Marketing has come off the age in the last couple of decades. Before e-commerce started, marketers predominantly used outbound marketing, both in B2B and B2C, to reach out to prospects via offline and electronic channels - mostly TV and Radio.
Last decade was the decade of "Internet of People". Marketers used closed loop digital marketing campaigns to achieve KPI's like click to conversion ratio and revenue targets The focus was on manual analysis of customer data (mostly structured data) to create segments based on customer attributes, execute multi-channel campaigns (email, websites, mobiles, social), do test on small audience, measure the effectiveness and derive insights through analytics and fine tune the campaign for better outcome. Marketers have also begun to optimize the marketing budget with basic attribution capabilities and do simple marketing and media mix modelling to maximize revenue.
With the evolution of Internet, e-commerce, proliferation of digital channels and payment services in the last 10 to 15 years, marketers have started to do both inbound and outbound channel marketing. This has given birth to digital marketing paradigm, where marketers owns the budgets for paid, earned and owned media; generated leads and assisted in cross-sell and up-sell.
Digital Marketing has reached a stage, where we begin to see a degree of maturity in multichannel campaign management.
Happy days, RIGHT!! Not Really. Fast forward the clock into a not so distant future and we see a decade of "Internet of things" where there will be explosion of mobile, sensor, wearable devices and explosion of data emitted and consumed by them. This throws a great challenge to marketers, where in, they now have to deal with huge volume of structured and unstructured data and do marketing based on events and triggers generated in real-time.
Following are the key technology trends and high value banking use cases which will drive next generation digital marketing (i.e. data driven real time) where we will see rise of the predictive & prescriptive analytics:-
Natural Language Processing (NLP)- Most of the technology vendors for social and text media sentiment analysis and trend spotting use one-dimensional text analytics and NLP techniques. While this has an accuracy level of approx. 70% to detect emotion and intent, it still does not detect important nuances like irony or sarcasm. But once those capabilities mature, marketers will use these for branding and sales to derive customer intends and predict segments for targeting more accurately.
Use-case- Potential high value use-case for banks will be ability to do product ideation or services through social intelligence (where banks start to mine and own the data and create the new product or service - a very different model than crowdsourcing).
Big Data & In-memory Analytics - With the increasing maturity to handle large volumes of structured and unstructured data in real-time through Apache Hadoop based MapReduce parallel processing framework; efficient distributed storage systems (file system and in-memory database) like HDFS, SAP HANA, Cassandra, real-time in-memory computing like Apache Sparks, enormous capabilities are being provided to drive real-time prediction and advising.
Marketers can now start to discover data, spot statistical correlation and apply propensity modelling to create Need (profitability) or Value (lifestyle) based segmentation. It will also allow advanced attribution forecasting algorithms and "what-if" analysis to drive marketing and media mix optimization. It will be the beginning of meaningful predictive & prescriptive analytics marketing triggered by event & real-time data.
Use-case-Potential high value use-case for banks will be the ability to use mobile wallet as a trusted payment advisor driving higher sales through loyalty, rewards and discounts.
Machine Learning And Behavioural Science - With the evolution of psychology and other social sciences, it could become a source for business intelligence and help marketers do marketing better. The Neuroscience is still not matured enough to be adopted at large scale but marketers can start to use machine learning algorithms to understand behaviour through engagement techniques such as gamification and refine segmentation and attribution modelling to improve real-time marketing outcomes
Use-case- Potential high value use-case for Banks will be Gamification of a customer engagement scenario to understand psyche & emotion. Once outcomes are measured, feed the outcome into re-designing the customer experience with web and call centre interactions.
Internet of things- We will soon be surrounded by mobile devices, wearables, sensors and intelligent display boards everywhere - every device will emit data and marketers need to capture them to create a 720 degree view of customers. With evolution of API management, big data, in-memory computing, NLP, Machine and Cognitive learning capabilities, marketers will start to truly realize dream of anytime, anywhere marketing. Ability to detect an event, create recommendations and push the content in real-time through devices like smart phones, google glass, advertising board, smart TV or smart watch will be achievable.
Use-case- Potential high value use-case for Banks will be real-time marketing through wearables like google glass; where mobile wallet and smart device will act as an information processor and receiver; but glass will be the used as content delivery platform or even the intelligent delivery display boards will do marketing on real-time based on the audience profile.
Predictive analytics will give rise to Digital Marketing 2.0, a beginning of anytime, anywhere marketing paradigm.