"Everyone knows they need more data scientists," Cisco's VP and GM of data analytics, Mike Flannagan, tells Essential Retail. "But industries are having an increasingly difficult time finding them."
Flannagan is referring to the global shortage of data scientists, which is unsurprising when you think that between 2015 and 2020, global M2M (machine to machine) data connections will grow from 604 million to 3.1 billion. But even today, there is a desperate requirement for people to analyse those millions of data points.
"Everyone of those connections in itself creates no values, it's about when you analyse that data, that is where the value comes from," explains Flannagan.
Data in retail
For physical retailers in particular, newly connected assets include video analytics, Wi-Fi, customer counters and POS data, add in the millions of online touchpoints as well and retailers can easily get lost in an ocean of data.
"Just because I can show you heat-maps, footfall and pathways through a retail outlet, that doesn't tell you anything, you have to put that together with things on shelves and where the displays are set up – what can I put on this endcap that will be different? – All that requires business knowledge and every vertical is having the same exact challenge: I need more people to understand data and my industry."
Flannagan says data scientists are going to be the most in-demand jobs, but also present the greatest shortfall. And businesses especially don't want their prize data scientists doing mundane jobs like data cleaning – around 80% of a data analyst's time is spent preparing data for analysis "You want them focused on actual analysis of data," says Flannagan. "Because that's where the insights come from that help you drive your business forward."
He describes how video analytics can provide retailers with customer information, including sex, age range as well as stock levels. "That's a lot of data – and turning it into information requires having someone who understands what demographic you are targeting and none of this has anything to do with statistics or mathematics, it really has to do with knowledge of the business."
He adds: "For example, footfall data tells you where people have gone, but the question isn't where they have gone, but it's how are you going to get them to go to the most places where they will pick up the most things to put in their basket?"
It is also increasingly important to analyse data in real-time. "What retailers are really aspiring to is not what happened yesterday, and three days back, but what is happening right now, so I can influence somebody's buying decision today."
He says a store manager looking at last month's data is no longer of interest, what is much more powerful is giving that person an iPad to see in real-time what is happening in store. To enable this, analysis needs to move away from being conducted at HQ, but analysed at an individual store level using a cloud-based solution. This saves time moving the data back and forth between store and analysis.
"Tiny bits of data like transmitting addresses from a mobile is fine, but video analytics is a huge amount of data because to get them really done well it has to be HD," explains Flannagan. "And a normal retail store is not going to have the kind of infrastructure to move that much data out of the store and get insight back in seconds – it will take hours to move the data."
Flannagan also says it is important for retailers to prioritise collecting data in the best way possible.
"More data is not always better, but better data is always more," he explains. "And you need to make sure the data you're using is something you trust. For example, if you're relying on an associate to take information down by hand, and another to enter it into a system, an opportunity for some error is quite high."
He points out having reams of data doesn't necessarily mean a retailer will be any smarter at running their business. "What is the data which is really relevant for you to solve the highest value business problem that you have?"
Additionally, data science should be adopted across the entire business, not just a single project.
"I think most companies today know it's important to start doing analytics, but I'm not sure they've really looked at the potential to improve every part of the business," he explains. "If you think about becoming a digital company, it's not about having analytics as a project, it's about embedding analytics in every process and using data pervasively. Make it part of the fabric of your company. It should be part of everything you do and inform all your decisions, from employee and customer experience, to driving top line."
Outsource data science
But he argues retailers don't have to hire the best data analysts. "Is your primary business really about data science, or is your primary business about being the very best experts in retail?"
Flannagan believes retailers have to make a decision about whether they want to have data science as a core competency or find a partner who does that for them.
"You focus on what you're best at, and let them focus on what they're best at and build that partnership, in the same way most retail stores outsource cleaning services - it's just not core to their business."
He continues: "In Cisco's case, we take the data science and put it into software, more so than people," he explains. "If a retailer is large enough, there's absolutely an opportunity to build a centre of excellence – even hiring data scientists and analysts and developing a service they can provide to other retailers."
Flannagan says it's also important for retailers to look after their data science employees if they choose to hire them. "There are things that you know and learn from speaking with colleagues, share information and techniques and tools, and that's how you become better at what you do, if you're the only data analyst working for a retailer, you're not going to grow yourself. The challenge becomes for those retailers, they have to think very carefully, are they going to be able to retain those good people that they hire?"
For more information, click below: