Taking data for granted

Hello dear readers. I haven't written in a very long time. Life can get busy. And on top of that, the German heat doesn't help me to have clear ideas. But today I felt like sharing with you, my invaluable audience, a situation I experience every day. So I hope my neurons are still working and I get to write something entertaining.

First let's take a very quick English class. I'm not a native English speaker and many of you might also not be. So I though copying the definition of "taking something or someone for granted" is a very good first step for my article:

1. To consider something as being innately or unfailingly true, correct, real, or available. 

2. To underestimate or undervalue someone or something; to not properly recognize or appreciate someone or something. 

I find these definitions more than suitable to describe the attitude some people have towards data. Because people is taking data for granted, they believe that it's always correct and available. However, they clearly underestimate the effort that needs to be done in order to get to a state where data is actually true, reliable and explorable. Many technical and business people assume that developing an application is a synonym of having data. Or, what's worse, people presume someone must have taken the responsibility to capture data properly.

In this last sentence I incorporated two English verbs that complement the title of this article. To assume and to presume. I don't want to all of the sudden become an English teacher, I just ask Google to make sure I use the right words:

Assume is a verb that means to suppose, to take for granted, to take upon, to don, or to undertake. In the shared meaning of “to suppose,”  presume is usually used when you suppose based on probability, while  assume is used when you suppose without any evidence.

Usually people assume (yes, suppose without any evidence) that because everyone knows data is the new oil (such a cliche phrase), a huge part of the development of a new capability, is with no doubt tagging, data capture and data integration. Some people even take their assumptions to the extreme and think that consumer centric data is generated without even asking for it. Business people presume, maybe based on the fact that 1 out of 10 (10% probability) of the developments before included data capture, that data will be there. Maybe is it because people is in general religious? Is the faith on Data quite similar to the faith on God? People cannot see it but anyway strongly believe is there? Well, I'm not religious and therefor I also do not believe that data will be there, no matter what. I always ask for it.

While doing research and making sure my topic has not been discussed numerous times (or at least to bring a fresh view on it), I found this article very representative of my work life (although I'm not a data solutions vendor):

"people may assume they can easily access all data. In reality, if data connectivity is not managed effectively, we often need to beg borrow and steal to get the right data from the right person. If we are lucky. In less fortunate scenarios, we may need to settle for partial data or a cheap substitute for the data we really wanted. And you know what they say, the only thing worse than no data is bad data. Right?"

You might not believe it, but actually begging takes considerably a high portion of my day. "Can you please tag the link so we can track it?" "Would it be possible for you to integrate your data to to the data base so we can easily access it?" "May I kindly ask you to share the data structure of the application?". And keeping on the track of properly using English vocabulary, do you know what beg means?

"To ask for (something) in an urgent or humble manner"

And again, the use of this word in its full sense is more than suitable. We have to always ask urgently because we get the question from our business stakeholders, that assume the data is there, one day before an important board update. And humbly, because apparently asking for data is such a terrible thing to do.

But where are the processes failing? Why do developers do not realize that even before writing their first line of code, data flow design should be covered?

Well I strongly believe it's based on the lack of clearly defined KPIs and the involvement of analytical teams or people in the early stages of the conception of a new capability. Or even the true reason is that, even in the era of Big Data and AI, developers and business owners are not analytical enough, yet.

Usually, or at least in my experience, analytics teams are involved once the strategy has been defined. We are called once our stakeholders already worked on the design and when it's time to measure some KPIs. "Ah, you don't have the data? but we generated data, we have it in the backend. Is it enough for you to have a TXT file? I'm sure you can parse some XML".

And why is it so? In my opinion, the main reason is because in some companies Analytics is too far away from the development and has too scarce resources to cover all the needs of data driven thinking. Too far meant in terms of organizational structures. And the second reason is because Analytics is in fact another team, when every person should be an analytical team member.

So how can we get to a stage where there's no need to involve another team to make sure there are no assumptions made on data? By actually educating every company employee on the need of data, on how to think a new development based on the business questions they would like to answer in the future and what it takes to get there. Basically exposing people to the ugly world of data.

When I joined Unilever, millions of years ago as a Trainee, I spent one week replacing and cleaning products in the shelf of a supermarket to understand the value of presenting products to the consumer in the right way. In the era that we are living, I believe that any employee joining any organization, on any level, should spend one week at least on doing a basic training on data analysis with a real case in the company. You would expect this to be covered by basic education, math class somewhere?, however you should never take analytical mindset for granted!

(Opinions are my own and not the views of my employer)