There is a growing focus among marketers to leverage customer transactional and behavioral data that is already available within an enterprise through advanced analytics. Micro-segmentation, acquisition models, lead qualification models, identifying cross-selling and up-selling opportunities through market basket analysis are all tools available to make the marketing task more focused.

However, and this is a big however, most of these tools are predictive tools based on past behavior. Reasons for a particular behavior can, at best, be inferred ie. we assume that consumers will behave in a certain way in future because of the way they have acted in the past. The big fly in the ointment is  that we do not know why they did so or why  they would do so in the future. This, to me, is incomplete understanding – because circumstances that drove that behavior may change or marketing initiatives to drive a particular behavior may be wrongfully directed. Which is why, there is a crying need to integrate attitudinal research into the behavioral analytics model.

Let’s say a supermarket has data to show that Hot Wheels cars and candy are purchased together. The easiest decision here is one of placement – place the 2 categories together for maximum uptake of both categories.  But do we need to stop there? Let’s say, the store was running a promotion targeted at kids. Should we place the promotion at this point – hypothesizing that the kids’ pester power was driving sales here? Or should we assume that the buyer is a young mom shopping alone - in which case, a promotion targeted at her is more likely to catch attention.  Questions like these can and should be answered with simple, on-ground research surveys.

To carry the same analogy further, data could throw up a high correlation between purchase of Hot Wheels and contraceptives. Obviously, common sense dictates that the placement decision is irrelevant here. But how could this information be leveraged for better offtake? Researching habits and attitudes even in a qualitative way armed with this information could provide significant insights. Very little granularity would derive from mere data mining.

The reverse is also true. A customer satisfaction survey would certainly provide inputs into evaluation and correction of internal and external processes.  But over time, smart businesses would find it much more optimal to link the feedback mechanism to customer value/business transactions so that they can act on the critical processes and not necessarily only the ones with the highest visibility.

While traditional market research as defined by the 45 minute, pen & paper interview is certainly becoming redundant in the pace of today’s world, the explosion of alternative media and technology has thrown up a whole clutch of other touchpoints for marketers and researchers to structure smart research around. Building a marketing recommendation around a strong foundation of attitudinal research led analytics will make the difference in the future.

Analytics is hot! It’s the latest buzzword. And as with all new trends, there is a great deal of confusion on what exactly ‘analytics’ is. How is it different from Business Intelligence? Where does market research come in? This is an attempt at throwing some light in that direction..

Definitions and the like…The Wikipedia definition for analytics is – “… how an entity (i.e., business) arrives at an optimal or realistic decision based on existing data”.  The definition for Business Intelligence – “…. refers to computer-based techniques used in spotting, digging-out, and analyzing business data ….” indicates degrees of overlap but also differences.  Both aim to analyze existing enterprise data and are widely used for decision-making.  The difference is that analytics aims to take up where BI leaves it.

Focus: BI has traditionally been a top management view of the health of the company – not a mere bird’s eye view but again, not exactly a nuts and bolts analysis. The focus has been on automating and consuming aggregated data for monitoring performance and for early warnings. This has been through various interface applications like scoreboards and dashboards on Key Performance Metrics. 

Analytics, on the other hand, aims to dig deeper; to help the line manager take relevant action based on the data. Automation figures in the scheme of things but the need for flexibility and human insight is critical as business dynamics change. This is a much more ‘end-to-end’ solutions game – the tool (BI or other) is relevant for its credibility and reliability but the critical factor is the insight and consulting edge.

Who are the players? The platform vendors(Oracle, SAS, IBM SPSS, SAP, MS ) of course – new products and platforms developed to perform complex statistical models; Software/ Software as a service  (SaaS) vendors providing end-to-end solutions; consultancies and market research agencies providing cutting edge insights are all competing in this space. 

Applications: In view of its broad ranging need for operational excellence, analytics finds a use across functions.

Marketing Analytics aims to help optimize marketing ROI – through better segmentation and targeting, through up-selling and cross-selling opportunity analysis, through campaign tracking, pricing & channel analytics.  In partnership with the Customer Service function, they also deploy customer analytics to lower acquisition cost for customer retention and to minimize churn.

 Financial analytics focus on credit risk management, fraud and risk management, equity research, claims analyses and the like. HR analytics can help gain better insights into identifying and nurturing talent, identifying potential attrition and root-cause analysis for this and so on.. Similarly supply chain analytics.

Web Analytics is, of course, an area of far-reaching importance today. Understanding how consumers navigate the web, your web-site, what facilitates conversion and what impedes, are now critical elements of any corporate’s web strategy.

In subsequent posts, we will try and take a deeper look into each of the above application areas as well as analytics principles and techniques.