Omnidata e o fim do funil de compra Algumas provocações FGV/EAESP, Junho 2016 marcelo.coutinho@fgv.br @mcoutinho
REFER FRIENDS POST REVIEW JOIN GROUPS TRACK ORDER RECEIVE PACKAGE MISSING ITEM SETUP PHONE BILLING ISSUE MAKE PAYMENT PHONE DAMAGE TERMINATE SERVICE NETWORK ISSUE CHANGE ADDRESS RECEIVE OFFER DISCOVER NEED RESEARCH RECEIVE OFFER TRACK ORDER RESTART SERVICE SHOP & BUY SHOP & BUY RECEIVE PACKAGE MISSING ITEM POST REVIEW SETUP PHONE BILLING ISSUE MAKE PAYMENT PHONE DAMAGE TERMINATE SERVICE JOIN GROUPS NETWORK ISSUE CHANGE ADDRESS RESTART SERVICE RECEIVE OFFER REFER FRIENDS WEB PRINT DIGITAL ADS WORD OF MOUTH SOCIAL TV RETAIL STORE WEB SHOP REVIEWS SEARCH KW/ADS CONTACT CENTER WEB SHOP EMAIL SOCIAL RETAIL STORE WEB SHOP CONTACT CENTER CONTACT CENTER RETAIL STORE CONTACT CENTER CONTACT CENTER SOCIAL WORD OF MOUTH SOCIAL EMAIL SUPPORT PORTAL? SUPPORT PORTAL? BRANDED COMMUNITY? AWARENESS INTEREST DECISION ACTION USE ACTION ADVOCACY ADVOCACY DECISION USE Fonte: SAP
O loop do comércio COMPRA USO Fonte: adaptado de A Jornada do Cliente, apresentação do Bradesco no 7º Congresso Brasileiro de Pesquisa de Mercado, 2016
Fonte: IDC
Objetivo da Mensuração: KBOs Despesa Receita Fluxo de Caixa Satisfação do cliente
Causalidade no digital: crash course no Google Analytics
How Programmatic Media works besides several processes in order to connect buyers and sellers and coordinate the bidding process and ad placement, it uses tracking software to identify content consumption and web behavior attributing unique identifications to them (canvas fingerprints, evercookies, tracking pixels, etc)
Programmatic Media
Sharing Context Content Personal
Source: Simbiose
Data don t buy. People do Market Research = Big Data Catalyst
Our idea: Programmatic Sampling Use behavior instead of sociodemographic variables Partnership with IPSOS and LatinDMP 80 million of uniques IDs from 1.2 million domains.br
Our methodology 9 visits to fashion sites AND 6 visits to health information sites (Comscore average x2) We used the same wording in order to compare with similar clusters in EGM In 4 days we had 1,159 clicks from 2,315,506 impressions, resulting in 64 complete and valid questionnaires at a U$ 1K cost (there was no incentive to answer the survey)
Health: top 2 box I always look for info about diseases/medicines on the Web I try to have a healthy diet I do frequently check-ups I always use diet/light products 0% 10% 20% 30% 40% 50% 60% 70% 80% IPSOS EGM Progm. Sample
Fashion and Beauty: top 2 box I like to visit several stores before buying I am influenced by the brands that my friends recommend I like to try new products and brands I like to buy brand clothes 0% 10% 20% 30% 40% 50% 60% 70% 80% IPSOS EGM Progr. Sample
Shopping: bought in the last 3 months
Demographics (IPSOS GMU: sampling to small for sociodemographic comparasion) Age Prgm. Samp. IPSOS EGM Between 13 and 17 yrs. 9,4% 10,7% Between 18 and 24 yrs. 17,2% 19,18% Between 25 e 34 yrs. 28,1% 26,5% Between 35 and 44 yrs. 6,3% 21,94% Between 45 and 54 yrs. 18,8% 13,83% 55+ 20,3% 7,86% Gender Prgm. Samp. IPSOS EGM Female 67,2% 80,49% Male 32,8% 19,51% Education Prgm. Samp. IPSOS EGM Some High School 14,1% 28,4% Some College 35,9% 51,66% College/Graduate 50,1% 19,95%
Conclusions so far We do have a more extreme behavior of general consumption of Fashion/Health information and products as expected Programmatic Sampling can be a viable (and cheap) technique for fast surveys if you have a good programmatic structure and/or you are a publisher/advertiser with a large base of visitors. But it is always good use some kind of benchmark against your results (ratings vendor or single source mkt/media survey) Possible uses for Programmatic Sampling fulfill difficult quotas/behaviors in more traditional web-surveys identify and have a better understanding of early-adopters of new kinds of content make easy the creation of specialist pannels integration with online panels will enrich their results
The convergence of technology and market research: from sociodemographics to behaviors?... the reality is often neither as wonderful as the proponents of the technologies argue, nor as dire as the major detractors fear. Each new technology enhances and extends the range of possibilities and opportunities for survey research, but also often introduces new challenges and issues for further research - Couper, 2002: Designing a Strategy for Capturing Respondent Uncertainty in Web-Based Surveys. Social Science Computer Review, 20 (1): 3-9
Dados NÃO são o novo petróleo
Atenção é o novo petróleo In an information-rich world, the wealth of information means a dearth of something else: a scarcity of whatever it is that information consumes. What information consumes is rather obvious: it consumes the attention of its recipients. Hence a wealth of information creates a poverty of attention. Herbert Simon, 1971
Personalização é uma demanda econômica diante da crescente concentração de renda Fonte: O capital no Século XXI
Um comerciante, qualquer que seja, não é mais que um servidor do público, ou de um público; e recebe uma paga, a que chama o seu lucro, pela prestação desse serviço. Ora toda a gente que serve deve, parece-nos, buscar agradar a quem serve. Para isso é preciso estudar a quem se serve mas estudá-lo sem preconceitos nem antecipações; partindo, não do princípio de que os outros pensam como nós, ou devem pensar como nós porque em geral não pensam como nós, mas do princípio de que, se queremos servir os outros (para lucrar com isso ou não), nós é que devemos pensar como eles: o que temos que ver é como é que eles efectivamente pensam, e não como é que nos seria agradável ou conveniente que eles pensassem (...) o comerciante não tem personalidade, tem comércio; a sua personalidade deve estar subordinada, como comerciante, ao seu comércio; e o seu comércio está fatalmente subordinado ao seu mercado.
A ESSÊNCIA DO COMÉRCIO Revista do Comércio e Contabilidade, Lisboa, 1926
Omnidata e o fim do funil de compra Algumas provocações FGV/EAESP, Junho 2016 marcelo.coutinho@fgv.br @mcoutinho