bEST PRaCTICES

Brian, your expertise could help a person who runs an NGO or government agency under-

stand social capital a bit better. Let’s start with the basic question: What is social capital?

Theeasiestwayforpeopletothinkaboutsocialcapitalisthatitisthevalueofthecontactsintheirnetwork.Acon-tactnetworkincludeswhoyouknowandhowthepeopleyouknowknoweachother.Youcanlearnalotaboutaperson’ssocialcapitalbylookingatthatstructurealone. Whenyouthinkaboutrelationshipswiththepeopleyouknowandhowthosepeopleknoweachother,Isug-gestyoufocusonthreeessentialthings.Firstisthekindofinformationyougetfromthosepeople.Thesecondisthekindofskillsetsthosepeoplehavethatyoudon’tpossessandwouldprobablyhavetomakealargeinvestmentto

acquire.Thirdisyourinformalpowerasaleader—areyousomeonewhocangetpeoplecommitted,whocanhelppeopleinterpretmandates,orwhocanhelpothersdrawinferencesfromdataormakesenseoutofwhatsomeoneelsehastoldthem?Positivesocialcapitallieswithsome-onewhoisabletogetrichinformationfromhisnetwork,whosecontactshaveadiversityofskills,andwhogetspoweroutofhisnetwork. Thefirstissueiswhatkindofinformationyouwant,keepinginmindwhatkindofinformationisgoingtobereallyvaluableinyournetwork,fortherearemanydif-ferenttypes.Onetypeofinformationiseverythingavail-ableontheinternet.Youcanscrapeeverythingoff,andit’spubliclyavailable.Itturnsout,though,thatthatisnotthekindofinformationthatgivessomeoneapositionofpowerandedge.Thereasonisthatifeveryone

« an interview with Brian Uzzi »

Paul Godfrey discusses the proper building of social networks with expert

Brian uzzi. uzzi’s award-winning and frequently cited research uses social

network analysis and complexity theory to understand human achievement.

His research has appeared in numerous journals—domestic and interna-

tional—and has won many awards and grants. uzzi is the Richard l. thomas

Distinguished Chair in leadership at the Kellogg school of Management

of Northwestern university, where he has innovated teamNet©, leadNet©,

six Degrees of separation Worksheet©, and Build your Personal Board of

Directors©, each analyzing different facets of networks and giving feed-

back on how to improve social capital through the networks. uzzi advises

major firms worldwide, including Baker and McKenzie, Deloitte, Pepsico,

Creditsuisse, and the World Bank.

Keys to Understanding Your Social Capital

PHoto: RoBB HIll

PH

ot

o: B

oB

Mo

oR

E

5

hasaccesstothatinformation,itgivesyounocompara-tiveadvantage. You’renotlookingforthatkindofpublicinformation;youarelookingforwhatpeoplecallprivateinformation.It’sunstructured,it’stacit,it’snotdocumented,andyouwon’tfinditontheweb.It’softenemerginginformation.Youspendalotoftimetryingtogetitandtryingtofig-ureoutwhatit’sallabout.WhataresomegoodexamplesapersoninanNGOmightthinkof?Privateinformationmightinclude:Aregovernmentstakingsomeoftheirequityandbudgetoutofoneareaandputtingitintoanewarea?Aretheygoingtoshrinkeverythingalltogether?Aretheygoingtogrowsomethingnewthatmightcanni-balizeonaparticulararea?That’sthekindofinformationthatdoesn’tappearinthewrittenrecordanywhere,butit’sbeingsharedsomewhereelse.That’sallprivateinforma-tion,andthat’swhatflowsthroughyournetwork.

How do you get that through your network?

Whileyouexamineyournetworkandcontactsyouneedtorememberthatthebigissueistrust.Thatmaysoundself-evident.Youmightask,whatcanyoutellmeabout

trustthat’sdifferent?Whatwehavelearnedabouttrustinnetworksisthattherearereallythreeimportantthings.Oneistheself-evidenttruththatIjustmentioned—you

havetohavetrustinyournetwork.Whenyoutradeyourprivateinformationwithsomeoneelse,yougetvaluebackbasedontrust. Therearealsotwotrusttraps.Oneisthatmanypeo-plejuststartingtobuildtheirnetworksseemtooverbuildtheirnetworkbasedontrust.Theytrytobuildalotoftrustintoit,butwhenyoubuildalotoftrustintoit,youtendtocrowdoutanotherdimensionofyournetwork.Andonceyoucrowdoutthisotherdimensionofyournetwork,youhaveaweaknetwork.Thatotherdimen-sionisdiversity,andtheproblemiscalledthetrust-diversityparadox. Theothertrapisthatpeopleneedskillstoturnthosewhoarepotentialrivalsintocollaboratorsandtrustingconnections.Manypeoplewalkawayfromthosewhoviewthemwithdistrust;it’salostrelationship.Butthosepeoplewhoconvertthosekindsofdistrustingrelation-shipsintotrustingonesreallygetabigbangoutoftheirsocialcapital.Soyouhavetolearntodeveloptrustwithotherswhoyounaturallydistrust.

Is trust enough?

No,thesecondthingyouwanttokeepinmindaboutnetworksistheconceptofdiversity.Doyourcontactshavedifferentskillbases,background,training,andwaysofseeingtheworldthanyoudo?Iftheydon’t,youbuildredundancyinyournetwork,andthatturnsouttobeabadthing.There’saphraseusedinbusiness:anechochamber.It’sshorthandforsayingthatifpeopleinyournetworkhavesimilarbackgroundsandtraining,thenreallyallyougetisanechochamberaroundyou.Itdoesn’texpandyoursocialcapital;itclosesyouofftothings. Thereisaverysimplementaltesttofindouthowdiverseyournetworkis.Agoodthingtoaskyourselfisthis:OfthepeopleIgotoforadvice,howmanyofthemalsogotoeachother?It’saproportionequation.Ifyougototenpeopleforadviceandfiveoutofyourtengotothesamepeople,thenyou’vegotaproportionoffiveoutoftenor50percent.Sevenoutoftenis70percent,andtenoutoftenis100percent.Asthatproportiongetsclosertoone(or100percent),thediversityinyournetworkgetsclosertozero. Manypeopledonothavediversenetworks.InarecentsurveyofAmericanbusinesspeople,thediversityproportionswere8.2outof10.Sotheaveragebusinesspersonyoumightgotoforcounselgoesto80percentof

Private information flows through your network.

� ESR—Fall 2008

thesamepeoplethatyoudo.Thequestionis,why?Well,thisisthetrust-diversityparadox. Mostpeoplehavemuchhigherexpectationsfortrust-ingpeoplewhoarelikethemthanunlikethem.Wethinkthatwhenpeoplearelikeuswecanpredicttheirbehavior;weknowwhattheirinterestsare,howtheytreatproblems,andwehaveasenseofhowtheywillbehavewhenwe’renottheretoobservethem.Thosebuildingnetworkswhereprivateinformationandtrustgohandinhandtendtospendalotoftimeselectingpeoplewhotheyreallythinktheycantrustbecauseitlowersthecostsofmanagementandofgettingprivateinformation. Ofcourse,whattheytendtodoisoverbuildtheirnetworksbasedontrust.Consequently,theycrowdoutdiversityandareoftennotevenawareofit.Thereasontheybecomeoblivioustoitisthatwithoutdiversityintheirnetworkstheydon’tevenknowwhatthey’remissing;everyonearoundthemsharesthesamepointofview,givingafalsesenseofsecurity.Butinfact,it’snotsecure.It’squitetheopposite.It’saverynarrowpointofviewthatdoesn’tincludealotofthecomplexityyouwouldotherwisewant.Ifyouhaveanetworkwherepeoplesharethesamepointofviewasyou,thenyou’rejustwastingyourtime.Youalreadyhavethatknowledge.Yougetalotoftrustbutnotenoughdiversity.Thisparadoxisabigtrapinpeople’snetworksandmakesthenetworkslowinsocialcapital.

Are there ways out?

Asawaytogiveasolution,letmedescribewhatitisthatmakessocialcapitalsoimportant. I’vetalkedaboutpri-vateinformationandtrustanddiversity.Thethirdimpor-tantthingfornetworksispower.Howdoyougetpowerfromyoursocialcapital?Imagineinyourmindpicturesofnetworkswiththepersonatthecenterofaclusterandthepeopleheknowsaroundhimandthepeopletheyknowarearoundthem.Yougettheseclusters,cliques,orpock-ets.Whenyoumapthesocialworld,youfindoutthat,infact,that’sthewaylotsofnetworkdiagramslook. Itturnsoutthatalargeportionofpeopleliveinthesecliques,butthereareafewpeoplewhobridgethecliquesandbringthemtogether,oriftheydon’tbringthemtogether,theyseetheinformationandtheideassimulta-neously.Thisallowsthemtobebetterandmorecreativeinnewwaysthatnooneinanysinglecliquecouldonhisownbecausehelackstheinformationthatsomeoneseesinanothernetwork.Thosepeoplehaveanameina

network:brokers.PeoplewhohavethetypeofpowerI’vementionedtendtobebrokers.

Thereisareallywell-knownresearchstudythatgoesalongwiththisthatsomeofyourreadersmightbeinter-estedin.1Itwascompletedbyoneofthegreatgeniusesofthelastcentury,StanleyMilgram.Milgramdiscoveredthisideaofbrokers,andalongwithit,hecoinedthisfamiliarphrase:sixdegreesofseparation.Brokersanddegreesofseparationwindupgoinghandinhand. Letmetellyouabouttheresearchbecauseitwillhelpputthisinperspective.Milgramhadtwopassionsinhislife:researchandtravel.WhenhetraveledhelikedtogototypicalplaceslikeParisinthespringtimeorNewOrleansforMardiGras,andhelikedtogotoexoticplaceslikeMadagascarorPagoPago.Wheneverhewouldgoonthesetrips,hewouldplayagamewithacompletestrangerhe’dmeet.So,maybewhileonatourinBorneo,he’dbesittinginarestaurantintheeveningandwouldseeacouplesittingatanothertable.HewouldgouptothiscoupleandintroducehimselfasStanleyMilgramandthenaskoneofthecoupleifthey

Learn to develop trust with

others who you naturally distrust. Turn those who

are potential rivals into

collaborators and trusting connections.

7

best practices

wouldplayagamewithhim,whichwaswhetherthetwoofthemcouldfindapathofpeoplethatcouldconnectthemtohim.Milgramgotthisideabecauseinhistravels

thiskindofconversationwouldusuallyarisespontane-ouslyandtheywouldfindthattheyactuallydidhaveasetofcommonconnections.Theconversationwouldendin“Gee,isn’tthisasmallworld?”Infact,thatisthenameofMilgram’sstudy. Milgramstartedtothinkthatthisishappeningmorethanitshouldhappenjustbychance.HegotbacktohisofficeatHarvardUniversityinthesociologydepartmentafteronesuchtripanddecidedtoconcoctthisingeniousexperimentaboutwhatasmallworlditreallyis.HeopenedhisphonebookinCambridge,Massachusetts,andchosethenameofasinglestockbrokeratrandom.HethengotaphonebookfromasmalltownoutsideofOmaha,Nebraska,whichwasaprettyremoteplaceintheearly1960s,andchosethenamesof160peopleatrandom.Theyarebutchers,bakers,homemakers,cross-ingguards,lawenforcementofficials,andteachers—awiderangeofpeople.Hesendseachofthese160people

aletter.Intheletteristhenameofthesinglestockbrokerandinstructions:Ifyouknowthestockbroker,thensendtheletterdirectlytohim.Ifyoudon’t,sendittosomeoneyouknowwhoyouthinkwouldsendittosomeonewhowouldeventuallygetitbacktothestockbroker. Milgramwastryingtocountlinksordegreesofsepa-rationbetweenpeople.So,hegotallthetransmissionsback,andhefoundoutthat,onaverage,ittakessixinter-mediariestoconnectpeoplewhoareessentiallychosenatrandom.Thisiswherewegettheexpressionsixdegreesofseparation.Hefoundoutsomethingelse,though.Sixtypercentofthetransmissionspassedthroughthesamefourpeople.Nowponderonthat.Thatsuggeststhatwe’renotreallyallconnectedtoeverybodyelse,butthereareafewpeoplewhoaredisproportionatelyconnected;itisthroughthemthateverybodyconnectstoeverybodyelse.Milgramcalledthosepeoplebrokers.Brokershopfromclustertoclustertocluster,andthat’swhatmakesthemsoeffectiveinorganizinganetworkthatmeetstheirneeds. Sowhatisimportantforyourreaderstorememberabouttheirnetwork?Youwanttrustandprivateinfor-mation,butrememberthattrusthasthistrust-diversityparadox.Youwantdiversitysothatyouhavedifferentskillsets.Finally,youwantbrokersbecausethatgivesyoupower.

The big question is, how do you do it?

Youhavetounderstandwhatpeopleusetobuildtheirnetworkbecausesomepeopledoitrightandsomepeopledon’t.Youwanttolearntheprinciplesthatpeoplewhodoitrightuse.Therearebasicallythreeprinciplesforbuild-inganetwork.Firstisself-similarity;wetendtochoosepeoplealotlikeourselves.Secondisproximity;wetendtochoosepeoplewhojusthappentobearoundus.Thirdisthesharedactivityprinciple. Self-similarityandproximityarenotthebestpro-cesses.Self-similaritygivesalotofpositivefeedback,lowercosts,andlessconflict.Thereislessdisagreementandlessneedforyoutochangebecauseyou’renotgettinginformationthatyoudon’talreadyhave.Self-similarityissomethingthatpeopleusealot,andsomeofthat’sokay.Everybodyhassomeself-similarityintheirnetwork,butwhenyouhavetoomuch,thediversityproportiongoeshigh,andthat’sabadthing. Proximityisreallyaconsequenceofhowtheworldorganizesitself.Theworldorganizesthingsbylikeness.That’swhywehavedepartments,units,disciplines,silos,

Without diversity in your network, all you have is an echo chamber. Everyone around you shares the same point of view, giving a false sense of security.

PH

ot

o: J

aR

EN

WIl

KE

yP

Ho

to

: Bo

B M

oo

RE

� ESR—Fall 2008

andalltheseotheraggregations.Imagineitthisway:Mostpeoplelivetheirlivesinanaggregationwithotherpeoplealotlikethem.Asaconsequence,theirnetworkbecomesareflectionofthemfilledwithpeoplejustlikethem.Andmostpeopleareinsensitivetoitbecauseitisjustaconse-quenceofthehabituationofourlives. Thosetwothings,self-similarityandproximity,buildclique-likenetworks.Wehavefoundthroughresearchthatthethingthathelpsbuildpowerfulnet-works,highintrustanddiversity,isthesharedactivitiesprinciple.Thesharedactivitiesprincipleisverypower-fulbecauseitbreaksdownthetrust-diversityparadox.Basically,itworkslikethis:Peoplewhogetinvolvedinsharedactivitiesofallsortstendtomakecontactwithpeopleunlikethemselves,andtheactivityitselftendstohelpyoubuildtrustquicklyandinalow-costwaywiththosepeople.Itgivesyouallthebenefitsofself-similar-itybutwithnoneofitsdownside. Whataresomegreatsharedactivitiesforsomeonewho’spartofanNGO?Servingonnonprofitboards,

communityservice,volunteering,teamandpartnersports,cross-functionalteamsandcommittees,etc.—allofthosethingsthatbringpeopletogether,whethercommerciallyorientedornot,tendtoplayaveryimportantrole.Why?Becausepeopleofallbackgroundsandwaysofviewingtheworldcanbeattractedtothesamesharedactivity.Thesharedactivityputsyouintoadealstreamwhereyou’recomingincontactwithdiversepeople,andproximitygetsbrokendown. Onceyouareinthatsharedactivitywiththesediversepeople,howdoyoubuildtrustwiththem?Well,itturnsoutthattheactivityhelpsyoudothatquickly,andthat’sthemainbenefit.Youalreadyextendsometrusttothatpersonjustbybeingintheactivitywiththem.Youcouldn’tparticipateinthatactivityunlessyoutrustthatperson,andviceversa. Also,researchshowsthatwhenyou’reinvolvedinthesharedactivities,youareprofilingthepeopleintheactivity.Youarelookingattheircharacterandqualities,howtheyactundervariouscircumstances,andhowthey

best practices

dealwiththeunexpected—evenifyouarenotconsciousofdoingit.Likeasubliminaladvertisement,thisprofilingleavesanemotionalfootprintorsignaturethatcreatesasmalleffectonhowwethinkofthatperson.Manypeoplebelieve,againonasubconsciouslevel,thatwhatyouobserveinthesharedactivityisinsomesenseatruerpictureofthatperson’sunderlyingcharacterthanwhatyougetwhensittingacrossthetableordeskfromeachother,timeswhenyou’vehadtimetoscriptbehaviorandmanageimpressions.

So this is why we play golf or basketball with other people?

Exactly!Ithelpsreallybuildtherelationship.And,ofcourse,thisisbuildingarelationshipbeforeadealcomesalong.Therelationshipprecedesthedeal,whichmakesthedealgosmoother,moresimply,moreeffectively,andlesscostly.Andthroughthisprofiling,yougettranspar-ency.Theotherthingisalotofthesharedactivitieshavethingsatstake.Ifyouwinyoubreakarecord;ifyouloseyoudon’tbreakarecord.Youarguewiththeteamtomotivatethem,andthistypeofworkhelpsbonding—weeithercommiserateinlossorcelebrateinvictory.Andbondingbuildstrust.

What else should our readers pay attention to?

Sharedactivitiesareparticularlyimportantintwoaddi-tionalways.ManypeopleinNGOsarebuildingnetworksandtryingtogetinfrontinverylittletime.Theyarepressedandhaveamillionthingstodo.Sohowdoyouconnectwiththepeoplewhoarealreadyreallybusy?Whatyouwanttodoispiggybackonthingsthattheyhaveadeeppassionfor,athingtheyalwaysfindtimefor—andthatisusuallyasharedactivity.Theycareaboutthisthing,andtheirpassionbringsthembackoverandoveragain.Consequently,yougetsomeopportunityforfrequencyofinteractionthatdoesn’tdrawthemawayfromtheirscheduleorroutine. Thesecondthingisthatsharedactivitieshelpyoucommunicateonamoreevenlevelwithpeopleofdiffer-entranksthanyourself.Insociologywehaveaconceptcalledthescript.Basically,inascripteveryinteractionyouhavewithsomeonefollowscertainexpectationsofwhatisappropriateorinappropriate.Wealldoitallthetime.Ascriptdoestwothings:First,ittendstomakeussimilartoanybodyelsefollowingthesamescript.Second,ittendstosuppressqualitiesorcharacteristicsthatmightbeimportanttothepersonthathe’stryingtoreach. Imaginethatinthemiddleofthisinterviewyouallofasuddenstarttalking;youstartsaying,“Allthismakesmethinkofsoandsoinsuchandsuch;letmesketchoutmynetworkandlookatthis.”Yougoonforaboutahalfanhour,anditturnsouttobeanabsolutelybrilliant,Einstein-likeinsightaboutnetworks.TheproblemisthatthescriptwouldmessupmyabilitytoseethatbrilliantinsightbecauseIwouldbecomeuncomfortable.Iwouldfeelyouwereviolatingthescript,andthatviolationofthe

Shared activities break down the trust-diversity paradox. Shared activities help you communicate on a

more even level with people of different ranks. Shared activities allow you to step out of one script and into another script. Shared activities are an engine for building powerful networks rich in social capital.

10 ESR—Fall 2008

scriptwouldoffsetanyvalueofyourinsightsandcharacter. That’stheproblemwithscripts.Ifyou’rerunninganNGOandyou’retryingtomeetagovernmentofficial,animportantdonor,oracorporationthatyouwanttoinfluence,youarefollowingascriptwhenyoumeetthatotherimportantperson.Thatpersonisalsofollowingascript,andwhentheyseeyou,youarelikeeveryotherNGOthattheyhaveeverseenbecauseofthescript.It’ssuppressesyourspecialqualities.Asharedactivityallowsyoutostepoutofonescriptandintoanotherscriptwhereyoucanshowotherqualitiesaboutyourselfthatcouldbevaluabletothetargetperson.Ifyou’regoingtoplaysquashorgolforbeonthenonprofitboard,theseotherscriptsallowyoutoshowthingsaboutyourselfthattheprofes-sionalorhierarchicalscriptdoesnot.Thoseareopportuni-tiestobuildtrustwithpeopleunlikeyouandshowthingsaboutyourselfthatyoucannototherwiseshow. Let’srecap.First,youhavediversitybecauseyou’reinadealstringwithasharedactivity.Second,youbreakdownthetrust-diversityparadoxbybuildingtrustchieflywithpeoplewhoareunlikeyou.Third,yougetbrokerageasanaturalbyproductoftrustanddiversity.Sharedactiv-itiesareanengineforbuildingthesepowerfulnetworksrichinsocialcapital.

Many NGOs work in areas of the world where insti-tutions are weak and institutional capital is weak. Governments may even be corrupt, and the rule of law may be suspect. How can positive social capital help people overcome weak institutions?

Thefirstthingthatcomestomindisthatalotoftheproblemsyoujustmentionedare,infact,duetosocialcapitalbeingusedinirresponsibleways.Socialcapitalandnetworksprefigureinsomanydifferentcontextsofhumanachievement.Farmersinsub–SaharanAfricawhowerethefirsttoadoptpest-resistantcropshadthesepowerfulnetworks.Thedoctorswhogetthelatestlife-sav-ingdrugsseemtohavethesebetterkindsofnetworks.Inthe1990s,crimeinsomeinnercitiesoftheUnitedStatesdroppedmorethan60percentbecausepoliceofficerschangedtheirnetworksfrombadonestoonesrichinsocialcapital.Ifyouwanttodosomethinggreat,youneedtohavesocialcapital. Unfortunately,criminalorganizationsandcorruptionalsorevolvearoundnetworks.That’showthingsgetdone.Thenetworkisnotbadinandofitself;thedifferenceishowreasonablypeopleapplyit.WhenIthinkabouthow

socialcapitalcanhelpovercomeweakinstitutions,whenIlookattheliteratureandtalktopeople,Idiscoverthatsocialcapitalhasbecomealotmoresensitivetomeritandtothevaluepeoplebringtothenetworkduringthelastfiftyyears.Thissensitivityismakingitpowerfulvis-à-visweakinstitutions.Creditcirclesareanexampleofpeopleusingtheirsocialcapitaltoovercomecorruptorweakfinancialinstitutions.Therightkindofsocialcapitalallowsthemtocombineresourcesthattheymightnotindividuallyhaveorthattheystrugglewithtosubstitutefortheseweakinstitutions.

What’s the dark side? What should people be aware of as they go out and build these networks?

Thereareafewdarksides.Oneisthistrust-diversitytrap.Ittendstohavepeoplebuildnetworksthatarenotenrich-ingthemselvesorothers;itdoesn’treallyaddvalueforanyone.It’ssomewhatofabenignprobleminthesensethatnooneintentionallytriestofallintothetrap. Ithinkthatanotherdarksideofnetworksalsohasabenignqualitytoit.Despitehavingpositivecharacteris-tics—smallworlds,clusters,andbridges—thesepowerfulnetworksturnouttobeextremelyfragiletothelossofjustafewsuperconnectorsbecausethosesuperconnec-tors,likethefourintheMilgramstudy,arethetiestoeverything.Ifthosepeoplearenoteasilyreplaced—ifotherpeoplecan’tstepintoreplacetheirlinksquickly—thenetworktendstocollapserapidly. Thelastnegativethingtoconsideristhatnetworksarenotgoodorbadinandofthemselves—thepeoplewhousethemdeterminethatworth.Aswelearnmoreabouthowpeopleusenetworksandmoreabouthowpowerfulnetworkscanbeintheirstructures,theywilleventuallybeadoptedbygroupsthatwillusethemforselfishpurposes.Ifwewerehavingthisconversationafewyearsagothefirstthingthatwouldcometomindwouldbeterrorists.Terroristcellsarecliquesandconnectedbybrokers,andtheyareabletohaveaveryfluid-likestruc-tureanddothesortsofthingswesawthemdobecauseofthenetwork.That’sanextremeexample,butmembersofNGOsneedtomakesuretheyareusingtheirnetworksinappropriateways.

Endnotes 1StanleyMilgram,“TheSmallWorldProblem,”Psychology Today

(May1967):60–67.

11

best practices