­cal cam­pus, Go brings to­gether what’s pos­si­ble hor­i­zon­tally when Ama­zon puts all the pieces to­gether.

e most important as­pect of it all is Prime. At its core, it’s a mar­ket­ing tool, a way to en­cour­age buy­ing—and gen­er­ate re­cur­ring sub­scrip­tion rev­enue (an es­ti­mated $10 bil­lion in 2017)—in a sim­i­lar vein to warehouse clubs like Costco and BJ’S Whole­sale. But as it grows, it’s surged into a way to shower its 100 mil­lion-plus sub­scribers with perks and priv­i­leges.

And those perks cre­ate end­less new ad­ja­cen­cies and busi­ness lines. Prime ex­plains why Ama­zon has started to en­croach on Net ix and is ex­pected to spend $5 bil­lion on pro­gram­ming this year, in­clud­ing the highly touted Marvelous Mrs. Maisel. In just three years, “Prime Day”—an an­nual 36 hours of special deals just for mem­bers, with par­tic­i­pants spend­ing bil­lions on more than 100 mil­lion prod­ucts this July—has be­come a shop­ping hol­i­day with a ma­nia sur­passed only by Black Fri­day and Cyber Mon­day. Ac­cess to Prime is how Ama­zon gets its out­side re­tail­ers to pay more for Ful ll­ment by Ama­zon.

Prime also un­der­pins Ama­zon’s brick-and-mor­tar strat­egy. Prime’s same-day de­liv­er­ies and pick­ups re­quire more phys­i­cal beach­heads, which in turn help un­der­write ex­pan­sion into all

“TRUST I S WHAT AL­LOWS

YOU TO EX­PAND THE BUSI­NESS,” BE­ZOS SAYS.

sorts of elds Ama­zon could never have justi ed, such as food, the kind of per­ish­able prod­uct that doesn’t t into the clas­sic Ama­zon model. A er years as a book­store killer, the com­pany now has 16 per­ma­nent Ama­zon Books lo­ca­tions across 11 states. And Ama­zon re­cently opened a sec­ond Go in Seat­tle. “ ey have to have some­thing that’s special and new,” Be­zos says. “And that’s what you’re see­ing if you’re pay­ing at­ten­tion to our phys­i­cal-stores strat­egy.”

Prime, in short, has be­come Ama­zon’s cen­tral ner­vous sys­tem, con­nect­ing ev­ery­thing in the com­pany—and giv­ing Ama­zon a path to ex­pand into new mar­kets while also juic­ing its core re­tail busi­ness—with­out re­ally be­ing it’s own thing. “It can’t be a stand-alone busi­ness be­cause it’s com­pletely tied in to our con­sumer of­fer­ing,” Be­zos says.

Sim­i­larly, as seen with Go, arti cial in­tel­li­gence will in­creas­ingly con­nect what pre­vi­ously seemed like dis­parate prod­uct lines, from cloud to re­tail. Take the wildly suc­cess­ful Echo, which has made Je Wilke’s ex­is­ten­tial ques­tion about Ama­zon— is it a hard­ware or a so ware com­pany?— moot. It’s an e ec­tive piece of hard­ware em­pow­ered by Ai-driven so ware, which drives Ama­zon’s re­tail sales, lever­ages Ama­zon’s con­tent and so on. “ e most in­ter­est­ing thing about ma­chine learn­ing as op­posed to a lot of other tech­nolo­gies is just how hor­i­zon­tal it’s go­ing to be,” Be­zos says. “ ere’s not a sin­gle cat­e­gory of busi­ness or gov­ern­ment or any­thing, re­ally, that can’t im­prove it­self.”

Even the much bal­ly­hooed “HQ2” sec­ond head­quar­ters bake-o , which Ama­zon in­sid­ers con­tinue to say will be nal­ized by year’s end, speaks to Ama­zon’s hor­i­zon­tal out­look. ( Forbes is bet­ting on Wash­ing­ton, D.C., where Be­zos al­ready has a base and which has the tal­ent, lo­gis­tics and ac­cess to in uencers; by short­list­ing both the Vir­ginia and the Maryland sub­urbs, as well as the Dis­trict of Columbia, Be­zos can also play three gov­ern­ments within one metropoli­tan area against each other for the best sweet­heart deal.) In­creas­ingly, Be­zos says, ge­og­ra­phy at Ama­zon is be­com­ing ir­rel­e­vant. When sep­a­rate teams need to work to­gether, “what you want is those two groups to have in­fre­quent meet­ings and set a road map of the fu­ture.” He adds:

“If you or­ga­nize cor­rectly, peo­ple do not have to be in the same build­ing or the same city or even the same time zone, be­cause you can work o a road map.”

And make no mis­take, that road map is “pretty much all” that Be­zos works on, del­e­gat­ing the day-to-day busi­ness op­er­a­tions. Long-term think­ing has been Ama­zon’s hall­mark, and the cash spigot from AWS em­pow­ers Be­zos to keep run­ning the re­tail jug­ger­naut with near-zero mar­gins and unas­sail­able e cien­cies, as well as in­vest in far- ung ar­eas. “I very rarely get pulled into the to­day,” Be­zos says. “I get to work two or three years into the fu­ture, and most of my lead­er­ship team has the same setup.”

“Friends con­grat­u­late me a er a quarterly-earn­ings an­nounce­ment and say, ‘Good job, great quar­ter,’ and I’ll say, ‘ ank you, but that quar­ter was baked three years ago.’ I’m work­ing on a quar­ter that’ll happen in 2021 right now.”

It’s that mind­set that has com­pa­nies in dozens of in­dus­tries run­ning scared. “I come up with ideas. We could sit here with an idea, and I could ll this white­board in an hour with 100 ideas,” Be­zos says. “If I have a week with no brain­storm­ing meet­ings, I com­plain to my of ce, like ‘Come on, guys, help me here.’ ” Cor­po­rate Amer­ica, take note—ei­ther in­no­vate or Je Be­zos will do it for you.

F

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.