What The Data

There has been a lot of buzz around data for a very long time and yet many feel intimidated by the sheer reference of data management and analytics. Why is that? Aren’t we literally living and creating a data world every day? From the neatly organized personal finances in a spreadsheet to our social media posts, we are creating data at every given instance. It is predicted that the data sphere will grow to 175 zettabytes by 2025. This article is creating a dataset. It has a name, a time stamp, a category, a theme, and going further, a sentiment. The scale of data is now at an unforeseen levels in the entirety of human history along with wide variance in the type of data. No wonder it has caused generic intimidation in our ability to tackle it.

They say information is power. In other words, knowing your data can be powerful. But how do you know your data? To understand the data analytics landscape, it is critical to have understanding of data types. Let’s start with Structured and unstructured. I love structured data (it’s my kind of data :)). Neatly organized in a tabular fashion with easy to derive relationships and insights by simply slicing and dicing in multiple ways. And when you put this into a beautiful visual, Voila. Everything you need to make that impression.

 I wish though, all data was that simpler. 93% of all data by 2022 is going to be unstructured; examples of which are audio/video files, your emails, tweets & pictures with tags and/or geo location.

So why are companies investing large sums of dollars in building and/or modernizing their data infrastructure? 3 key reasons:

1.  Understand the past: To make progress, it is critical to know what is inhibiting progress. Acquiring and analyzing data (operational, financial & organizational) will lead you to the root cause of the problem. Thus, narrowing down the focus on burning issues and giving you a roadmap for further improvements.
2. Predict the future: Finding the right signal in the noise is a scientific experiment. You start with a hypothesis, follow the data trend, if you like where it leads you- you won. If not, build on it or pivot to a new hypothesis. Based on these insights, data scientist can build predictive and prescriptive models for proactive decision support.
3. Make money: At the end of the day, it’s all about shareholder value. Driving improvements or transforming the current modus operandum based on (near) real-time visibility coupled with recommendations based on prescriptive analytics, equals better decision-making to enable revenue goals.

So how do you take advantage of this data? Let me give a very cliched but relevant response. Start with a Data Strategy. Easier said than done, you say. I know there are millions of articles written on data strategy but the goal here is to provide a big picture outline of what an organization needs and in which order.

1. Know the Why: Understanding the business problem you are trying to solve is the key to success. Any experiment starts with the Problem Statement. This problem statement drives the corporate priorities and investment decisions.
2. Set up the goal post: Once you start to uncover the problems you will see there are many tentacles to it and each one has some sort of dependency to another. But you cannot solve all at once. Set some measurable goals for yourself, some near term (your quick wins), and some out in time which is your north star.
3. Plan to Execute: Lay out the deliverables and milestones against a reasonable timeline. Understand your human capital, technical skills and map them to the needs. Call out your risks using a ROAM board. Basically, project management 101.
4. Organize and Simplify: The real battle lies in making sense of the unstructured data or getting the unstructured data to a state of being structured. Converting and classifying the unstructured data using advanced analytics tools leads to in-depth market and customer insights. Try the visualization tools to create dashboards. A dashboard shouldn’t just be charts and pretty pictures but weave in the contextual story to want to narrate. Read your audience and iterate.
5. Maintain and Govern: Staying current is essential when it comes to data as market trends and technology changes rapidly. When the strategy is in place, the whole engine needs to operate like a well-oiled machine without glitches. Build a maintenance plan and right level of security so that right people have access to right data on time.

Now all you need to do is find the right tools to execute that will enable you to make data-driven decisions. Keep in mind, people use data to make decision. Hence, the tool you use to solve your specific problem; from statistical programing, BI stack to machine learning; is only good when it simplifies the data with the intent to solve business problem. Stay flexible on your tool choice, to ensure adoption. Ultimately, we have now reached an inflection point, wherein data strategy and its deployment has not remained a “nice to have” but a “must” to survive the dynamic market forces. Is your organization ready?