Cognitive Dissonance between Usability and Engineering: How it applies to Big Data Outcomes
My Artificial Intelligence professor had a very interesting artifact on his desk that looked like an ordinary piece of rock to me. He had labeled it “The world’s fastest computer”. Underneath he had a note indicating that it had the world’s worst I/O. The lesson he wanted us to take away from it was that Engineering and Capability is wasted if it is not usable without ET like powers.
The other day I was in a very nice rental car with a colleague. This nice car has a sophisticated radio with touch and other tactile controls. The radio came on when we started the car and we both spend about 30 seconds each trying to figure out how to turn it off. One of us finally figured out how to lower the volume. For the two days we had the car we never touched the radio or volume control again.
My TV Set-to-box is probably a very fine piece of engineering. It allegedly allows me to control what I watch, view my phone logs, decide what I do with my phone calls, manage the alarm and security settings for my home, take a look at my basement with the included night vision pan and tilt cameras, view recorded shows; and many more things I will never use it for. I prefer instead to use my computer or tablet to do most of these functions including searching for and playing shows on my TV just because I would rather not deal with a 1940s type user interface designed originally to switch between 12 or so channels.
These are just three examples that demonstrate the important role that User Interface (UI) and User Experience (UX) play in technology’s growth and adoption.
In deploying Big Data Proofs of Concepts (POC) and Projects, I have noticed a pattern where the best tool or the perception of success from a business perspective may have less to do with the engineering or the capabilities of a tool and more to do with the ease at arriving at a pre-determined outcome.
I have been in many environments where there is a consensus about the infeasibility of an outcome or the incapability of a tool based on a “POC” or trial using downloaded software executed by a non expert. There is nothing wrong with executing a trial for experimentation. However, I would challenge that there is a stark difference in the ability of a tool to achieve an outcome when used by a trained professional vs. an individual trying to prove the capability of a tool while learning to use it. In the later case, I would suspect that often training as well as the trial turns out to be ineffective. Guess who gets blamed? The sour grapes being the tool of course.
There is the other scenario where an IT department purchases a tool based purely on a checklist; baking tools off for alleged technical capabilities. The typical scenario is to retro fit business processes around a COTS package and make it work. However the reality is that if usability from a business perspective isn’t factored in from the beginning, the success of such initiatives and sometimes the success of an organization may become suspect.
Agility and usability is the key to success in Big Data and Analytics, and yet IT organizations continue to ignore UI & UX at peril to their own success and the business as a whole. The impact and cost of ignoring usability is clear if you know where to look. It comes in many forms starting with unrealized ROI on Data Warehouse and Analytics projects and continuing deep into the business where Shadow IT and other technical groups thrive without support from Enterprise class tools, processes, procedures, or compliance. Finally there is the cost and risk that comes with unrealistic policies forcing the business to adopt unsupported tools and processes.
An outcome that is achieved deep within the bowels of IT while failing to achieve business adoption is meaningless. Success is when an end is achieved allowing its exploitation by the greatest number for the greatest good of the organization. Good outcomes are achieved in a collaborative manner where an organization bring in its best, its brightest acknowledging, its gaps, and filling them the right level of qualified even if external support and then working on adoption in a transparent manner. A good UI that provides a strong positive UX is key to adoption. Without adoption, there isn’t much point or a business case for a purchase or an implementation.