1. Data Analytics
  2. 87 views

About a week ago I was meeting with some of our analytic staff discussing differentiated strategies for distinct markets. It was one of those rare, fun conversations that extends beyond tactical needs of a project and really allows you to dig into strategy and (gasp) academic theory behind a methodology.

During the course of the conversation, the word ‘analytics’ was tossed around as it so often is. After what I considered to be enough patience, I finally interrupted the conversation to make a point. So many people use the term analytics for such a broad range of activities that it has functionally become meaningless. I have heard the term used as a description of the most basic reporting function; I have heard it used to describe the aspirational, strategic goals of some of the smartest marketers I have met. I wanted to point out to my young Padawans the impotence of throwing that word around, especially in the presence of an ‘analytic’ audience.

The response I received surprised me:  I was asked, “So what does analytics mean (to you)?” I am sure many of you have been in a similar situation, where you have adopted a position (and probably it was well founded or researched) and had your talking points, but you haven’t really exhaustively thought about that position for quite some time. I didn’t have a rote answer or elevator speech to describe the essence of analytics. Sure, it is generically used to describe anything that touches data, and sure it is often used in a way that minimizes what it is that I do for a living, but when asked directly how I define it, I didn’t have a rehearsed response.

That forced me to consider what it is that analytics is…and do it quickly (can’t look unsure in front of a team that thinks you have all the answers). What I realized was that I am not so much opposed to calling something analytics incorrectly as I am opposed to calling everything analytics simultaneously. Analytics is not a generic term that fits in all situations, and from that is where my discontent with the use of the word stems. Analytics is a process that includes the ever-maturing use of data and statistics (maturing in the sense that it starts with a very simplistic representation of data and grows in sophistication and application) to truly understand and solve real-world issues.

Luckily, I came up with that answer in just a few seconds because I would’ve looked very foolish lamenting the use of a term if I did not have a practical definition myself. As we discussed the analytics process with the team, I realized my definition of analytics was actually a hierarchy that included four discrete activities. When combined, these activities result in the best that practitioners of analytics have to offer. When viewed discretely, none of these can stand-alone as a shining example of ‘analytics.’ Therein lies my dissonance with the standard use of the word. It minimizes the process and the value when analytics is decomposed to its base elements.

Functional Analytic Hierarchy

Similar to Maslow’s Hierarchy of Needs, analytics has an evolving hierarchy and the higher level needs cannot be met unless lower, foundational levels have been addressed. Each of the four analytic needs plays a key role in the delivery of value to an organization. The four levels, in reverse order of sophistication include:

1.      Reporting

  • Defined: Putting numbers on paper (or screen)

  • Role: Significant effort and skill go into using data to accurately create valuable tables, charts, and reports. Beyond the technical skills required, there is much value in an individual that can concisely and intuitively communicate information to the lay audience.

  • Shortcoming: Reports are a summary of data. They provide a foundation for an individual to understand and draw conclusions. By themselves, they offer only as much value as the design allows – enabling review, understanding, and development of action plans.

2.      Analysis

  • Defined: Circle the highest number on a piece of paper (and tell the audience to ‘look at the circle’)

  • Role: Identifying what is important vs. non-important often requires a more educated and technical critique than most imagine. Data and reports can be misleading and/or erroneous based on oft-overlooked items like sample size, practical validity, relative importance, etc. Knowing when results are worthy of further exploration or action is a critical part of analytics.

  • Shortcoming: Analysis is a way to understand what is important. When done correctly, it preempts any spurious conclusions and focuses on items with real impact or importance. It does not however offer any explanation as to why the world is the way it is.

3.      Insight

  • Defined: Explain why you circled that number on the piece of paper and why you think it is important

  • Role: Insight is the bridge between data analytics and the business. It involves understanding not just the data at hand, but also the larger context of the business and creating meaningful hypotheses or conclusions that not just explain trends, but also suggest action.

  • Shortcoming: Insight does not prescribe future action, rather it leaves it up to the business community to approve or reject the hypotheses and theories and determine the appropriate market response.

4.      Consulting (Application)

  • Defined: Explain what the results mean in context of the overall situation and recommend what the business should do as a result of the findings.

  • Role: Recommend action and create the appropriate strategy (with test design) to execute and optimize. Requires a full understanding of the business, including the industry, competition, and internal concerns such as operations and politics. In a fully mature organization, this role has a seat at the executive table.

  • Shortcoming:  Scarcity of applied analytics professionals that have the data background, relevant business experience, and depth of organizational knowledge to be effective in this role. Additionally, this requires an organizational belief (with follow thru) in data driven decision making, which is easier to preach than to practice.

Effective applications cannot be discussed nor effective tests designed without the benefit of insight and the foundational pieces of the analytic hierarchy. Conversely, reporting without insight or application is little more than a data buffet where the business can (often errantly) choose to see and act on what it feels like in that moment. None of these pieces are independently sufficient.

Unfortunately, in reality, most of our ‘analytic’ time is spent lower in the hierarchy that we would like. It takes considerable skill, effort, and investment to derive meaning and value out of the multitude of sources and Terabytes of data that is present in every organization. We (speaking generically for the entire analytic profession, if you please give me the liberty) all want to be living in the top of the hierarchy all the time.

While data scientists and other analytics experts want to be operating at the top of the hierarchy, it is often premature or impossible until lower level needs have been met.  This introduces real challenges when building internal analytic capabilities. A senior level analytics hire is often impotent without a team of competent data professionals. But how do you hire and build an analytics team before you have a leader to guide them?

Analytics is not a ‘thing’ or an activity. It is a process, and a rigorous one at that. That (re?)definition has direct implications on that way we approach our work, on the way we hire and build our teams, and on the way we nurture and professionally develop each team member. In the latter case, teaching the approach and giving each team member the context of the business while they mature their hard skills is critical, as is transparently communicating expectations of each level of the hierarchy and coaching each individual.

Circling back to the conversation at hand, I asked my team to look at what they do in the ‘process of analytics.’ If they want to label or define themselves or what they do, I asked them to understand themselves in the context of the bigger picture. Are they closer to reporting or consulting? Is that where they want to be in their career? What value are they providing, and what would they have to do to provide value at the next level of the hierarchy?

If you are an analytics professional, where do you spend your time? Are you getting an opportunity to engage at the level that is satisfying for you? Does your environment allow you to not just practice foundational analytics, but to grow in sophistication and maturity?

If you are a purchaser (or employer) of these services, do you know where you are at in the hierarchy of needs? Are you getting what you need from your team or supplier? Do you often feel like you get a data dump and are left to do the higher-level analytic work on your own? That should have implications for who you hire (staff or supplier) and what you ask them to do.

So, what is analytics? It is the application of skills combined with a way of thinking that has become critical in Our Connected World. What it means to you is based on where you are at in your current company and/or professional development. But know, it is larger than a defined set of activities. And if it is not transformational for your business you are likely doing it wrong.

View Original Article

Recent Content