But investments, whether in talent, technology, or partners, need to be carefully considered and maintained. Brands need to continue investing in their new hires’ growth and data skills to maximise their contribution to marketing objectives. While data mining and statistical analysis are hardly new in the world of marketing, the content, platforms and applications (mixing online and offline data, or integrating different devices) for these are, and they’re changing all the time. Today’s statisticians and data professionals need to wrap their heads around a whole raft of new metrics relating to online performance, social media resonance, video viewing and location-tracking data, all across multiple devices, to uncover the impact on each to driving digital and in-store sales.
Due to the newness of availability and access to all these variables, analytic talent with deep expertise in making connections across these types of data points is scarce. The ability to accurately interpret data analysis and translate insights into the right action requires input from individuals with a unique blend of industry expertise and analytic prowess – and marketers need to continuously test and learn.
First Steps In Predictive Analytics
With the right analytic expertise and a test-and-learn culture in place, marketers can even begin to venture into the arena of predictive analytics, leveraging those “must-have” metrics to anticipate which audience segments or consumer actions will result in a specific outcome. Marketers can work with their teams and their knowledge of predictive analytics to drive brand equity and sales across retail and communication channels; and use data-based evidence to support further investment in the right areas.
Working with a European confectionary brand recently, Millward Brown combed through multiple data sources to quantify the value of paid, owned and earned activity on social media in driving both short-term sales impact and long-term brand health. By evaluating consumer interaction across key touch points and quantifying the causal relationship each interaction (or sequence and combination of interactions) had on sales and brand equity measures, the brand was better able to prioritise its future activities, and maximise sales. Similarly, we worked with a US telecoms company to examine the on-site and competitive site activity of current customers to identify online predictors of churn. By continually tracking churn indicators, the telecoms company is able to detect early warning signals of customer defection and employ counter-measures to mitigate risk.
Marketers are currently facing the front end of a steep learning curve as they themselves test and learn how to harness the power of big data efficiently. Through further experimentation, the focus next year will shift from “big” to “intelligent” data, and ultimately, human knowledge and expertise applied to big data assets will be recognised as the key to translating the hype over big data into reality – and results.