Bloomberg Businessweek - USA (2019-07-22)

Bloomberg Businessweek July 22, 2019

47

customers.Finallytheysettledonusingcomputervision,a
relativelynewtechnologythatallowsdigitalcamerasandcom-
puterstoidentifyitemsbytheirvisualappearancealone,with-
outanyspecialtrackingchipsorcodes.
Kumarhiredcomputervisionandmachinelearningscien-
tists,oftenfromotherpartsofAmazon,withouttellingthem
exactlywhatthey’dbeworkingonbeforehand.Hesetdead-
lineafterdeadline,usingupcomingpresentationstoBezos
orKesseltomotivatethem.Engineersputin 70 to 80 hours
a week,answeringemailsandwritingAmazon’sclassicsix-
pagedocuments,narrativememosthatoutlinea proposal,
inanyillusoryfreetimetheyhadonnightsandweekends.
“Wewerealllivingina cave,”saysone.
Atfirst,theIHMteamenvisionedlarge-scalestoresof
about30,000squarefeet,roughlythesizeofa suburban
supermarket. But after a few months, the group decided such
a megamarket was overly ambitious and cut the size of the
proposed store in half.
Puerini’s group created the first models of stores using
kids’ blocks, bookshelves, and other items lying around the
office. As the project neared a hoped-for introduction in mid-
2015, the company retrofitted the warehouse in south Seattle
for a mock-up to show to Bezos. For its first actual store, it
also anonymously leased the ground floor of a new luxury
apartment building in Seattle’s wealthy Capitol Hill neigh-
borhood. Permits filed with the city included plans for large
produce and dairy coolers and an on-site kitchen for prepa-
ration of fresh foods.

ButthenBezosvisitedthemock-upwiththecheesemongers

andputa halttotheprogress,bettingcustomerswouldbe

drawntoa morestreamlinedexperience—thephysicalequiv-

alentofthecompany’sfamousone-clickordering—even if

it lacked the personal touches of a farmers market or bou-

tique butcher shop.

Kessel convened a team meeting after the Bezos demo

andbrokethenews:Theywerepivotingtoconvenience

stores.Someengineerswererelievedthattheywerereduc-

ingcomplexity by eliminating items of varying weights, such

as produce and meat. Others were crestfallen and left the

project, either exhausted from the nonstop pace of work or

disappointed by the scaled-down vision. For the next three

years, the storefront on Capitol Hill would sit abandoned, in

the heart of one of Seattle’s most well-trafficked neighbor-

hoods, its windows mysteriously covered in brown paper.

Bezos and Kessel were growing impatient. So in March 2015,

while Puerini and Kumar reworked their concept, they

formed a separate group under Kessel to open bookstores.

Books were the opposite of food—nonperishable, consistently

priced, easy to stock, and, of course, the product category

Amazon had the most history with. And because people tend

to shop bookstores at a more leisurely pace, it wouldn’t be

necessary to try to displace cashiers with technology.

That fall, as the company prepared its first Amazon Books

outlet in an upscale mall in Seattle, speculation over how the

company would enter physical retail was so feverish that a

COOKIES CRACKERS SANDWICHES WRAPS SUSHI DELI SIDES SALADS COLDDRINKS

CANDY YOGURT COFFEE BREAKFAST DRINKSCOLD DAIRY M E ATS ICECREAM FROZENMEALS

BAGS

MEALKITS H E AT& E AT DESSERT PRODUCE DELISNACKS COLDDRINKS

POPCORN SNACKS BREADSSPREADS& PANTRY DRINKS QUICKPICKS

BAKERY

BAGS ENERGYBARS BREAKFAST

ACMESALAD,WHITEFISHDEAN’S

FRENCHDIP,OLLIONIONGENOA

SALAMI,HUMMUS,SABRAREADY

SETSALMON!

EPICUREDSTRAWBERRY

ROSEPUDDING,CHIAPROPERSEED

FOODBREAKFASTCHIQUITO

BURRITO

PREPRERIBEYENEWYORKSTEAK,

STRIPBURGERSTEAK,PATTIES,PRE

OSCARORIGINALMEYERCENTER

CUTBACON

STOUFFER’S FIVE CHEESE

LASAGNA, AMY’S BEANS & RICE

BURRITO, MARIE CALLENDER’S

CHICKEN POT PIE

HERSHEY’SCHOCOLATEMILK

BAR,ORGANICYUMEARTHLICORICE,

SOURALTOIDSPATCHMINTSKIDS,

③

CHIPS

①Customersscan
their phone and receive
a digital identifier the
software uses to track
their visit.

②Everyspotis covered

bymultiplecamerasto

makesurea customeris

trackedfromoneframe

tothenext.

③ The software decides

which items have been

selected using camera

depth sensors and weight

sensors in the shelves.

④If thesoftwareloses

track of someone, it uses

appearance, trajectory,

and other inputs to figure

out who’s who.

⑤In-shelfcamerasverify

when a customer puts

back an item, and trigger

notifications for staff to

restock things.

⑥ As customers leave, the
software bills their credit
card. If there’s a problem,
it sends the order to a
human to review.

DUNCAN HINES FUNTASTIC

CONFETTI CAKE FOR ONE,

CAMPBELL’S CHUNKY CHILI

WITH BEANS

PROPER FOOD CHICKEN & HUMMUS

ON CIABATTA, TUNA SALAD

ON CIABATTA, CHICKEN SALAD

ON CIABATTA

KASHIRAISIN FLAX OATMEAL

COOKIES, NUTTER BUTTER BITES,

NUTELLA & GO! HAZELNUT SPREAD

& BREADSTICKS

PEPTO-BISMOL, GLIDE FLOSS PICKS,

EXTRA STRENGTH TYLENOL

CAPSULES, CLASSIC CHERRY

CHAPSTICK

CAJUN-STYLE POTATO SALAD,

CHICKEN CAESAR SALAD, BROCCOLI

CRUNCH POWER BOWL WITH

CHICKEN

READY, SET, GO