In our second Machine Learning episode, Dr. Glenn explains what role big data will play in the future of machine learning and personalised technology. Find out what effect this will have on advertising and what marketers can do to prepare themselves.
In this episode:
- The technologies that will enable the future of supermarket shopping and highly specialised advertising
- The role of big data in personalised advertising
- The effects of machine learning on the customer service industry
- How and why small and medium sized business need to get on board with big data and personalised advertising ASAP!
So far on our Big Data journey…
- Video: An Introduction to Big Data
- Blog: Understanding Big Data as a Complete Journey by Dr. Glenn Bayne
- Video: Machine Learning- The Future of Supermarket Shopping with Cat Matson and Dr. Glenn Bayne
Cat Matson (Chief Digital Officer, Brisbane Marketing): Welcome back to Traffika TV and welcome back to you, Dr. Glenn, this time to talk about personlized advertising and more about machine learning.
Dr. Glenn: Thanks Cat, good to be back again.
Cat: I’m very excited about this particular conversation. Last episode we talked about the future supermarket, the supermarket of five years time where our shopping trolleys are going to look different, they’re going to scan our products as we put them in, and we’re going to be able to be advertised to as we’re walking down the aisles. What I want to focus on today is the technology behind that and how personalised advertising is going to work. So before we get into the geeky stuff, can you tell us what personalised advertising is?
Dr. Glenn: Well Cat, personalised advertising is really taking your consumer data (this could be your purchases that you make whether online or in store, and perhaps through a FlyBys or rewards card), that data gets stored and then it’s remarketed to you based on what your preferences are and what your behaviours are, and essentially on what products you like best.
Cat: Nice. I already know how that works if I’m doing my shopping online, but if I’m actually pushing my trolley down the aisle, how am I going to see those ads as I’m walking through?
Dr. Glenn: Well Cat there are probably two ways that you could do it. If you look forward a couple of years, we could use flexible displays in shops or supermarkets. Currently they are available in the labs; they are only semi-transparent though, but they are very bendable so you can bend them around anything you like. Just imagine in the shops, you can have just a strip along a shelf or wrapped around the end of the aisle to display some specials. They can be displayed virtually anywhere.
The other type is holograms. You could have a display of a person or a product advertising to you as you walk past or essentially you could walk right through them, couldn’t you?
Cat: Obviously the role of data is going to become critical in order to advertise appropriately. Tell us more about how that data will need to be used in order to give us that quality, personalised advertising experience.
Dr. Glenn: Let’s take a step back and consider this term that we often hear- what is actually “Big Data”?
What it is, is a collection of everything that we do. At the moment we can collect data very well and very fast. We can store it in huge volumes. But we can’t quite analyse all of that data just yet. So that’s where we are in the evolution of the big data timeline, if you like. And I think that’s where the key is, to analyse that data. Because there is so much data that is generated each day by each person, we need machine learning algorithms to actually go through the data and pick out the golden nuggets of information that we could use perhaps to remarket to our customers.
Cat: It’s boggling and breathtaking at the same time. Now for all of this to work, when I walk into a store, the store is going to need to know it’s me and not some other nondescript middle-aged woman. How’s it going to do that? How’s it going to know who is walking into the store?
Dr. Glenn: There’s one technology we could actually use for that, and it’s facial recognition. You’ve probably been to the airport and have been through customs, where you have to stare at that light glows at your face and it looks back at you. That’s just picking out the biometric features of your face, so why not do that in a supermarket? The technology is already there. Currently in Japan they are able to do 64,000 facial recognition images a second. So that kind of technology, again research based but then applied in the commercial sector into a supermarket, can then just pick out anyone in a supermarket and match them to your customer database.
Cat: That is just stunning. And I guess to scale it back a bit, there’s no reason why when we pick up that shopping trolley, we don’t use something like our thumbprint to log in and pull up our shopping list, right?
Dr. Glenn: That’s correct.
Cat: So we have many variations of how sophisticated technology is going to get in the coming years.
Dr. Glenn: Really what it comes down to is cost effectiveness. It’s just taking that commercial and scaling it down to the smaller ones.
Cat: So if this is the future of machine learning, and the machines are getting smaller… On one hand it seems to me that we’re going to have a smaller need for humans. What’s the impact of all this going to be on customer service, particularly in the bricks and mortar stores?
Dr Glenn: I think bricks and mortar stores actually become more important, Cat. At the end of the day it’s the interaction with the person that you like and that you go back for. If you go into a store and you don’t like the store or the person, you tend to not go back.
When we think about online, at the end of the day the shopping process is kind of similar. You have items in the cart, you put in credit card details and you make your purchase. But when you go to bricks and mortar stores, it’s quite different; you’re interacting with a person and they’re building that customer relationship with you and that’s what drives people back to bricks and mortar stores. I think that will become even more important as we go forward.
Cat: We’ve covered a lot of boggling information in this episode. What do we need to be doing right here right now with our data so we can at least start preparing for this fantastic future?
Dr. Glenn: Cat, I think we need to start with just making sure we collect the data in the first place, otherwise we can’t use it. That’s the basics. For the smaller and even medium businesses, make sure you’re collecting your data. If you’ve got a website, you want to be collecting that data most likely through Google Analytics, and make sure things are tagged correctly. Once you’ve got that data, you can then go through and analyse it to find out what your customer journey is, what they’re purchasing, what they prefer, and then you can add some smarts to it and say, “Ok, I’ll put a machine learning algorithm over this to find out how to best target and market to these people.”
At the end of the day, the big players are already doing this. It’s filtering down from the big players, to the medium and it’s starting to get down to the smaller businesses. So you’ve got to start getting prepared because if it’s starting at the top levels, you’re going to be competing against them very soon. It pays to get onto it sooner than later.
Cat: And the beautiful thing about this technology is it’s not just for the bigger players, as you say it’s Google Analytics and we all have access to that, so it’s just getting smart about how we’re using that data.
Dr Glenn: That’s correct, like anything in technology, as you go forward it gets cheaper to implement and over time it gets smarter, but you have to get onboard. You have to make that effort to start collecting and analysing that data. That’s the key to it all.
Cat: Dr. Glenn, thank you once again for coming onto Traffika TV and talking about the very juicy stuff when it comes to personalised advertising and machine learning.
And thank you viewers for watching. If you have any questions for Dr Glenn please leave a comment and don’t forget to subscribe to Traffika TV for more of our Machine Learning series.