Last week we had the pleasure of sitting together with László Kovács, Managing Director at Cyberport who has been described as a man who not only does big data, AI and digital transformation, but lives these topics, too. We chatted about a topic that’s on everyone’s mind: ‘Big Data’ and what it means for the future of retail.
Generally when it comes to big data, there are two main approaches: gathering data to build products vs analysing the product to understand the data. So László, my first question for you is: from data to product or from product to data?
Well this is a great question. It’s not that easy to answer because, at Cyberport, we are doing both. Since companies started talking about Big Data over a decade ago, there’s been the belief that what is needed is to gather data, build an infrastructure to understand the data and then tell the marketing department to make use of it. At Cyberport, we think it’s better to work backwards.
First, we understand the customer through the data we have, starting to create a unique view of the customer and the different customer segments. Then, we start seeing, through our data, what their preferences are – as the end goal is always to understand how we can get better for our customers. Let’s also consider that in our industry, consumer electronics, it’s really different than other industries like fashion where trends, inspiration and the “artistic touch” are core drivers of the industry. For us it is mainly a matter of understanding the demand and how to fulfill it in the easiest way possible for our customers. Therefore, we work with data to understand what factors are relevant for different customer segments, like prices, availability, delivery terms etc.
Nevertheless, the approach of gathering data to build products is becoming more and more important in order to understand the data flow in real time, without even necessarily storing the data. This can be extremely useful to analyse and understand your processes and how you’re performing throughout your organisation.
Another point in big data, other than what you mentioned about using data to build products, is using it to predict trends. How does that work in your specific case?
Let’s talk about something easy to relate to: this year.
2019 is forecast to be a difficult year due to different micro trends we can’t fully forecast as of now. For example, the unclear effects of Brexit and how this will affect the value of the Pound, or the uncertainty with Italy and how the rising debt of the country will affect the Euro sector. Still, these are factors we somehow understand and we can try to predict.
On the other hand, there are other completely unpredictable factors like the climate because we don’t know how the summer will develop. In 2018 for instance, we had one of the hottest summers in years, in Germany. How’s this related with eCommerce Sales? For Cyberport, the high temperatures of last summer had an impact on the number of visitors in the physical stores. Not only that, but we also witnessed a decrease in online purchases that we can attribute to weather conditions. This is a clear example of something you couldn’t have forecast in any way.
As you see, there are certain and uncertain elements: the goal of working with Big Data is to try to understand trends in order to be quick and ready to adjust your strategy.
That’s why we are working a lot with simulations. Based on historical data, we can identify trends and prepare a plan of action in response to specific inputs we receive from the market.
Careful, though! Trends can be misleading. You need to be able to distinguish between macro and micro trends. For example, a macro trend is that the notebook market is decreasing year by year. But if you take a closer look at this market, you will see that the gaming notebook market is rapidly expanding each year, and this is a micro trend that’s going in the opposite direction of the overall market trend. Nonetheless, it can be a huge source of profit for you, if you’ve been able to recognise and understand it. At the end of the day, we can try to simulate how all the different factors will affect our customer base by combining all of these trends. The result of the simulation will help us to create different plans of action based on the clues we recognise in the market!
Unfortunately, it’s not this easy. As Mike Tyson once said: “Everyone has a plan until they get punched in the face”.
Creating different strategies according to different trends won’t make you successful. Being the first one to recognise and react to trends, this is what will build your competitive advantage. Once you see the first clues of a trend you simulated appearing in the market, you can quickly respond and be more agile than other companies who don’t do these simulations.
Again: data, data, data.
This poses a challenge, though. Every year there are more factors to take into consideration and more data that can be analysed. So the question is, where do you stop?
Well, it’s simply impossible to take everything into consideration. When a leaf falls from a tree, it’s impossible to predict where it will land. There’s no mathematical model to forecast this outcome because there are too many factors influencing the event. Luckily, there are also general factors you always need to consider such as market development, competition, innovation cycles in different companies, market price development, etc.
I would say that at the end of the day, when it comes to planning, every company is pretty much at the same level. No one, ever, can forecast with precision. You’re never going to find yourself saying: “Well, we planned this to happen in November and it’s happening precisely when we were expecting this to happen”.
Hence, being able to understand what’s happening in the market in real time becomes a crucial factor for success.
That’s how you build an organisation which is fast and agile to respond to the inputs you receive from the market with the right output. It’s good to have a plan, but it’s always more important to understand what’s currently happening. To do so, you can’t rely only on the data you gather. Benchmarking becomes as important as understanding your company and customers. If I’m growing at 10% on the prior year but the market is growing at 30% on the prior year, I’m not doing a good job.It’s like in sports, no? You always look how to improve yourself, but at the end of the day, you have to do better than your competitors to win the game.
As you mentioned before, big data started being on everyone’s mouth around 10 years ago. In which phase are we at the moment, in terms of owning, understanding and making the right use of big data?
If I would be Google or Facebook, I would tell you that we’re getting better and better.
Reality is, the rest of the world is still far away from unlocking the full potential of big data.
To understand the potential of big data, we can’t overlook Artificial Intelligence, which is the main technology to improve and deploy the usage of big data.
Consider the fastest computer in the world: the human brain. We’re still years away from that capacity, even with quantum computing. It’s only in the next 10-20 years that hardware development will really take a leap forward and drive the effectiveness of big data further. So we can definitely say we’re still far from the maturity of big data, and in particular for the retail segment, I think we’re still in our childhood.
Is it already relevant? Yes, absolutely. Understanding your past and present will definitely help to predict and understand your future better and better.
And for the last question: what’s the number one piece of advice you would give to the retail sector as a whole, regarding big data?
Don’t underestimate the power of AI.
AI will be a really big game-changer, it’s one of my deep beliefs in the last few years.
If you don’t invest in AI now, there’s a high chance you will disappear from the market in a few years.
Secondly, if possible, try to make alliances. We’re in a sector which is demanding higher and higher investments in manpower and expertise. We have an extremely high demand for limited resources, so it’s good to find allies to share projects, knowledge and learnings. Not only between retailers and solution providers, but also between retailer and retailer.
The giants like Amazon, Facebook and Google are investing billions and billions in AI and data processing. Don’t think they’re doing it to make the world a better place.
They’re in it only for the money, and if they have more money, there’s less money for the rest of us. The only chance to survive is to start looking for collaborations in key sectors.
Don’t be afraid of that, because collaborations and partnerships will enrich you. The most similar companies are not necessarily your biggest adversary.
Thank you, László!
If Big Data, AI and Digitalisation are topics you are interested in, check out the programme of Savant eCommerce Berlin on the 5th and 6th of February. Our carefully curated audience means a more dynamic experience for attendees: get to know all the right people on a first-name basis; establish contact with top decision makers and learn from experts in the field.