Finding Value in Analytics, Part 1: The Journey
Frustration with Analytics
As an IT professional with a BI/DW focus, I am frustrated by our current analytic-crazed industry. I feel like we are losing critical business perspectives on why and how to apply technology to help our organizations. It often seems like technology for technology's sake! …especially with analytics. If the analytics is possible, then we must do it! What happened to good olde fashion business-savvy for running our companies?
In response, I am writing a blog series that will focus on Finding Value in Analytics. I am proud to share these thoughts with the assistance of my colleagues at the Eckerson Group. This series will be a journey of searching as to why we should invest resources on analytics and, in fact, any form of IT. This series will set the context, outline the principles, and suggest practical ways for investing those resources.
I hope to relieve my concerns, especially with the use of analytic systems at scale. Maybe some of this will help you think through the same issues and do something amazing for your organization.
Take-Away: Have the same feelings about Finding Value in Analysis? Then, share your thoughts and experiences with the colleagues around you and with the readers of this blog.
Let’s start by setting the context for this journey… What is the economic value of analytics and any other form of IT?
The journey to find this value began many centuries ago. As soon as humans measured and recorded data in conducting commerce, information was made tangible, solidified from verbal conversations and individual perceptions. When used to conduct commerce, that tangible information generated value.
Since the 1950s, organizations have spent considerable resources on IT to digitalize, organize, integrate, describe, and generalize information about their operations. Those early days were simply motivated by increasing efficiency and reducing errors in what was purely manual clerical processes. Since then, IT has evolved and matured in numerous and amazing ways.
Take-Away: So, you ended up in a career caring for a bunch of computers! Remember and share some of the amazing (and not so amazing) ways that this technology has changed over your career.
However, the economic principle remains the same. As data refined and consumed by humans, information has economic value only if organizations are enabled to make actionable choices that yield higher expected payoffs than if those choices were made in the absence of that information. The implication is…to have value, analytics should enable the organization to execute improved actions that, in aggregate, modifies organization behavior toward customers and the like. The objective is a positive net value.
Take-Away: Poke around your organization for ways that analytics is being used. In each situation, what are the new actionable choices that analytics has enabled? Were these choices NOT possible before? What were the actions taken? Was there a real change in organizational behavior? Did these actions yield better payoffs for the organization? …for your customers?
Be Aware (not beware) of Turbo-AI
Recently, the ability of analytics to transform raw data into information has improved so much that payoffs are exceeding Human-Level Performance (HLP) for specific tasks. Research into neural networks now documents such examples almost weekly. To state it concisely, we humans have created tools smart enough to out-smart ourselves! Let’s refer to this type of artificial super-intelligence as Turbo-AI.
Take-Away: Pay attention to Turbo-AI developments, by regularly scanning AI coverage in Forbes, O’Reilly Media, and MIT Technology Review. Note instances of controlled comparisons with HLP for use cases relevant to your organization. Imagine the possibilities, both positive and negative.
Hence, questions about analytic value have a deeper significance than ever before. As society increases in complexity, one can imagine a future in which turbo-AI as artificially augmenting human intellect becomes as valuable to human existence as water or air. If you are feeling uncomfortable with this statement, then you are understanding the implications. We need to take this topic seriously. It is way more important than justifying your pet neural net project to the boss!
As a society, we can view the emergence of Turbo-AI as both a positive and negative force, much like electricity or nuclear energy. IT professionals are today making decisions daily about whether and how to invest in analytics within their organizations. On what basis are those investment decisions are made, especially amid the rapid flux within analytic technology?
Take-Away: Do you have this same uncomfortable feeling? Are you torn between Turbo-AI’s extreme potential for good for society and its dark opposite? We all are! Engage with your colleagues in discussing these issues. Make it a safe and open discussion to have within your organizational culture.
Where are We Going? Are We There Yet?
Where will this value journey take us in future blogs?
In the next, we will review the last 70 years of IT evolution, with a focus on the value of information that results from IT investment. This retrospect has clues as to where our journey SHOULD take us to find value. The value journey will continue by examining topics like these …
What are the cultural disconnects within our current IT organizations that prevent organizations from realizing the value from analytic systems?
What is an analytic? Where is the boundary between describing known data versus generalizing beyond known data?
Should we create new forms of data literacy dealing with analytics, as targeted toward executives, IT professionals in general, and business/data analysts specifically?
What are the inconvenient social implications and ethical issues that we face using analytic technology? How should the value proposition change? Like the role of carbon credits in managing climate change, what is the equivalent of carbon credit for regulating Turbo-AI?
Are there any practical suggestions for monitoring and managing analytic systems at scale, especially as Turbo-AI technology becomes compelling?