2016: The Year of Data and Relevance
Last week, I read a post by Tim Crawford, a former CIO, who proclaimed 2016 as the year of data and relevance. Although Tim meant many things from his pronouncement, I was left wanting to answer an interesting question. How can CIOs and their teams create the data needed to achieve business relevance? My assumption here is that CIOs taking this step will increase the business relevance of their IT organizations.
I believe that there are two elements to achieving this end. CIOs need to create relevant and trustworthy data. Relevant means data is put together at the right level and the right time for decision makers but trustworthy means that the data is safe for decision making. Both are required.
Making Data Relevant
The first step in making data relevant—especially, in the era of big data is making sure that the right problems are addressed by a firm’s data and analytics. This is the case whether the data is for descriptive, predictive, or prescriptive analytics. This does not mean that every piece of data or every data relationships needs to figured out in advance. Instead, it means that there are substantial business problems that if solved will materially change business outcomes. As I have indicated previously, this requires that business problem versus applications determine the data collected. Increasingly being relevant involves as well means appealing to the needs of new data stakeholders and this can mean bringing in new sources of data including the so called IoT (Internet of Things) data.
Many of the so called big data problems are focused on being personalized and relevant to the customer. This requires data but as well it requires a fix customer data. Here mastering is critical to seeing cross business unit relationships. And more importantly, it is needed to make sense out of customer social data. Clearly, relevant data enables analytics and improves the productivity of data modelers. It sounds amazing but most of a data scientist time is actually in prepping data and then discovering data relationships. Relevant data is ready to go and has the critical statistics regarding itself and show the correlative effects which are necessary to building an effective predictive model.
Making Data Trustworthy
With relevant data, it is essential that we do one more thing and that is that we make the data trustworthy. There are two elements to doing this. First is we need to be make data consistently timely. I learned this in a startup that I co-founded. At the startup, we did manual pulls of data weekly. I was then responsible as the head of products for putting the data into a visualization tool for our customers. As I discovered from my early adopter customers, today’s business problems want to be solved with data and this means the currency of data has to change.
Over time, our customers asked whether we could provide the data daily and eventually in a semi real time fashion. They want this because they want to respond to business changes as they happened. With consistent data, we need to tackle next the quality of data as well. Most data has inaccuracies in it or with new data sources data can even lack metadata or have source inaccuracies. Furthermore, data has a life and over time becomes without a system more and more inaccurate. Recently, the CDO of major company said that data is like a dead fish, the older it gets the more it smells. If you want trustworthy data, then you need to fix both issues.
So there you have it, CIOs that create relevant, trustworthy data drive relevance for their teams and themselves. These CIOs will achieve what Tim Crawford proclaimed for 2016. They will achieve the year of data and relevance.
Blogs and Articles
Hypercompetitive business demands data driven orgs and CIOs
Will the data driven inherit the enterprise?
Taking charge of your digital transformation starts by fixing the data it is built upon
Your data is speaking. But are you listening?
How to be a data ready enterprise?
The nature of business is changing. Are you ready?