Big data analytics balancing acts for the new c#o

  • BLOG
  • July 29th, 2019

You’re in charge of big data analytics in a large company. Are you the CIO? Chief Analytics Officer? Chief Data Officer? Chief Data Scientist? How about the CMO or CFO? You’re likely one and more of them at the same time, thus my new title for you: C#O. 

Companies like yours have made enormous strides in getting a grip on their data, analyzing it, and monetizing the resulting insights. But today’s big data analytics climate still feels like the Wild West. Models for governance, prioritization, big data analytics tools, and information architectures, talent development, and approaches to problem-solving, are still ill-formed or even absent. If you’re that C#O, the CEO is likely pointing at you now to figure it out. And you’re not focused in the right areas.

At Mu Sigma, we conduct a lot of big data analytics, machine learning consulting, and decision sciences workshops. In the lead-up to each workshop we survey participants and ask, among others, this open-ended question, “What’s your top challenge with respect to big data analytics?” Two types of response usually emerge: concerns over data and concerns over big data and concerns over decisions – that soft stuff. The following table captures the phrases that emerged most commonly from one-hundred responses across 2015:

Is it all about data?

Or decisions?

Disparate sources of data

Prioritizing where to apply analytics

Multiple sources for one metric

Finding scalable tools without high costs

Legacy vs. new data systems

Scaling our big data analytics capabilities

Usable data in a consistent format

Best practice sharing across silos

Adding new data when busy with existing

Identifying the correct metrics

Master data management

Creating actionable insights

Information governance

Change management and talent

Table 1: Top Data & Analytics Challenges (N=100)

Your background – whether from IT, marketing, or a new analytics group – might determine where you concentrate your efforts. But we see a lot more emphasis being placed on Data at the expense of that Decision stuff on the right, and usually under the guise of “we can’t get anything done until the data is all set.”  As C#O you can’t afford to accept that premise. So I write this post as a set of guideposts for you to use in striking a balance between big data analytics and Decisions.

There are five balancing acts.

#1 Balance creation with consumption.

You’ve seen some form of a chart that shows how the production of big data is far outpacing our ability to use that data. The same holds for big data analytics. Various teams continue to create more metrics, more dashboards, more supposed insights, but not all of it is desired nor driven by those folks making business decisions.

Whether you recognize them or not, there are thousands of decision supply chains that operate in your business each day, each chain spanning from problem generation to data gathering to insight generation and evangelization, with other steps in-between. As C#O you should formalize such a construct, to ensure visibility across all links in the chain, so that you’re only creating analytics that involve business stakeholders upfront, and you’re feeding your learnings from how the business leaders use your tools and insights, back into the creation work.

#2 Balance answers with questions.

Your team is under pressure to provide accurate answers, and fast. But do you know if they’re asking the right questions? If you’ve ever played the game 20 Questions with a child, it can go something like this:

“Is it a person?”


“Is it alive?”


“Is it a cucumber!!!”

As humans with hectic lives, we jump too quickly to answers without slowing down and asking the right types of questions. But asking better questions can lead to a better solution design for a problem. In some cases, you’ll want to expand your view of the problem, rather than keeping it narrowly focused. In others, you may want to challenge basic assumptions or affirm your understanding in order to feel more confident in your conclusions. Give your teams the time and training on what good questions look like.

#3 Balance solving problems with problem solving.

Similar to the constant pressure your teams are under to find answers, they’re under pressure to keep up with a daily barrage of problem requests. With their heads down, who is responsible for helping them hone their own craft? Who is responsible for the art and science of problem solving? It’s you, the C#O. That craft may look unique to your company, but to us at Mu Sigma, it involves problem-solving processes like:

  • Always beginning with measurable outcomes in mind, and charting the behavioral changes required to drives those outcomes
  • Encoding problems in a consistent manner, one that emphasizes asking good questions, and drives community of practice across your organization
  • Mapping and understanding the interactions between business problems
  • Integrating different forms of big data services and analysis – descriptive, inquisitive, predictive, and prescriptive – problem by problem, so as to drive consumption and also avoid analytics silos

#4 Balance governance with empowerment.

Governance in an analytics context can be as convoluted as the job titles that funnel into the C#O, especially because analytics governance must now dovetail with information governance and sometimes data governance. Regardless of model, most will focus on matters of risk mitigation: decision rights, policies, and standards. And this is appropriate, considering the number of threats looking to exploit your big data analytics or even legitimate users making poor choices with respect to your data.  

But keep in mind the role that Gartner calls the Citizen Analyst. They’re out there. They’ll create new Tableau cockpits, hire their own big data consultants, and procure their own modeling tools. You won’t stop them. But you must find a way to give them some fuel to go along with the guardrails you set up for their work. That fuel can manifest in the form of making a clear litany of services and tools available, offering flexible resource capacity, helping to assemble funding requests, even training on different problem solving techniques.

#5 Balance technology focus with a focus on culture.

Technology and culture aren’t exactly flip sides of the same coin, but it’s easy in the manufacturing analytics Wild West to fixate on tangible items such as software tool functionality, training, and support. The shift from data to decision science requires a lot of cultural introspection. Seek a learning mindset, ways to experiment in a fail-fast-and-often-but-cheaply mode, make sure your analytics efforts constantly reinforce tight feedback loops, and don’t take a discrete project mindset but a problem-solving mindset that will last indefinitely.

Data Orientation

Decision Orientation

Create analytics and insights   

Drive consumption of analytics

Find answers quickly

Ask the right questions

Solve specific analytical problems

Improve the approach to problem solving

Govern and mitigate risks

Empower the frontlines

Scale technology effectively

Drive a culture of experimentation

Table 2: The Five C#O Balancing Acts