EM­BRAC­ING DIG­I­TAL DIS­RUP­TION

Dig­i­tal trans­for­ma­tion is pos­si­ble for ev­ery or­ga­ni­za­tion. And it’s be­com­ing more nec­es­sary by the day.

Rotman Management Magazine - - FRONT PAGE - By Maxwell Wes­sel

these days — but the DIG­I­TAL TRANS­FOR­MA­TION IS A HUGE BUZZ­WORD fact is, it’s not a new topic. More than 30 years ago, Har­vard Busi­ness School Pro­fes­sor Michael Porter wrote about how in­for­ma­tion tech­nol­ogy (IT) would trans­form com­pet­i­tive ad­van­tage. Pro­fes­sor Porter saw what was hap­pen­ing with the de­moc­ra­ti­za­tion of IT, and pre­dicted that this new ‘IT layer’ of ev­ery or­ga­ni­za­tion would trans­form busi­ness, en­abling in­for­ma­tion flow to scale in ways that were not pos­si­ble be­fore.

Pro­fes­sor Porter was right: Thanks to tech­nol­ogy, to­day’s com­pa­nies can reach around the globe to find cus­tomers in new mar­kets, and IT has be­come the ‘bind­ing glue’ un­der­pin­ning all sorts of tra­di­tional-style in­dus­trial growth.

How­ever, when we talk about dig­i­tal trans­for­ma­tion to­day, there is much more to the story. To­day, we have Amazon, the big­gest re­tailer by mar­ket cap­i­tal­iza­tion, with only a hand­ful of ex­per­i­men­tal stores; Uber, the big­gest liv­ery ser­vice in the world by fleet and by coun­tries reached, with no ve­hi­cle own­er­ship; and Airbnb, the big­gest hos­pi­tal­ity ser­vice by num­ber of lo­ca­tions, with no owned prop­er­ties. Clearly, the way com­pa­nies are em­brac­ing dig­i­tal to cre­ate busi­ness mod­els and at­tack the mar­ket- place is very dif­fer­ent from the way we talked about dig­i­tal trans­for­ma­tion over the last 30 years.

In this ar­ti­cle I will dis­cuss three key as­pects of dig­i­tal trans­for­ma­tion: What used to mat­ter; what mat­ters to­day; and the key chal­lenges or­ga­ni­za­tions face.

What Used to Mat­ter?

To de­scribe what used to mat­ter, I will use a sim­ple ex­am­ple: Soap — the pro­to­typ­i­cal prod­uct of the in­dus­trial process. At the turn of the 20th cen­tury, Proc­ter & Gam­ble was build­ing a mas­sive busi­ness — and it was not based on tech­no­log­i­cal in­no­va­tion. Sure, Ivory was ‘the soap that floats’; but the most in­ter­est­ing thing P&G did was to mas­ter a set of func­tions that were largely un­avail­able to the av­er­age com­pany at the time: It mas­tered the sourc­ing of sup­ply to scale; dis­tri­bu­tion across the na­tion and ul­ti­mately the globe; and, most im­por­tantly, the process by which a com­pany can mar­ket to and cre­ate a cus­tomer base from a dis­tance.

When P&G first tack­led the art of print ad­ver­tis­ing, there was no prece­dent for a com­pany to reach around the world to

cre­ate a crit­i­cal mass of cus­tomers. Prior to that, peo­ple had to visit their lo­cal gen­eral store and have a piece of soap carved from a mas­sive bar for in­di­vid­ual use. P&G suc­ceeded due to its mas­tery of process at scale — some­thing that had never be­fore been achieved.

When the Dupont com­pany came along, it too was able to man­age a con­glom­er­ate of or­ga­ni­za­tions, pri­mar­ily be­cause it de­vel­oped met­rics to al­low for the re­mote au­dit of its pro­cesses. At the turn of the 20th cen­tury, the best com­pa­nies in the world com­peted with the knowl­edge of what would al­low them to per­form — even when their man­agers weren’t around. It-driven met­rics al­lowed for that, and it was trans­for­ma­tional.

In short, the 20th cen­tury was all about in­dus­trial scale. That is what mat­tered, and great global busi­nesses were built on the back of a mas­tery of those pro­cesses.

How times have changed. Just look at a com­pany like Borders — an ex­am­ple of a busi­ness that had a per­fect mas­tery of in­dus­trial scale, Borders had re­tail lo­ca­tions across the coun­try; a keen un­der­stand­ing of how to price and man­age its in­ven­tory; and re­la­tion­ships with dis­trib­u­tors and lo­gis­tics com­pa­nies that al­lowed it to ac­tu­ally move phys­i­cal ma­te­ri­als like no one else in the in­dus­try. But none of that stopped Amazon from up-end­ing its busi­ness.

Put sim­ply, all of the things that al­lowed com­pa­nies to build mas­sive, valu­able in­dus­trial-scale busi­nesses in the 20th cen­tury are now ‘for rent’. Ap­ple is the largest con­sumer elec­tron­ics com­pany in the world. The­o­ret­i­cally, as such, it should have some man­u­fac­tur­ing scale to pro­vide the types of re­turns and scale ad­van­tages that peo­ple cov­eted through­out the 20th cen­tury. In­stead what does Ap­ple do? It rents man­u­fac­tur­ing ca­pac­ity from Fox­conn.

The fact is, if we can rent scale from Fox­conn, com­mu­ni­ca­tion in­fra­struc­ture from Tril­lium, com­put­ing ca­pac­ity from Amazon and lo­gis­tics ca­pac­ity from Fedex or UPS — we have ef­fec­tively de­com­posed what it means to be an in­dus­trial busi­ness.

What Mat­ters Now?

So, what mat­ters to­day? The an­swer, I would ar­gue, is data. IBM has a tagline, ‘Data is the new oil’ — mean­ing that the sub­stance by which we pow­ered in­dus­trial change will power dig­i­tal trans­for­ma­tion. The point is not that data is valu­able in and of it­self. The only peo­ple who profit from telling you that are those who sell data stor­age in­fra­struc­ture. Data is specif­i­cally valu­able to­day be­cause it has three very in­ter­est­ing prop­er­ties that lend them­selves to com­pe­ti­tion — and en­able text­book cases of dis­rup­tion:

To­day you can rent scale from any­one, but you can’t rent ter­abites of data from your com­peti­tor’s server.

Data is in­fin­itely scal­able, with lit­tle to no DATA IS SCAL­ABLE. marginal cost. Typ­i­cally, when I ask an au­di­ence if Uber is a Big Data busi­ness, ev­ery­one raises their hand. It seems so ob­vi­ous; how­ever, I would ar­gue that Uber is a small data com­pany, and here’s why: When I wanted a taxi, I used to have to raise my hand and wait for a taxi driver to see that hand. In or­der for that sys­tem to work, you needed thou­sands of taxi driv­ers, driv­ing around the city look­ing at the side­walks, pro­cess­ing reams of vis­ual data in a par­al­lel com­put­ing sys­tem that is known as the brain. We ba­si­cally had thou­sands of brains com­put­ing huge amounts of data to iden­tify whether some­one’s hand was risen.

To­day, I can send one sig­nal, and that sig­nal says, ‘I am look­ing for a ride, and I am in this lo­ca­tion’. That is a far smaller piece of data, and the fact that it is cap­tured dig­i­tally al­lows Uber to repli­cate it at no cost to thou­sands of driv­ers around a city. Be­cause data is scal­able at no cost, we are see­ing this through­out the new dig­i­tal econ­omy: Com­pa­nies tak­ing ad­van­tage of the fact that they can get a sig­nal and they can scale that sig­nal very eas­ily. Ba­si­cally, data’s scal­a­bil­ity al­lows a busi­ness like Uber to cre­ate a new op­er­at­ing model — and also en­ables it to im­prove very rapidly.

The sec­ond thing about data is that it DATA IS REINFORCEABLE. gets bet­ter over time. Think about the Netflix rec­om­men­da­tion en­gine. Ini­tially, those rec­om­men­da­tions weren’t very good, but to­day, I’d ar­gue that Netflix un­der­stands the type of con­tent that I want. It has fig­ured out how to de­com­pose the rec­om­men­da­tions al­go­rithms based on time of day and what I’m look­ing for, be­cause over time, it’s been watch­ing how I be­have.

In Netflix’s early days, ev­ery time I rated a movie that I had viewed, it ob­served my be­hav­iour and the al­go­rithm re­in­forced it­self over time. This feed­back loop and the value it cre­ates is

fun­da­men­tal to­day. While I can repli­cate my com­peti­tor’s in­dus­trial scale, I can­not repli­cate the data they have col­lected, or the time they’ve had to re­in­force what they’ve built. Data be­ing reinforceable al­lows for cheap prod­ucts to re­place concierge ser­vices. I can walk into a Neiman Mar­cus or a Saks and get rec­om­men­da­tions from a highly paid in­di­vid­ual who has ex­pe­ri­ence in the fash­ion in­dus­try; or, I can go to Stitch Fix and get the same types of rec­om­men­da­tion with no marginal costs. At first, that rec­om­men­da­tion al­go­rithm will not be nearly as good as what comes out of Saks, but the fact that it re­in­forces it­self over time — im­prov­ing very rapidly — al­lows it to be dis­rup­tive.

To­day, you can rent scale from any­one, but DATA IS DE­FEN­SI­BLE. you can’t rent (or steal) ter­abites of data from your com­peti­tor’s server. In the past, if you hired an en­gi­neer from a com­pet­ing or­ga­ni­za­tion, they would bring with them an un­der­stand­ing of ‘trade se­crets’. Part of the rea­son that we have Sil­i­con Val­ley to­day is that there were no en­force­able non-com­pete clauses, free­ing peo­ple to move from one or­ga­ni­za­tion to another and bring best prac­tices for build­ing a Big Data in­fra­struc­ture into an up­start like Face­book.

When you’re talk­ing about as­sem­bling tech­no­log­i­cal in­fra­struc­ture, that is doable; but when you’re talk­ing about, say, build­ing an AI sys­tem that dif­fer­en­ti­ates be­tween a good rout­ing of ‘how to drop some­body off at the edge of the city’ and a bad rout­ing that wastes 15 min­utes, a par­tic­u­lar in­di­vid­ual’s un­der­stand­ing of in­fra­struc­ture no longer plays a role. It’s the data it­self that al­lows Uber to co­or­di­nate its trans­ac­tions so spec­tac­u­larly, and you can’t sim­ply hire some­one out of Uber that has mem­o­rized the bil­lions of records in a given city sys­tem. As such, it be­comes im­pos­si­ble to poach ca­pa­bil­i­ties from com­peti­tors in the same way.

In­for­ma­tion-based Dis­rup­tion

The dis­rup­tion en­abled by data is a new form of dis­rup­tion. We used to talk about ‘low-end dis­rup­tion’, whereby the dis­rup­tor is fo­cused ini­tially on serv­ing the least prof­itable cus­tomer, who is happy with a ‘good enough’ prod­uct. We also used to talk about ‘new-mar­ket dis­rup­tion’, which oc­curs when a prod­uct fits a new or emerg­ing mar­ket seg­ment that is not be­ing served by ex­ist­ing in­cum­bents in the in­dus­try.

To­day, we are see­ing in­for­ma­tion-based dis­rup­tion. When data is your core as­set, change hap­pens more rapidly. That’s be­cause, once a com­pany like Netflix has the in­for­ma­tion it re­quires to build a good rec­om­men­da­tion al­go­rithm, it ac­tu­ally gets bet­ter at do­ing lots of other things, too.

When Netflix ac­quired the rights to pro­duce House of Cards, it was thanks to its rec­om­men­da­tion al­go­rithm. By mon­i­tor­ing and cre­at­ing pro­files for it users, the com­pany un­der­stood what types of con­tent were in high de­mand (and what was in low de­mand.) Po­lit­i­cal thriller con­tent, it turned out, was in high de­mand, and the plat­form of­fered low amounts of such con­tent. Netflix used the same in­for­ma­tion — data that was re-en­force­able, scal­able and de­fen­si­ble — that had pre­vi­ously al­lowed it to pro­vide concierge-like ser­vice to its users to cre­ate orig­i­nal con­tent that was in ex­traor­di­nar­ily high de­mand.

Com­pa­nies that be­have like this — that har­ness data, build net­work ef­fect, lever­age the scal­able na­ture of those as­sets and re­in­force their data over time — can rocket for­ward in their abil­ity to dis­rupt. And, when an in­cum­bent busi­ness em­braces them, that busi­ness can rocket for­ward, too.

Even if we ac­cept that all of this is true — that what mat­ters to­day is the abil­ity to har­ness data to cre­ate feed­back loops that re­in­force your com­pet­i­tive ad­van­tage — there are still many com­plex­i­ties in­volved in mak­ing this trans­for­ma­tion. Fol­low­ing are some of the key chal­lenges or­ga­ni­za­tions face.

CHAL­LENGE #1: AC­CEPT­ING THAT MANY PEO­PLE NOW CON­TROL YOUR FATE

If you are P&G, you have a re­liance on other firms for dis­tri­bu­tion. CVS is a great, long-stand­ing dis­tri­bu­tion part­ner, but it has very dif­fer­ent goals from P&G. If you’re the Dol­lar Shave Club — which was re­cently pur­chased by Unilever to at­tack the Gil­lette busi­ness — you have a di­rect re­la­tion­ship with your cus­tomers. You can ob­serve their be­hav­iour, see­ing when they or­der more, when they drop off your plat­form, and how long it took from when the first time they logged on to when they made a pur­chase. You even know where they came from, prior to en­ter­ing your vir­tual store.

If you’re Gil­lette sell­ing prod­ucts through a CVS or a Safe­way, those re­tail­ers don’t nec­es­sar­ily have the same vested in­ter­est in track­ing that data for you. At the front en­trance of the store, imag­ine be­ing asked by a store at­ten­dant whether you came from Whole Foods be­fore walk­ing in the door or if you came straight from home. CVS has a vested in­ter­est in mak­ing the in-store ex­pe­ri­ence as pos­i­tive as pos­si­ble. So, for many busi­nesses to­day, the chal­lenge is fig­ur­ing out how to work within an ecosys­tem of dis­tri­bu­tion part­ners that don’t have the same in­cen­tives to col­lect the type of data that will al­low you to trans­form your busi­ness.

CHAL­LENGE #2: VARY­ING IN­CEN­TIVES

Even within your own or­ga­ni­za­tion, there are peo­ple with a va­ri­ety of in­cen­tives. We re­cently had a guest in our class who runs a multi-hun­dred-thou­sand per­son or­ga­ni­za­tion, with a large num­ber of union­ized em­ploy­ees. The fact is, these em­ploy­ees have very dif­fer­ent in­cen­tives when it comes to em­brac­ing dig­i­tal trans­for­ma­tion over the long term — es­pe­cially with re­spect to the dis­place­ment of jobs. Man­ag­ing a va­ri­ety of dis­tri­bu­tion part­ners, sup­ply part­ners and em­ploy­ees with vary­ing in­cen­tives is a real chal­lenge. One thing we rec­om­mend is to clearly un­der­stand what your core busi­ness is. Fun­da­men­tally, you need to un­der­stand what points of lever­age you have: What you rely on, who your dis­tri­bu­tion part­ners are, what their in­cen­tives are — and whether their busi­nesses will be around in the fu­ture.

Once a com­pany builds a good rec­om­men­da­tion al­go­rithm, it ac­tu­ally gets bet­ter at do­ing a lot of other things, too.

CHAL­LENGE #3: THE NEW COM­PETI­TORS PLAY BY DIF­FER­ENT RULES

If you speak to any­one at Gen­eral Mo­tors, Ford or Daim­ler, they will tell you that ‘the fu­ture is elec­tric’. New­comer Tesla knows that the fu­ture is elec­tric, as well. But some­how, Tesla is al­lowed to make a 100 per cent bet on win­ning in 2025 or 2030. It’s al­lowed to lose huge amounts of money to­day and raise cap­i­tal from pub­lic mar­kets in pur­suit of its 2030 vi­sion. Whereas, if you’ve been an in­dus­trial busi­ness for 100 years, your share­hold­ers will have very dif­fer­ent ex­pec­ta­tions. Put sim­ply, when your com­peti­tors play by dif­fer­ent rules, it makes it very dif­fi­cult to man­age change.

How to Move For­ward

Go­ing for­ward, what is the best way to ad­dress all of this? Fol­low­ing are five guid­ing prin­ci­ples.

1. The stark­est dif­fer­ence be­tween the lead­ACCEPT RE­AL­ITY. ers who have been able to drive large-scale trans­for­ma­tion and those who have not, is that or­ga­ni­za­tions that ad­mit re­al­ity do bet­ter. They rec­og­nize that the things that made a com­pany great in the 20th cen­tury will not make a com­pany great in the 21st cen­tury. If you be­lieve that all the things that made you great are in­sur­mount­able, then you may very well rel­e­gate your­self to be­ing the next Borders, which con­tin­ued to build more stores as Amazon tor­pe­doed the in­dus­try, be­cause scale be­ing prox­i­mate to cus­tomers was thought to be the Holy Grail. But, when you can ship a book next day any­where in the coun­try, the fact that a book­store is half a mile away from my house doesn’t mat­ter. We all have ad­mit to re­al­ity: The game has changed.

2. Too of­ten, we BE VERY CLEAR ABOUT WHAT YOU ARE OF­FER­ING. fool our­selves into be­liev­ing that the cus­tomer buys what we sell to them. But, as Har­vard’s Ted Le­vitt has pointed out, when a cus­tomer buys a quar­ter-inch drill, what they are ac­tu­ally look­ing for is a quar­ter-inch hole — and if you lose track of that fact, you miss the point. Many of our or­ga­ni­za­tions be­lieve that cus­tomers care about per­for­mance the way we de­fine per­for­mance. If that were true, then a com­pany like Stitch Fix — which sends per­son­al­ized boxes of clothes to end-users — would not be grow­ing at the rate at which it is grow­ing. Of course, it is not the same type of shop­ping ex­pe­ri­ence that you get when you go to a high-end re­tailer like Neiman Mar­cus and ask for opin­ions; but, it turns out that fig­ur­ing out what looks good in dif­fer­ent sit­u­a­tions can be achieved via an al­go­rithm. If you re­ally get to the core of what you do, it will be much eas­ier to think about how data can sup­plant those jobs.

3. GE’S Beth Com­stock, who over­sees ES­TAB­LISH A NORTH STAR.

GE In­no­va­tions, has re­counted the story of a 2008 off-site in which for­mer CEO Jeff Im­melt forced his lead­ers to es­tab­lish a vi­sion for what would dif­fer­en­ti­ate their busi­ness in the fu­ture. All of GE’S lead­ers agreed that 15 to 25 years in the fu­ture, soft­ware-en­abled in­dus­trial prod­ucts were go­ing to be per­va­sive, be­cause it made sense to col­lect data to

pre­dict break­downs and op­ti­mize us­age pat­terns us­ing con­nec­tive de­vices. And, once they had es­tab­lished a 15-to-25year goal for the busi­ness, it be­came much eas­ier for GE ex­ec­u­tives to dis­cuss what they needed to be done in a threeto-five year time­line.

Any­one can ar­gue about whether a change like elec­tric ve­hi­cle pro­duc­tion in the au­to­mo­tive in­dus­try is go­ing be preva­lent next year, a year af­ter, or a year af­ter that; but if no one can ar­gue that 20 years from now, the in­dus­try will be reliant on elec­tric ve­hi­cles, then the com­pa­nies that will win 20 years from now are those who lead boldly in that di­rec­tion. Any­thing they do that sub-op­ti­mizes for that 20-year vi­sion will be coun­ter­pro­duc­tive. If you have such a ‘North star,’ you can ex­e­cute and avoid mak­ing haz­ardous de­ci­sions at the ex­pense of your long-term goals.

4. There are BUILD NEW OR­GA­NI­ZA­TIONS, MET­RICS AND PART­NERS. peo­ple who will not pros­per in the new econ­omy and part­ners — whether they be dis­tri­bu­tion ven­dors or sup­pli­ers — who will have to be left behind. If you ad­dress the first three con­sid­er­a­tions, it should be­come ob­vi­ous which part­ners need to be left behind. This is im­por­tant, be­cause if you tie your­self to bring­ing a num­ber of peo­ple with dif­fer­ent in­cen­tives along, it will be­come ex­traor­di­nar­ily dif­fi­cult to make the re­quired changes.

You will still have man­agers who are in­cen­tivized to op­er­ate your core busi­ness, and share­hold­ers who de­mand that you do that job well. At the end of the day, your ex­pan­sion into new mar­kets will be funded by the core busi­ness that you’ve suc­cess­fully ex­e­cuted. How­ever, you need to rec­og­nize that the lead­ers who are fo­cus­ing on ROI to­day can­not com­pete with com­peti­tors who play by very dif­fer­ent rules. The lead­ers who are great at run­ning to­day’s busi­nesses will sys­tem­at­i­cally de­pri­or­i­tize in­vest­ments in the types of things that will fend off the next Uber in your in­dus­try, and as a re­sult, you need new or­ga­ni­za­tions and met­rics to move ahead.

Visa did this quite well by split­ting off the part of its or­ga­ni­za­tion that was fo­cused on build­ing mi­cro-ser­vices and de­vel­oper-fo­cused so­lu­tions atop its pay­ment plat­form.

If Visa be­lieves that e-com­merce en­abled pay­ments are the fu­ture, it’s vi­tal that part of the or­ga­ni­za­tion thinks about that all the time — even if it’s not nec­es­sar­ily the core of the busi­ness to­day.

5. The good news is, dig­i­tal is not a zero-sum EN­LARGE THE PIE. game. The econ­omy con­tin­ues to grow glob­ally, and if it grows at three per cent, it will dou­ble in 20 years. Even if it grows at two per cent, it will dou­ble in 30 years. When you sit down and strate­gize with your lead­er­ship team, you need to fo­cus on what you can do to en­large the to­tal pie avail­able to you. What new ser­vices could you add with dig­i­tal? Which new user groups could you en­gage? Who was pre­vi­ously priced out of the mar­ket that you can bring on board if you have a sim­ple, easy sales process through a dig­i­tal chan­nel?

In clos­ing

By em­brac­ing the five prin­ci­ples out­lined herein — ac­cept­ing re­al­ity, be­ing clear about what you do, es­tab­lish­ing a North star, de­sign­ing or­ga­ni­za­tions to tackle new op­por­tu­ni­ties and fo­cus­ing on en­larg­ing the pie — dig­i­tal trans­for­ma­tion is pos­si­ble for any or­ga­ni­za­tion.

Em­brac­ing dig­i­tal trans­for­ma­tion has not been easy for any of the firms that are now lead­ing the way, but it has be­come nec­es­sary. Lead­ers who deny or ig­nore what is hap­pen­ing around them will not stop the mar­ket from fol­low­ing its course, and no amount of de­nial will keep the world from chang­ing.

Maxwell Wes­sel is the Gen­eral Man­ager at SAP.IO, a di­vi­sion of SAP that sup­ports early-stage start-ups that will cre­ate value for SAP cus­tomers. He is also a Ven­ture Part­ner at Nextgen Ven­ture Part­ners and a lec­turer at Stan­ford’s Grad­u­ate School of...

Newspapers in English

Newspapers from Canada

© PressReader. All rights reserved.