Big data blues? It's all about the 'so what' question: Page 2 of 2

Know what you need
 
DNA: Specifically, what kind of technology has contributed to big data? Could you share with us what you think is the inflection point for big data?

Finger (pic,  right): I think the big change was the onset of the Hadoop and NoSQL database, which allowed us to store massive amounts of data and still process it in a quick and agile manner. What has changed is the technology to store more. It has changed the way we work, access, and store data. Analytics, per se, hasn’t changed that much. The big step is our ability to digitise behaviour, content, and the ability to access that data.

[Basically] you need tools that are commensurate to the task. If a company does not know what it needs, it can’t possibly know what tools it needs to complete those tasks.

But that isn’t the most important issue to grapple with. There are a lot of people who try to sell technology which you don’t really need. For example, if you need to commute, you may need a car. But to commute on the roads, you don’t need a Lamborghini. You may only need a bike. Companies may not need those tools. What they want is money, they want better business models, improved processes, reduced costs, to be faster than their competition. Right fitting the companies with tools is more important than the technology itself. If the company needs a bike, don’t buy a Lamborghini.

DNA: But yet there is so much hype surrounding big data as there is for many other nascent technologies. Hype is a consequence of newness but it shouldn’t overshadow the reality of the technology. What is your view on this?

Finger: It’s not as if having big data will save your life. In fact, the truth is that having more data is painful. Why? Because you’ll need people like me – data scientists who are very expensive to hire – to actually make sense of this data. An organisation can have a nice big data dashboard but so what? As long as it can’t answer the ‘so what?’ question, I would argue that it doesn’t need big data. You need to understand the ‘so what’ question first.

Some would argue and ask: ‘But what about the knowledge discovery process? Isn’t that important to find data correlations and get an idea of what to do with the data?’ The problem with knowledge discovery is that you need to have an idea of what you’re looking for rather than just finding things that correlate and concluding from there. Knowledge discovery is useful to generate ideas – I too practice it to get ideas. But the thing about the business world is that you can do this to get ideas but you still have to match a business plan to it. You still have to know where you’re going with it.

DNA: So how do you match ideas to business plans? What are the best practices?

Finger: Firstly, you want to listen to the data insights and make your decisions measurable. Develop a way to control how you make those decisions in order to become data driven. Secondly, organisations must enable data people to talk to business people about the data. Forget the big data buzz. As soon as a business person talks to a data person who can help you with the data, wonders will happen.

A business person will have questions and the data person might have ideas on how he might be able to answer those questions. That’s what you need to do. Enable them to work together. Very often a business person would need to open a ticket just to speak to a data person.  Break down the silos and just get the two to talk and figure it out together.

Once they have an idea, they’ll need to make the decisions measurable. If a company embarks on building a new product, they should want to know how it sells. So we should apply the same rigour for everything we do [in big data] and constantly re-evaluate whether it is looking at the right things in a measurable way.

DNA: What do you think are some of the impediments that stymie big data projects?

Finger: Not knowing what they want in the first place. If someone wants to be data driven, don’t tell him about Hadoop. Ask him what he wants to do. How does he want to improve the business? How does he want to make money? If he doesn’t know go back and ask. Think about it. If a customer doesn’t know what he wants to do, how will technology vendors know? This is why big data projects fail. Because they have no clear idea of what they want to do in the first place.

DNA: So how do organisations avoid such pitfalls?

Finger: Focus on the ‘ask’ question first. Ask yourself the ‘so what?’ question first. This is why you have product managers in any engineering company. When engineering something, you need to know where you’re going. You just can’t keep changing the design. It’s very expensive if you keep doing this. This is why you have product managers – to think about such things. The same with analytics and big data.

DNA: What is your opinion of the state of the adoption of big data in the region?

Finger: No matter where I look, everyone I come across is just starting to use big data for business. If someone says he has loads of big data, my question would be: ‘Is he making lots of money? Or is he going to conferences and telling everyone he’s big on big data?’

The thing is, if people are already making money with big data, they would not be telling others about it. They would just be making money. For example, there is a trend where people try to use what is said on social media to try and predict share prices. A lot of companies have spoken about this. If anyone can predict share price [movements], would he speak about it? He would just make money. But if he just talks about it and how amazing big data is, he’s probably not figured it out yet.

Related Stories:

Malaysia focuses on building a data driven Asean

Big data benefits recognised in Asia but potential untapped: EIU study

Malaysia needs to accept failure: Teradata CTO

 
For more technology news and the latest updates, follow us on Twitter, LinkedIn or Like us on Facebook.