Wednesday, June 25, 2014
How many of us have endured endless wait times when calling customer service numbers? Or suffered through long lines at a POS counter? Or even waiting for a website to load and log you in for online self service. All of these are cases of "traffic jams" in the customer services of an organization.
Tuesday, June 17, 2014
But what happens with this data?
In high school my physics teacher defined data as the "raw form" of information. He said, "Information is useful, but data on its own can't get you anywhere." These words ring so true with the large amounts of data that we encounter every day.
How do we transform this data to generate actionable information? (or do we?). Well, that depends on the type of data that you are looking at and more importantly, the type of information you want to get. This second one is tricky and I will talk about it in a later post.
I like to divide data into these buckets as it helps to understand where this data can be used and what type of information can be derived from it.
Descriptive (Adjective) Data:
This type is data collected about someone or something. This is like the "adjective" in a sentence. It describes a noun. This would include data like: Name, Address, Phone, Email, Height, Weight, etc about a person. If we were talking a product or service, this data would describe the product or service. For example, the photo of a product can be Descriptive Data.
This data is a great source for Category Information. It is great to use when we are segmenting information or trying to find relations between many sets of data.
Transactional/Operational (Action) Data:
Like the name says, this data is about an action or transaction. This can be from an documented operation or from ad hoc transactions. Some examples of this data are: sales transactions, website visits, financial investments, etc. Since these examples about the "what happened" aspects, there is a lot of this type of data out there. While the Descriptive Data does not change a lot, Transactional Data gets created and changed very frequently.
This data usually ends up as Measures in reports. Because there is so much of this data, there are different ways to work with it. But that is a story for another post. From the perspective of information creation, this type of data plays the part of describing the action. This data is usually summarized in some way during the information creation process.
Analytic (Strategic) Data:
Analytic Data is the twilight between data and information. It is so close to information that many people use just the analytic data as the definition of information. The fact we call it data means that it is still in its raw form. I like to explain this type as data about the relation between descriptive and transactional data. A very basic example of this would be the something like "teenagers spend more than 60% of their time in front of computers". You can see how this can easily be confused as information. In some way, it is information if you are looking for only this level of transformation.
However, to become truly strategic, analytic data talks again in "aggregations" of relations. We talk about relation between categories (from descriptive data) and measures (from transactional data, or sometimes descriptive data as well). The true information is derived from transforming these relations to real-world trends which can help make decisions. Now, you can make a decision based on the example I gave above as well... but strategic decisions are often not that straight forward. So I term this still as data rather than "information".
Predictive (Potential Action) Data:
Now, Predictive Data is unique in the sense that it is about transactions that "may" happen. This data is not yet a "fact" it is still a "theory" of an action that can happen. Given the human nature of wanting to know the future, this is the type of data that is most sought after. Again, this can be easily be confused with information as you could possibly make decisions based on this data.
Technology today has come a long way and quite accurate algorithms can be found that can predict actions based on trends in the market.
It is quite evident that there is a lot of data that is generated these days (a phenomenon that some have come to call "Big Data"). However, the challenge that I see today is the transformation of this data into actionable information. Until you sit down and sieve through the data and transform it, all that data is useless. It is like sitting in rubber boat in an ocean without a drop of water to drink, hence the title of this post.
Any reporting or data visualization initiative should always focus on transformation of the data into information that can be consumed and is relevant. Get rid of all the clutter and noise, data is the most important part of any piece of information.
Think about it... and share your thoughts... I would love to hear them.