The role of CPG analytics to tackle changing consumerism
Blog Posts:Mu Sigma
Published On: 09 June 2015
Changing population demographics are posing a challenge to CPG companies. In 2015, Millennials will overtake Baby Boomers as the largest living generation. As they enter the workforce, their rapidly increasing spending power will make them the most important consumers, for marketers, for many years to come.
With the Baby Boomer generation, marketers had to rely on segment insights primarily – methods such as archetypes and personas. But an avalanche of data exists on Millennials, and much of it is at the individual level. Why? Because they are heavy tech users, and their every move generates data. Marketers can use this data to generate insights and customize marketing.
Numerous tools and data analytics platforms have cropped up to help marketers collect, analyze and leverage this data. Hadoop, for instance, has given organizations new ways to handle the volume, velocity and variety of data. The number one barrier to marketers in leveraging this data? Privacy and data security policies. Google, Facebook, Amazon and Netflix may lead the pack in terms of the number of users who avail of their services on a daily basis, but the data they collect is essentially dark matter to third parties. Outsiders are only able to see glimpses of what lies within. An enterprise may spend billions of dollars on capturing data from these sources and leverage their big data infrastructure to generate insights, and still end up without a concrete solution to the Millennial equation – just the same segmented data they collected on Baby Boomers 10 years ago.
So what should marketers do? We know that broadcasting generic messages to Millennials is a costly and ineffective exercise. Customization of messaging and channels is the route to take. The problem now takes the form of identifying preferences of groups at a micro level to ensure success.
Ideally, social media could function as a massive focus group and a number of listening platforms from various vendors exist, which provide analysis along with monitoring services to calculate user sentiment, understand a user’s influence over a group and even measure the effectiveness of a digital campaign. The broad range of capabilities offered by these vendors is only limited by the lack of critical mass of demographic and geographical data. In the coming years, the yardstick to measure the performance of these vendors would be their ability to unify the data from multiple channels to provide a comprehensive database of raw data and useful insights. This will allow organizations to tie together data from various sources and analyze them to unlock a new depth of insights about consumer groups.
With the amount of data being generated showing no signs of slowing down, enterprises must learn to understand how best to leverage big data soon. The main aspects of the system are
1.Data Capture: Organizations can capitalize on the digital space by capturing all data from their owned assets, existing / new, such as CRM websites, mobile apps and social pages.
2.Data Management: Hadoop forms the basic framework that allows organizations to cost-effectively store the vast amount of data. Many companies including Hortonworks offer big data analytics software, support and services to business customers who wish to launch big data pilots and test the waters.
3.Data Analysis: Data analytics Companies will have to integrate this big data with traditional data to gain a 360-degree view of the landscape. Programming frameworks such as MapReduce and languages including Java and Python with their extensive libraries can be used to analyze the data. New platforms and languages such as Spark, Scala, mahout etc. are being developed, which are improvements to the existing systems.
Starting with pilot projects and building on their successes will help enterprises realize the benefits of leveraging big data with minimal disruption to the business. In part 2 of this series, we’ll dive deeper into the creation of an analytical roadmap.