Breaking down big data: How to move beyond the buzzwords to gain actionable insights

The disparate cobbling of analytics reports, and the muscle-memory media buys and martinis are starting to show its age. The pressure to find leverage in an increasingly competitive and digitally enabled market is encouraging brands and agencies to go deeper into customer and prospect data and make some sense of it.

Are you looking for some actionable ways to explore Big Data? Here are some emerging best practices to consider from Advertising Week 2013:

Big is relative. You don’t need a petabyte of information to start mining your customer data in new and valuable ways. Often times building basic relationships between data sets is more important than putting any one thing under a microscope or going wide and shallow that targeting begins to lose its value. Combining data dimensions into a few consolidated views can go a long way to inspire hypotheses on who to target, what to offer, and how to best convert. Dimensions such as time (year-over-year, monthly, etc.), performance KPIs, lifestyle segments, behavioural triggers (for example - contract expiration and new product releases), marketplace dynamics, and related campaign events are just a few examples of data that works well together.

Standardized KPIs and attribution metrics. Sounds simple, but most companies don’t have a common vocabulary and definitions around core performance indicators or metrics. Developing this is critical, especially before expanding into multi and omni-channel analytics.

Contextual Visualisation and Progressive Exploration. Traditional dashboards often lump all KPIs together without displaying contextually relevant sets of data with it. Sales not meeting goals? Don’t just show sales over time; dynamically show the campaign(s) that were running at that time against prospect web traffic, with conversion and volume of referring sources…you get the idea. Beyond contextually relevant information, ‘progressive exploration’ allows the primary data set to be interactively reduced into manageable data sets with shared attributes that can inspire much better market targeting around a business goal.

Not leveraging these two principles is probably one of the biggest reasons most analytics tools fails in being actionable. Think about it, in marketing we are often are trying to solve definitive problems (grow sales, reduce churn, etc.), but with ambiguous tactics and fuzzy audience targets. Without explorative analytic tools we are simply just staring at the facts or quickly become dumbfounded from the complexity.

Stay on target. As a marketer, it’s easy to get caught up in a campaign’s performance as the end goal, which can result in losing sight of the big picture. It wasn’t too long ago when troves of businesses signed up for a Groupon deal to drive walk-in traffic to then realise it was killing their margins and attracting the wrong customers. As big data is leveraged better, businesses can glean more real-time and predictive insights on how marketing decisions will impact the bottom line. One way to do this it to assign a LTV (Lifetime Value Score) to your base and look for related tells in your prospects that map to the same attributes in your base.

Got Site Personalisation? What is fueling it? Many companies are overhauling their dotcoms to enable deeper personalisation. However, it begs the question where is the actual personalisation logic coming from? Many industries have a treasure chest of data on their current base of customers. However, when looking to grab net new customers all that intelligence often gets over looked. The power to leverage Big Data to create new and virtual segments should be a boon to marketers, especially where prospect and third party segmentation can help bridge the gap between knowing who your best customers are and attracting more of them.

Factor in Omni-Channel. It’s probably an eye-roller to jam in another larger than life buzzword, but there is a compelling reason to do so. In addition to ongoing retail supply chain related advancements, Omni-channel is just starting to become being actionable for most marketers at a broader scale. Offer codes, cookie tracking, shared carts, and mobile campaign activation can provide some connective tissue, but bigger solutions like unique IDs, and sensor enabled retail environments are examples of solutions that will more fully connect each channel and enable omni-channel modeling and analytics.

Build brand love. Many industries that already thrive on building business through data analytics often forget the importance of building brand love. Good analysis and optimization can go a long way, but there is no substitute for brands that embrace shared values with their customers and kick-out a great product and service at the same time.