From Silos to Success: A Decision Maker’s Approach to Big Data

 

There are a million stories playing out on the big data landscape – not all of them pretty. Some marketers are working smoothly with data to drive their businesses to success.  But others are struggling to cope as data gets bigger and more ubiquitous.

The availability of data itself does not resolve marketers’ consumer information issues.  To reap the benefits of all this data, they need to turn it into actionable insights. Data must be integrated to fully realize the value of information.  The question is: How do you approach this task – of synthesizing big data and all kinds of other information? How can you get the most out of data?

 

1. Every good solution starts with a well-defined problem

All too often, data integration starts with someone in an organization –typically a senior manager — asking, “Why you don’t we take advantage of all the data available to us?”  The request often arises without context, process, or clear goal.

Then the research team goes to its partners and asks the same random, unfocused question.  If action is taken at all, the knee-jerk response is to bring data streams together with little planning and emphasize the technology.  Breaking down silos is important; but if that is all you do, you are left with the same problem you had before – lots of data, but no decisions coming from it.

Effective use of multiple data sources must start with well-defined business questions, problems, or issues.  Without these, you are simply creating a bigger haystack to search for the needle.  The process of articulating these questions does not need to be cumbersome. Simply writing down what you and/or your organization are working towards provides a great start. Then prioritize and dig into a topic to expose precise business questions.

 

2. The data you need is determined by the answers you seek

Literature is full of quest stories. In these tales (think Odysseus or Frodo), the protagonist is in search of something – he or she has a goal and is working towards it.

Data integration is no different; the business questions and issues help you define the goal.  Having multiple sources of information that can be combined in different ways provides many paths our hero can take.  Successful quests have a common thread – focus.

When Indiana Jones was searching for the Ark, he zeroed in on this goal. This same discipline needs to be applied to data integration efforts.  In the realm of data analysis, focus is achieved by having the right sources and amounts of data; too much can take you far off course.

By building off of precise business goals, you can craft a simple analytic model that will help you determine which sources to interrogate or combine.  For example, your business objective may be to increase share.  This leads to the question, “What are the factors that influence market share?”  There are market pressures – price, competitive products, and competitiveness of your value proposition.  There are internal factors, such as costs, service, and delivery. And there are customer factors, as well – satisfaction, perceptions, and needs.

From these “inputs” you can build a specific business question, such as, “Does our service model facilitate repeat purchase?” Then you can identify sources for the data you need to get to the answer.  Data will typically be sourced internally, from 3^rd parties and from custom, ad-hoc techniques, such as customer surveys.  Having a focused list of what you need will help you stay on target in your quest to address the important business questions of the day.

Make sure you pay attention to what you cannot learn from internal data. It tends to be harder to understand why customers make the decisions they make – but knowing the “why” is essential to effective messaging and targeting. This type of insight can be obtained through consumer surveys and focus groups.  Make sure you have sources to understand consumer motivations, perceptions, and decision making before you run off making plans to modify marketing, service design, or products.

 

3. Tools make the job easier – but they do not solve business issues

Data integration can be messy. Data sets from different sources do not fit neatly together. You could spend more time pulling data together and making it ready for analysis than actually analyzing the data. Today’s business environment moves too quickly for this kind of slowness.

This is where technology comes in. Across the value chain, programs and services can help you bring data together (think of Hive) and analyze it quickly (Hadoop and others).  But being able to run a report is vastly different from being able to interpret data.  And then there is the important step of communicating the findings and recommendations in a memorable, compelling way to spur change within your organization.

Do not mistake the ability to generate some output with the bigger task of developing business recommendations and motivating action. For this, you will need specialists who understand the business context, can derive meaning from data, and can effectively communicate to senior stakeholders. These people need to connect the dots, pulling insights from multiple sources, interpreting models, and tapping other aggregate sources.

Think again about Indiana Jones.  Much like his tools, such as his signature whip, big data technology is a tool.  It is his skill, understanding of the situation, and intuition that allow him to successfully find the treasures he is seeking.

 

4. Don’t review the past – see the future

When analyzing data, you typically are looking in the rear-view mirror at patterns that emerged in the past. This can help you understand what the future may hold – but there are also robust analytic tools that can take data, in real-time, and predict what will happen next.

Imagine if your retail managers had access to a dashboard that allowed them to see what aspects of the customer experience they should focus on today to maximize conversion.  And the focus areas for one outlet would be specific for that location, taking into account macroeconomics, geography, demographics, and historical trends.  Such a tool is possible through data integration and big data analytics. You will continue to need to review the past to make strategic decisions; but having tactical tools that drive the right behavior at the point of sale is going to become essential for a distributed sales force.

Making use of data from multiple sources is a must for every marketer. Going into the endeavor with a clear plan is key to success, as is maintaining what made you successful in the past – your ability to interpret information and use it to make decisions.

 

Photo by Franki Chamaki on Unsplash