The hype cycle sure is in full swing when it comes to the importance of data. We see headlines declaring that data is the new oil. We hear analysts talk about how to hire a good data scientist. The taunts of our peers echo off the hallways: "If you don't hire a chief data officer, you must not be one of the cool kids!" Well, ok, maybe things haven't gone quite that far yet. But still.
Let's get one thing straight here. Data is nothing new for IT. It is, after all, the "information" part of information technology. Yet we spend so much time contemplating data science, analytics, and data lakes. My observation is that very little time is spent on the basics of data.
It's a good question, perhaps the most important question about data. If we don't provide business value, why are we gathering data or engaging in analytics at all?
Why Data Is Like Raw Produce
Here are some equally important questions:
- How and where do you collect data?
- What is IT's role in collecting data and -- more importantly -- its empowerment in doing so?
My point to the #CIOchat group, and my point to you, is that if business value is the meal, and analytics is the cooking, then good data is the raw produce. How and when we collect data matters.
Other IT leaders chimed in on the #CIOchat.
David Chou, a healthcare CIO, said that the first part of data collection is to identify the problem that needs solving. "People want data, but they can't figure out what to do with it once they get it," he said.
Mark Thiele, a data center executive with Switch, said that opportunities for identifying data sources likely come from business units, and that, ideally, the CIO is involved in any system creating data.
Ideally. Aye, there's the rub. I have never worked at an organization in which IT had complete knowledge, or even authority, over every system that creates data. There is always a bizarre use-case, or a lack of acknowledgement that data may be valuable at other parts of the organization.
We hear standard lines associated with this bunker (and bunkum) mentality among our business colleagues: "We don't need IT. You'll just slow us down." "This system isn't on the network, so it's outside IT's purview. You're just wasting your time." "We're the only ones who will ever need this system. Why complicate things?"
Wrong, wrong, and wrong. IT -- or whatever IT morphs into -- is the only steward of data in the enterprise. Think finance or HR is going to take this on? Nope. No matter whether it's called IT Services or Digital Services, we're the digital advocacy department.
When we accept excuses for not sharing data throughout the organization we represent, we reduce the future performance of our organization.
Locked Data Blocks Opportunity
If you're not convinced, let me give you an example of opportunity blocked, versus the success that's possible if you unlock data.
First, the example of unrealized opportunity. A colleague was at an organization that replaced an antiquated pneumatic thermostat system with a modern electronic thermostat system, where everything can be monitored and controlled from a central PC. However, IT has no access to the system, PLUS the system is proprietary in nature. Good times.
To be fair, and to Chou's point, if there's no perceived business need, why should IT or digital services or anybody geeky get involved? The answer is simple: The data is there for more than the people who are running that particular corporate silo. Maybe the CEO's office or finance would like to know whether asking everyone to put their thermostat to 67°F during the winter would actually save any money. Maybe everyone already does, and this initiative would be pointless -- that's the power of data.
But here's how these data requests go: At a point in time, an executive requests data to inform a decision that needs to be made. The silo department says, "Beats me," when asked for the data. IT says, "We don't have it." Then the executive makes a decision by the seat of his or her pants. Not ok.
The raw data must be there when it is needed, or it is useless. You don't start a data collection program exactly at the moment when the question is asked. By the time the program is launched, the question will already be irrelevant.
Now compare this to an experience that my staff and I had with our transportation department, its garages, and its parking systems. The transportation department was hungry for data coming out of its proprietary system. This isn't because they're data wonks. It's because they had an additional public service in mind: showing citizens how many parking spaces were available in our garages, in real-time. The garages had hard-wired signs that showed customers how many spaces a given garage has available, but that doesn't exactly help people who are on their mobile devices.
The software vendor couldn't provide that, except in a PDF report.
Data does not mean a PDF report. Data, as the "good guys" in government now know very keenly, means machine readable, tabular data.
There was no API on this system that would let us easily grab the data. So, one of our enterprising analysts started monitoring the SQL backend and the file system for changes. He eventually figured out where the data was being stored and how we could store it elsewhere and make it available to others.
Because we did this, we were able to create a website to show current garage availability, and others in our community were able to create a nifty Web app based on that data -- one that also shows historical data -- for the public.
When decision-makers ask a question about parking, there's a wealth of data to inform an answer.
The takeaway? Too often, either because of a lack of priority or because it's perceived as a techie thing, business partners decide they don't need to inventory data in the same way that they might inventory fuel or supplies.
So, back to basics. Given the higher-order decision-making that is now possible with analytics, I believe that every IT department -- whether in government, healthcare, or the private sector -- has a call to action to ensure that, when data is being gathered, it's stored in a manner that allows for future decision-making.
That means tearing down silos. That means difficult conversations. The good news is that IT has been having these difficult conversations for years when it comes to app standardization and security.
Now it's time to gently, but firmly, curate your organization's basic data. Go.