金融大數(shù)據(jù)創(chuàng)業(yè)案例-Jun-Wang(Twitter)_第1頁
金融大數(shù)據(jù)創(chuàng)業(yè)案例-Jun-Wang(Twitter)_第2頁
金融大數(shù)據(jù)創(chuàng)業(yè)案例-Jun-Wang(Twitter)_第3頁
金融大數(shù)據(jù)創(chuàng)業(yè)案例-Jun-Wang(Twitter)_第4頁
金融大數(shù)據(jù)創(chuàng)業(yè)案例-Jun-Wang(Twitter)_第5頁
已閱讀5頁，還剩38頁未讀， 繼續(xù)免費(fèi)閱讀

下載本文檔

版權(quán)說明：本文檔由用戶提供并上傳，收益歸屬內(nèi)容提供方，若內(nèi)容存在侵權(quán)，請(qǐng)進(jìn)行舉報(bào)或認(rèn)領(lǐng)

文檔簡(jiǎn)介

金融大數(shù)據(jù)創(chuàng)業(yè)案例燒了五千萬美元的教訓(xùn)12ParallelUniverseGFW3Aboutme4BankingGoogleCheckoutAdwordsGoogleMapStartup(today’sstory)SocialmediaadvertisingStartupAdvisor5TheIdea6ABetterGroupon7What’sWrongwithGroupon?What’sWrongwithGroupon?PlatformSellersBuyersTheIdeaHelpretailers/merchantstosendhighlytargetedandrelevantpromotionalofferstoconsumerswhohavehighinterestsintheirproduct“Take$10offanypurchaseatXYZstore,offerexpiresin10days”Whereisthedata?Whereisdata?MerchantsthemselvesOnlinetrafficlogsPOSmachineBanks…Howdowegetthedata?Howdowegetthedata?Banksdon’thavetechnology&expertisetodoitrightPartnershipwithbanksWin-winproposalwithpartnersBothonlineandoff-lineWegotdata!Wegotdata!6outoftop8majorbanksCreditcard/debitcardAnonymizedtransactionleveldata2+yearsofhistoricaldata+dailyrefreshesNowwhat?Bigdata!

Machinelearning!

Bigdata!

Machinelearning!

Butbeforethat…

DataProcessingGeodata:locationofstores,locationofcustomersMatchtransactiontolocation/storeCategoryofstoresNatureofthetransaction(online,phone,mobile,etc.)BankregulationsAggregationlevelDataProcessingConsumerTransactionDataRetailerDataHDFSPostgresqlaggregationPredictiveModelingConsumerbehaviorpredictionRetailer1Retailer2Retailer3…RetailerNConsumer1???Consumer2??Consumer3???Consumer4??…ConsumerM??“Wheredoyourcustomersalsospendmoneyat?”AgreattoolforSales/MarketingIdentifypotentialclientsCompetitoranalysisStrong/weakstores/areasCross-categorypromotionsPredictiveModelingInputvariables:location,userspend,shoppingfrequency,categorypreference,timing,etc.Output:purchaseprobability,purchaseamountUser1User2UserM…Retail1Retail2RetailN…User1User2UserM…Retail1Retail2RetailN…Forevery(user,retailer)pair,everyweekPredictingprobabilityofapurchaseCouponvsOrganicOnlinevsofflineNewcustomervsexistingcustomerNewcustomerincategoryvsexistingcustomerincategoryShoppingfrequencydifferencebetweencategoriesBirthdays,annualeventsMatchrate:redemptionreallyhappened?PredictingspendamountRankingofhistoricalspendExistingcustomersIn-categorycustomersHowtodefinecategories?OptimizationProfitableoffers:Applyfilteratretailerlevel,dependingoncoupontypeWhatwecareabout:TotalRevenuesum((expectedspend–discount)*profitmargin*expectedresponserate)Sumoverallusers,maximumoneretailerperuserResponsevolumesum(expectedresponserate)Retailerlevel,maximumoneretailerperuserOtherfactorsOptimizationBalancebetweenmultipleobjectivesandconstraintsTextbookoptimizationproblemLargescalelinear/quadraticprogrammingmodelTheFinalProduct:ConsumerexperienceRegularlyusingcreditcardReceivedamessageaboutastore-specificcoupon:EmailWebMobileGeo-fencingNeedtouseinoneortwoweeksTransactionhappenasusualDiscountappliedoncreditcardstatementTheFinalProduct:MerchantExperienceRunaspecificpromotionprogramNotraininginvolvedEngagedcustomerscomeforpurchasesTheFinalProduct:BankExperienceDatasharingReceivedataandsendoffersouttoconsumersReceiveshareofrevenueFastForward3years…BanksMerchantsConsumersWe’rehereInvestorsWhereweareSiliconValleybased50+employeesfromGoogle,Visa,Yahooetc.3roundsof$50M+VCfundraised100+clients,10M+usersAccuratepredictiononconsumerbehaviororotherwiseHighlytargetedpromotionaloffersfromrelevantretailersWhatcouldgowrong?MeetwithReality–technologyBalancebetweenmultipleobjectivesandconstraintsMaximizetotalrevenuewhilekeepingresponserate>=x%Maximizeoverallresponseratewhilekeepingtotalrevenue>=$XMaximizetotalredemptionvolumewhilekeepingrevenue>=$XHowtobalance?N-1Constraints,1objectivefunctionLinearcombinationofobjectivesMeetwithReality–technologyCustomizedcomplexmodelsTakeshourstofinishInfeasiblestatesolutionaddapre-optimizationprocesstoevaluate(rough)feasibilitywithoutactuallygettingasolutionMeetwithReality–technologyRealdataisdifferentfromlabsimulationExample:RedemptionratealothigherthanorganicbehaviorSpendamountoncouponredemptionisverydifferenttooModelneedtohaveroomforsuchadjustmentsOverall,wehaveahighperformingproductwithhighaccuracypredictionandoptimizednetworkyieldMeetwithReality–BusinessModelCooperativePartnersareextremelyimportantBanks,merchants,investors,andconsumersAllcomponentsHASTOWORK!Significantchallengesduringproductdevelopment,launchandbusinessdevelopmentMeetwithReality–BanksRegulationsSpeedinproductdevelopmentDatasharing/pipeline,frequencyLevelofdatadetails,itemspurchasedCommunicationstoconsumersExclusiverelationshipMeetwithReality–MerchantsRegulations:tobacco,medicine,etc.ProductspecificpromotionsDouble-dippingAdoptionofnewtechnolog

溫馨提示

  • 1. 本站所有資源如無特殊說明，都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請(qǐng)下載最新的WinRAR軟件解壓。
  • 2. 本站的文檔不包含任何第三方提供的附件圖紙等，如果需要附件，請(qǐng)聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶所有。
  • 3. 本站RAR壓縮包中若帶圖紙，網(wǎng)頁內(nèi)容里面會(huì)有圖紙預(yù)覽，若沒有圖紙預(yù)覽就沒有圖紙。
  • 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
  • 5. 人人文庫網(wǎng)僅提供信息存儲(chǔ)空間，僅對(duì)用戶上傳內(nèi)容的表現(xiàn)方式做保護(hù)處理，對(duì)用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯，并不能對(duì)任何下載內(nèi)容負(fù)責(zé)。
  • 6. 下載文件中如有侵權(quán)或不適當(dāng)內(nèi)容，請(qǐng)與我們聯(lián)系，我們立即糾正。
  • 7. 本站不保證下載資源的準(zhǔn)確性、安全性和完整性, 同時(shí)也不承擔(dān)用戶因使用這些下載資源對(duì)自己和他人造成任何形式的傷害或損失。

最新文檔

評(píng)論

0/150

提交評(píng)論