Big data and Internet of Things carry the promise that some billion connected devices will change our life and the performances of our companies. Each time we put too much faith in the tool, rather than in the human governing it, we are at risk of huge mistakes.
What are we really talking about? Big data is about three major layers, the systems, the analytics and the interpretation.
The systems are all about the technical configuration of devices talking each other, the cloud, the storage, the security of data. People still struggle to understand that the Internet of Things is going to change substantially how data are produced and their quantity: from vast amount of transactional data created when a human interacts with a machine (I do my shopping and pay at till the products in my basket), to a continuous wave of data generated by machines sending information each other, every instant (a thermometer collecting the air temperature each second and controlling the air conditioning system).
The analytics are about the people able to understand the ocean of data, mine it, make it understandable and visually simple to managers and operations not familiar to statistical models and software. These resources already work in our companies and have developed analytics roadmaps along the way, that are not a recent discovery. There is a gap now between demand and offer in the labour market of these resources that are intermediaries between data and decisions.
The interpretation is about the managers asking the right questions, discovering hidden patterns and finally taking the right decisions about their business, operations and clients in the end. I believe the three have to work in synch to maximize the performance, but obviously if you have a car (the system) a tank full of petrol (big data), but nowhere to go (who decides the destination), your journey is meaningless.
|This website is free, if you like our post do you mind sharing it?|
A few simple holy rules to remember about big data
Real time is not a joke. First of all, it is not easy to achieve and it costs to implement. Second, data has to be stored (logically or physically) and if you don’t mine it, trying to get some knowledge out of it quickly, it becomes obsolete. If you replenish you granary with a lot of raw material, but then you don’t grind the wheat, it goes rotten. Near real time is enough most of the times.
Text data is just going to increase the amount of data in circulation. Billions of devices connected in real time are going to download and upload to the internet and to your data wharehouse a huge amount of data. This is going to worsen when the data transferred will be text. The number of words posted on (just…) twitter every day, could fill a ten million page book. And I guess human feelings are embedded more in words than in payment transactions or temperatures. Cisco calls it “The Zettabyte Era”.
Correlation is not causation. Today I read that an artificial intelligence engine has discovered that there is a correlation between the month you were born and the diseases people suffer when adult. I bet there might be a correlation between the weather in Milan and the performance of the Nasdaq. I can tell you that most of the time is meaningless and leads astray. Managers are there exactly to investigate and drive the analysis in the right direction.
Don’t underestimate small data. Size in itself is not a measure of value, what matters is having the data, of whatever size, that helps us solve a problem or address the question we have. Everything processed in Excel, for example, is small data and will continue to be important: we are moving into an era of distributed models not centralized ones, so having small chunks of data integrated with other small chunks of data and elaborated by different parties, will deliver better results than centralized monolithic structures.
Algorithms comes at the end, not at the beginning. Big data and artificial intelligence are two separate technologies. Until you have AI, the intelligence of your leaders is what you have. Don’t expect analytics can uncover hidden patterns and show you the way to save money or make more profits; if managers don’t ask the right questions, big data is useless. Algorithms are already driving 60 per cent of the rentals from NetFlix, but the idea, the business model and the value proposition behind it has not been conceived by a machine.
Customer journey is the ultimate reference to manage big data. As companies continue to add more data sources to the mix, the number of potential data sets grows, as do the opportunities for new and different insights. And different experts can lead to different conclusions with the same data. That’s when the expertise of the managers kicks in and their guidance is simple, the client, because, in the end, the only perspective that matters is the customer’s.
Newsletter: because there’s muc more than big data here!
Feel free to subscribe the Futurist Hub Newsletter. All the updates, once per month, free of spam. And it comes with a free ebook written by me.