In­no­va­tion’s Re­wards and Trade-offs

Rotman Management Magazine - - FACULTY FOCUS - In­ter­view by Karen Christensen

De­scribe the ‘norm of col­lec­tive­ness’ that is re­plac­ing the age-old tra­di­tion of the in­di­vid­ual ge­nius in the realm of in­no­va­tion.

Most in­no­va­tive and sci­en­tific en­deav­ours are now col­lab­o­ra­tive ven­tures, with many di­verse peo­ple con­tribut­ing to them. Our fa­mil­iar im­age of Ein­stein work­ing alone in his lab and com­ing up with the The­ory of Rel­a­tiv­ity doesn’t hap­pen nearly as of­ten to­day. Wal­ter Isaac­son’s book, The In­no­va­tors, em­pha­sizes that col­lec­tive de­vel­op­ment has be­come the norm, par­tic­u­larly for dig­i­tal and In­ter­net tech­nolo­gies.

What are the im­pli­ca­tions for or­ga­ni­za­tions who want to con­tin­u­ously in­no­vate?

The im­pli­ca­tions are that in­te­gra­tion and en­sur­ing cross­func­tional in­ter­ac­tion are crit­i­cal for in­no­va­tion. Si­los may be in­di­vid­u­ally cre­ative, but re­ally align­ing in­no­va­tion with your or­ga­ni­za­tion’s goals is a col­lab­o­ra­tive ef­fort.

You have noted that in this en­vi­ron­ment, knowl­edge work­ers face an im­por­tant trade-off be­tween credit and col­lab­o­ra­tion. Please de­scribe it.

De­spite the shift to­wards col­lec­tive in­no­va­tion, the re­wards from in­no­va­tion still ac­crue mainly to in­di­vid­u­als. If you are a busi­ness per­son form­ing a startup with some co-founders, you can agree up-front on an eq­uity split, and as things change, you can ad­just it. So, en­trepreneurs get quite a bit of dis­cre­tion as to ‘who gets what’ in terms of shares in a busi­ness. But when it comes to more re­search-ori­ented and sci­en­tific en­deav­ours, you don’t get to de­cide on those shares.

For ex­am­ple, the 2013 No­bel Prize in Physics was awarded to Peter Higgs and Fran­cois En­glert, who were re­spon­si­ble for de­vel­op­ing key con­cepts in the the­ory that pre­dicted the Higgs bo­son. Only these two in­di­vid­u­als were hon­oured, de­spite the fact that it took decades of ef­fort by thou­sands of sci­en­tists to find the bo­son. Work by sci­en­tists at CERN and at the Te­va­tron par­ti­cle col­lider at Fermi Na­tional Ac­cel­er­a­tor Lab­o­ra­tory de­vel­oped the search tech­niques and elim­i­nated a sig­nif­i­cant frac­tion of the space in which the bo­son could hide, al­low­ing ex­per­i­ments by Higgs and En­glert to fi­nally find the elu­sive par­ti­cle.

Do you see this as un­fair?

Yes, it is un­fair, be­cause the big re­wards in science come from this recog­ni­tion. But on the other hand, there is some re­flected glory, so it isn’t all that bad. This is a tough is­sue. Not ev­ery­one can get a prize, but per­haps these things should be awarded to groups.

Stud­ies — in­clud­ing your own — show that cre­ative out­puts ac­com­plished by a large num­ber of peo­ple tend to be of higher qual­ity; why is that?

Clearly, many heads work­ing to­gether can solve more prob­lems. In our study, we looked at projects that peo­ple had to ac­tu­ally choose to be a part of. They could have de­cided to

col­lab­o­rate or to tackle the sub­ject on their own. So, they had to con­sider the both costs and ben­e­fits of be­ing part of a col­lab­o­ra­tive pro­ject.

As in­di­cated, col­lab­o­ra­tive projects will likely have a greater over­all im­pact, so that’s part of what at­tracts peo­ple to them. But there are also some con­straints to con­sider. One is that col­lab­o­ra­tion is hard: you have to talk to peo­ple and par­tic­i­pate in meet­ings, and then there is the at­tri­bu­tion as­pect we just talked about — a con­sid­er­a­tion of, What re­ward am I go­ing to get as part of a team ver­sus be­ing out on my own? If you’re on your own, the credit goes en­tirely to you; but if you’re part of a team, the re­sults will likely be more im­pres­sive — and therein lies the trade­off.

Our study looked at the be­hav­iour of MIT sci­en­tists and the choices they were mak­ing. In any given year, sci­en­tists will choose to be part of some col­lab­o­ra­tive projects and to do some of their own. We were able to mea­sure the im­pact of those dif­fer­ent en­deav­ours by look­ing at ci­ta­tions, us­ing 50 years of data. We found that while col­lab­o­ra­tive projects carry more to­tal value, sci­en­tists were not join­ing them be­cause they were con­sid­er­ing how much credit they would get. Clearly, this has im­pli­ca­tions for the progress of re­search and in­no­va­tion.

Do you rec­om­mend agree­ing on credit al­lo­ca­tion in ad­vance?

One in­ter­est­ing thing is that it is not ob­vi­ous what the for­mula for the al­lo­ca­tion of credit is. If you have ten peo­ple work­ing on a pro­ject, you might think, ‘What­ever the value of the pro­ject is, each per­son should re­ceive one tenth of the credit’. But as in­di­cated, it’s not like divvy­ing up shares in a com­pany. Con­ceiv­ably, ev­ery­body could get the full value of the pro­ject credit.

What would that look like?

It is a bit like ev­ery­one get­ting a prize. Re­wards are just pa­per, so you can just print more. But like cur­rency, do­ing so can de­value it. More­over, you might end up with too many peo­ple col­lab­o­rat­ing in name only; so it is tricky.

You found that the best sce­nario is ac­tu­ally some­where in be­tween; how so?

We ac­tu­ally found a pre­cise for­mula for it: one over the square root of the num­ber of col­lab­o­ra­tors. You wouldn’t want it to be one over ten, be­cause that would mean there’s no ben­e­fit to col­lab­o­ra­tions — no syn­er­gies to be had; and you wouldn’t want it to be one, oth­er­wise ev­ery­body would be clam­our­ing to get in on ev­ery pro­ject.

In another re­cent study, you looked at how en­trepreneurs go about form­ing ‘co­op­er­a­tive strate­gies’ with in­cum­bents. What do these look like?

Ba­si­cally, if you’re an en­tre­pre­neur­ial firm, one of the choices you make in terms of com­mer­cial­iz­ing your tech­nol­ogy is to ‘choose’ who your com­pe­ti­tion will be: you can take your prod­uct di­rectly to mar­ket and com­pete with es­tab­lished firms, you can li­cense your prod­uct by form­ing an al­liance with Com­pany X, or you can be ac­quired by Com­pany Y. The lat­ter two choices are forms of co­op­er­a­tive ven­tures, and we regularly see ex­am­ples of both. For in­stance, What­sapp re­cently came out as a com­peti­tor to Face­book, but it was ac­quired by them. These broad choices are open to en­trepreneurs, and we stud­ied what drives them to go one way or the other.

You note in your pa­per that dis­rup­tive tech­nolo­gies can be dif­fi­cult to com­mer­cial­ize in a co­op­er­a­tive setup. Why would this be?

Dis­rup­tive tech­nol­ogy, ac­cord­ing to its tra­di­tional def­i­ni­tion, has two fea­tures. The first is, when it comes onto the mar­ket, it doesn’t ap­peal to most peo­ple. We stud­ied the speech recog­ni­tion in­dus­try — think Siri. In this in­dus­try, tech­nolo­gies were in­tro­duced that had faster pro­cess­ing: you speak into

it and it can re­spond more quickly; but these faster an­swers may not have as big a vo­cab­u­lary.

Imag­ine a new tech­nol­ogy comes into the mar­ket, and you get a quicker re­sponse from it, but it doesn’t have as many words. Over time, the prod­uct that pros­pers will be the one with a big­ger vo­cab­u­lary and the faster re­sponse speed at the same time. But at the out­set, you don’t know whether this will hap­pen.

If you are an in­cum­bent firm and you see an en­trant like this, what do you do? You can pour your re­sources into fol­low­ing it over time, which seems waste­ful. The al­ter­na­tive is to wait and see, and down the road, say, ‘Great, that tech­nol­ogy panned out. Now we can ac­quire that com­pany and en­joy all the ben­e­fits of it.’ It will cost you a bit more now, but that’s the price you pay for wait­ing.’ We found that, par­tic­u­larly for dis­rup­tive tech­nolo­gies, in­cum­bents tend to fol­low this wait-and-see ap­proach.

How do you de­fine ‘Model-based Prob­lem Solv­ing’?

Mod­els are the­o­ries about how the world works, and they can be used to solve prob­lems. If you say, ‘The sun al­ways rises’, that’s more of a pre­dic­tion — al­beit a good one. To cre­ate a the­ory, you have to ac­tu­ally un­der­stand a bit of the mech­a­nism: the sun al­ways rises be­cause the Earth is ro­tat­ing. That’s more of a the­ory. Even if you don’t know why it’s ro­tat­ing, you know that it is ro­tat­ing. Based on this the­ory, if the Earth stopped ro­tat­ing, we would not see the sun.

In the So­cial Sciences, we look at how peo­ple be­have, and what pre­dic­tions we can make based on that. With re­gard to in­ves­ti­ga­tions of en­trants com­ing in with dis­rup­tive tech­nolo­gies, we had to con­sider all the pos­si­ble op­tions. We came up with a the­ory that for dis­rup­tive tech­nolo­gies, the en­trants would en­ter and be­gin to com­pete, and the in­cum­bents would wait and see and then, when they prove suc­cess­ful, ac­quire them. That pre­dic­tion didn’t come about be­cause we thought it was a good pre­dic­tion. It came about be­cause we con­sid­ered all the el­e­ments — what new en­trants do, and what could they do? Well, they could com­pete or co­op­er­ate, and then they could change their mind. And, What could the in­cum­bents do? Well, they could try to track the en­trants or try and com­pete with them, or they could change their minds. As it turned out, our the­ory pre­dicted this pat­tern of en­ter­ing, then com­pet­ing and then co­op­er­at­ing.

In short, mod­els are de­signed to un­cover some truths of the world. Of course, the world changes con­stantly, so some of the pa­ram­e­ters of ev­ery model will change; but what we’re look­ing for are mod­els that can ex­plain a good chunk of what we see, so we can rely on them to make de­ci­sions.

De­scribe the model you cre­ated to de­scribe an in­di­vid­ual’s quest to man­age the col­lab­o­ra­tive as­pects of their knowl­edge-work port­fo­lio.

That model was pretty sim­ple: it spec­i­fied what in­di­vid­u­als get from col­lab­o­rat­ing with oth­ers, and what prop­er­ties might that en­tail? Also, what are the costs as­so­ci­ated with it? Es­sen­tially we came up with the math­e­mat­i­cal for­mula I men­tioned ear­lier, but it was from the point of writ­ing that down that we re­al­ized, Okay, now we’ve got to think about the in­di­vid­ual’s de­ci­sion to par­tic­i­pate in a pro­ject, as op­posed to do­ing some­thing else; we had to think about what that some­thing else might be. The data we ac­cu­mu­lated al­lowed us to ad­dress as­pects of the de­ci­sion, and it also in­cluded some un­known vari­ables. But be­cause we had a model, we could ask the ques­tion, If the data is what it is, and our model is what it is, and ad­mit­tedly, we don’t know ev­ery­thing, what lev­els would those quan­ti­ta­tive mea­sures have to be at, in or­der for the model to be con­sis­tent with the data? Whether you are an economist, an aca­demic or a busi­ness leader, you need to come up with the­o­ries about how the world works.

Higgs and En­glert shared the 2013 No­bel Prize in Physics, while the thou­sands who helped them went un­named.

Newspapers in English

Newspapers from Canada

© PressReader. All rights reserved.