Big Data is Big Miss
Information-based analytics trump Big Data
I don’t use the term “big data,” which long ago became an overused buzzword anyway along with its cousin “The Cloud”. It’s not that I dont think data is important. Its just that Big Data by itself is like a basket of whole wheat flour without eggs, butter and milk. Just like you can’t eat just the flour, you can’t consume Big Data. You need a recipe to take the raw ingredients and a means to transform it into something yummy.
Recently, the Altus Group CRE Innovation Report showed that 89% of the firms surveyed faced major impediments to collecting and utilizing data to drive improved asset and investment management decision-making. This is in line with previous research that big data may actually be slowing down decision-making, as opposed to making it more effective.
The No. 1 goal of any CIO right now shouldn’t be achieving some big data-tied deliverable. It shouldn’t be implementing a massive data warehouse or finding a big data platform that can gather millions of information points for future analyses. Rather, it is about being smarter. One of my mentors, another executive at Cushman & Wakefield, says that an organization is often like a large brain using about 10% of its capacity. Harnessing the power of the rest should be the goal of any analytics program. So much of the information you really need for effective decision-making lies in the heads of thousands of professionals in dozens of different offices (or homes … or coffee shops … or cars) all over the globe.
The primary goal of a CIO, then, is finding ways for those professionals to come together, share the information they have, and solve complex problems. That may well involve a platform approach, but it’s not necessarily a big data platform.
I prefer to start with the questions, as most good strategy does. In our business, it all begins with: What are some of the models we might be able to build that help our clients more effectively manage their real estate assets? Once we have the right questions defined, we build the data models and refine existing — or create new — data collection mechanisms. In this way, instead of “big data” we can focus on “information-based analytics” that more immediately drive value in decision-making.
Silos are a large part of the problem
The Altus report found a lack of integrated data approaches. 80% of firms surveyed said their business could eliminate or reduce data silos through better integration and standardization. Four out of every 5 people complaining about silos, especially in an era where firms compete largely on data, is a very telling number. But it is also a sad reality. Many organizations are formed from a series of acquisitions. Over time, legacy systems build up and a patchwork quilt of interfaces is developed to keep them humming in unison.
But these patchwork quilts do not have to constrain or define your analytics strategy and practices. Instead, I try to think like a startup. Startups often gain that disruptive edge because their decision-making is better and faster, and they can move to market (and refine once there) quicker than an enterprise, legacy company. In the course of this happening, startups are often building their own analytic systems — as opposed to relying on third-party vendors. Why can’t this happen in legacy companies? The reason you hear most is “process”. And that is in fact a major problem. Sometimes we let process overwhelm actual notions of productivity, which is a bad play for all involved. Process should only exist to better business performance, not to hinder it or run your people in circles.
Data integration and standardization is obviously a challenge for firms, but the bigger challenge is a concept we don’t discuss as much: data model and taxonomy standardization. If one business unit thinks in terms of cost per square utilized foot and another one thinks cost per square gross foot but both simply refer to their data as cost per square foot, the analysis will be off because the data is off. Disparate systems with a common data taxonomy can get you pretty far. On the other hand, one global system with multiple data taxonomies can lead to bad analysis. You may have won the battle but you will lose the war.
This may sound corny, but it’s totally true: data is power, but only if used for good. Stephen Dubner, famous from Freakonomics, has been discussing this idea for years. We have a tendency in business towards more, more, more, but in this case it doesn’t work. That’s another reason I don’t like the term big data. Just collecting data essentially for the sake of having it, with no end goal in mind around improved decisions or processes, is complete folly.
The challenge for our industry is this: how do we take the lessons of the investment shops, insurance brokerages, and even the residential real estate business and translate it into what we do, while at the same time not losing the connection with tremendous local leaders?
I think we’ll continue to see more approaches around data — look at a model like Zillow and the amount of data they crunch — but my hope is that this idea of “big data” fades into buzzword obscurity and we focus on the right things at the CIO level. We should be connecting stakeholders and moving towards information-based analytics.
Be well. Lead On.
Adam
Related Posts:
Disruption: Thinking like our ancestors
Innovations Changing Our Industry
Adam L. Stanley Connections Blog
Technology. Leadership. Food. Life.
Follow me on Twitter | Connect with me on Linked In | “Like” me on Facebook