（英文）TalkingData数字化运营方案-爱茉莉.pptx

1、12022-7-24Strategic Cooperation DepartmentTalkingData Digitized Operation Solution22022-7-241Digitized Marketing on Retail Industry Understanding of Current Situation32022-7-24Digitized marketing on retail industry Understanding of current situationThe lack of data accumulation,proper diagnosis and

2、assessment system on operation,causing no way to deploy digitized operation CRM has accumulated a huge amount of data,but how to locate a customer precisely,and to build user circle?The lack of proper operation assessment system after the building of APP,data from online and offline are not linked?o

3、ffline customer flow situation,the membership conversion rate,customers profiles?How to operate while putting customers in the core and using data as the basis?By building DMP platform,we use data and let data talk,transforming the company into a data-driven one.42022-7-24 Based on 360 customer prof

4、ile,conduct group classification and group enlargement Build multiple ways to reach customers:online Ads,SMS,calling centers,and etc.Connect to marketing effectiveness data,in order to generate an effective closed marketing loop Multiple ways for closed loop marketing Unawared in-store customers mob

5、ile devices ID collection off-line store customer flow statistical analysis Off-line store customer groups profile detecting Off-line store customer flow analysisBy establishing Data Management Platform(DMP),it can connect with first party data,off-line customer flow data,and labeled data from the t

6、hird party,combining all three to illustrate a unified and standard customer profile.Meanwhile,it can build up a off-line/in-store customer flow analysis model and marketing model upon on the DMP,to meet the needs of operation department for customer analysis,marketing,and etc.Combine first party&th

7、ird party data,through ID Mapping to create unified labels of customers Form 360 comprehensive customer profile360 customer profile Centralized and unified big data management platform Integrated current modules of big data platform Multiple sources of data infuseCentralized and unified big data pla

8、tformThe Overall Goals of Digitized Operation 52022-7-24TalkingData Digitized Operation Solution62022-7-24-外部数据-企业数据采集前移-New channel:mobile internet传感器手 势Off-site dataOn-site data交易门交互门公开市场门StorageDW&BIServer3 gates in data application for enterprises:big data accumulation&application still have a l

9、ong way to goOpen Market Gate-Data from outside-Interaction Gate-Enterprise data collection moves forward-Transaction GateSensorGesture72022-7-24Core Members Core Members SegmentationSegmentationkey value memberskey maintain memberskey developing membersKey retain membersNew membersActive membersSil

10、ence membersSleeping membersBusiness SceneBusiness SceneApplicationApplicationCost ManagementAnalysisMembership labeling managementOperation&sales managementCustomer relationshipmaintenanceData CollectionData CollectionLayerLayerNeedle（Wi-Fi）CRMIbeaconWeChatOfficial accountsAPPPOS&integration system

11、TD datasupplementTD datamarketData ProcessingData ProcessingL LayerayerCustomer operationplatformClose loop marketingplatformSMC（Smart Marketing Cloud）Customer flow operation platformStatistical analysis platformData Application Data Application LayerLayerData combining Data labelingClustering analy

12、sis modelOther algorithm modelsBusinesslayerApplication layerData layerBlueprint of the pan-membership management systemPerceptive membersCore membersHigh-value membersImprove Membership Converge RatePotential membersPan-membership Management System82022-7-24TDData ConnectorHive(Basic attributes lis

13、t)TD Bitmap数据源及数据属性管理标签及标签分类管理Off-line computationl Support from multiple sources of datal Codeless&visible interface of metadata deployment(TD ETL)l Flexible data structure l Based on TD Atom Cube,high performance on data assessmentl Batch updates based on timed tasksData MarketFirst party&third pa

14、rty data assessment92022-7-24Pan-membership main dataIDFAAndroid IDMACIMEIIMSILBSPhone#Citizen IDMembership#Open IDUsing ID Mapping technique,combining multiple sources of data and manage with unified ID,in order to achieve pan-membership managementTalkingDataTDIDIDFAIDFVAndroid IDMACIMEIIMSIPhone#M

15、D5IPLBSID correlated graphic techniquePCIPTVIMEIIPMACCookieMACClient Management SystemClientClient IDIDClientClient namenameBrandBrandCategoryCategoryRentRentAP Management SystemPhone#Phone#MACMAC OpenIDOpenIDMembership Management SystemPhonePhone#MemberMember IDIDCitizenCitizen IDIDTransactiTransac

16、tiononPointsPointsTalkingDatas DataTDIDMobile device#Mobile device attributePopulation attributeApp interestsGeographic locationConsumption customFinancial standingConsuming capabilityCredit assessmentInvestment preferenceVehicle conditionAchieve data property iterationEnsure the compliance of priva

17、tization deploymentNo data privacy violationNeutral third party standing Unification of usersData Matching and Connecting102022-7-24TalkingData User Operation Platform-labels Labels that enterprise generated through its own business operation system“First party internal labelsTalking Data cloud labe

18、ls and other labels from the third party“Third party labelsUsing DMP to create labels needed in business operation,including basic&system labels“Self-definedlabelsBased on labeling rule model,DMP integrates self-defined labels,first party internal labels,along with the third party labels on data,to

19、create elaborate definitions of labels on all people covered.TD now has 10 categories covering 900+labels112022-7-24TalkingData User Operation Platform-Group Insight Group Insight includes profile dimension and index analysis dimension Provide flexible setting ability to support profiling and index

20、analysis Business index analysis based on group Available for subdivision analysis based on group attribute,and sub-group can be generated122022-7-24TalkingData user operation platform-internal and external potential group mining Available to find similar group within the enterprise with LOOKALIKE a

21、lgorithm to cross-market Potential customers can be mined beyond the enterprise(TDs cloud market for example)through LOOKALIKE algorithm to provide precise group for marketing132022-7-24TalkingData customer flow operation platform-elaborate analysis ability on customer flow Division of 8 types of cu

22、stomer groups:Divide 8 types of standard customer groups such as new customer,old customer,project visiting,project staying,single store front,single store entering and single store staying Personalized setting of group definition(visiting,staying)3 categories 21 types 76 indexes in all:3 index dime

23、nsions:time category,space category and profile category,including 21 types such as customer flow,entering rate,staying rate,staying time,density of customer and app preference improvement degree Available for 1st party and 3rd party customer group index extension 142022-7-24TalkingData customer flo

24、w operation platform-Ability to generate customer profile on the cloud regularly Generate store profile regularly:Each store would regularly(3 months as default)form customer groups cloud data profile Availble for profile comparison of different stores as well as different periods in the same store

25、Offline customer group profiling based on cloud big data:Population attributes:gender,age,car ownership,marriage and children Consumption ability:phone brand,phone priceAPP preference:APP behavior preferance coverage degree,APP preferrence improvement degree(significant or not)152022-7-24TalkingData

26、 customer flow operation offline Working&living analysis Living place&working place:According to target customer groups off-site activities to calculate their living place and working place,then customer value and consumption ability can be inferredAvailable for comparison of different customer grou

27、ps TOP Living Place Recommendation:Provide TOP 10/20 community recommendation list and group percentage,based on geographical POI information and location information162022-7-24TalkingData customer operation platform-Competitive brands analysis functionAP/WIFI collection tool of competitive brands p

28、roject:Collecct WIFI environment(SSDI of AP/WIFI)of competitive brands project through APP provided by TDAvailable for customer flow data uploaded by mobile needle Distribution of customer group radiation:Present distance distribution of customer group source in your own and competitive brand projec

29、t,to evaluate customer group radiation ability of different projectsTOP 10 store ranking:Present customer flow ranking of sub-project(store)in your own and competitive brand projectCorrelation Index:Proportion represented by overlapping customer group between your own and competitive brand in each o

30、thers project;it can indicate competitive strength among projects,more relevant more competitive172022-7-24TalkingData SMC(E)Group constructionGroup profile insightAd service management182022-7-243Case Studies192022-7-24Bestseller CRMEvent ManagementMember InfoMember ID&InfoCustomerInsight/ProfileCu

31、stomerStore WIFIUser Info(Authentication)Device Infor(Not logging in)Traffic&MovementOnline AnalyticsCritical Step 1:Data CollectionMarketing Management On PremiseDevice InfoPhone#/WeChat IDDevice InfoDMP On PremiseCritical Step2Data EngineeringMarketing ScenariosIn Store SAAOffline/In StoreReal tim

32、e ContentsCritical Step 1:Data CollectionMarketing Cloud SAASAD Tracking SAASCustomerInsight/ProfileMarketing ScenariosTmallJDH5Online/M-APPsPush Messages AD ContentsCritical Step 3:CustomerAwarenessCritical Step4:AssessmentBig data marketing solutions for an internationally renowned clothing brand 2017/2/1820Thanks!