 Hello, everybody, and thanks for joining us at Think Tech Hawaii. This is Security Matters, and I am your host, Andrew Lanning. Today we're going to be talking about retail security, and I really want to dig in a little bit with our guest about kind of what's going on behind that browser. I'm sure a lot of you are used to using the browser, but you may not really know what's happening while you are. Our guest today is Chris Schwartz. He is a business intelligence analyst. Thanks for joining us. I know you've been at this for a long time, and I appreciate you taking some time off to come in here and talk with us about this. I think I'm like a lot of consumers out there where I'm using online services and shopping on Amazon and shopping on, I need things. I hate to go to the store. I love the convenience of online. Even when I'm going to go to the store and pick it up, I'll buy stuff and then have it waiting for me when I get there. But a lot of us really don't know how that works and what's happening. So today I wanted to talk about that experience first. So once you give us a little bit of your background, you know, kind of where you grew up, went to school, how you got to Hawaii, and then we'll get kind of get into the browser after that. All right. So I grew up in South Florida, Fort Lardale, went to school at Georgia Tech in Atlanta. Oh, nice. And spent about, I'd say, 20 years in Atlanta. My degree was in electrical engineering. I don't actually use my degree at all. Okay. So you're one that got freed up from into the EE. Yeah, yeah. It's funny how a lot of EE's do other things. But what the school did is I met my future wife at Tech. Awesome. So, but she ended up finding a job out here in Hawaii. Okay. And we had always thought about moving somewhere else other than Atlanta. Atlanta is a nice city. Lots of opportunity there, but I missed the warm weather. And so I know your wife. Did she grow up in Atlanta as well? No, she grew up all over. Okay. And then she ended up in Georgia Tech. Okay. Awesome. So that brought you to Hawaii. And how long have you been here? I've been here about almost seven years now. Awesome. So your work has been in the database sort of behind the scenes of a retail organization. So I actually don't deal directly with security. So I'm afraid I can't go into a lot of detail about what companies are doing relative to security. I can talk in generalalities, but not in the specifics. I deal mostly just with the actual data generated, everything that a company does. I see. So, you know, when you say data that's generated, you're referring to sort of what the user has done while he's using that browser. And that data, let's talk about what that is, first of all, and what it means for the development. So I have an interesting statistic for you I just recently read. This is according to something I read online on IBM site 2017 that I think, and I don't know how they come up with this number, it's got to be a pretty big estimate, but 2.5 quintillion bytes of data are generated daily from online activity. Oh, quintillion, which is 30 zeros. 30 zeros? Yeah. That's every day. And they just send you that and you figure it out. Then there was some other statistic, and this was back in I think 2012 when this one came up, was like 90% of all the data that was stored had been generated in the previous two years. Oh my God. And that was five or six years ago, right? And so it's this exponential growth in data. Everything we do, when you're on your phone clicking around, that's recorded if you make a phone call, there's records of data being recorded by your provider, right? When you're online, every single click in most cases is recorded. Wow. That doesn't mean they have your personally identifiable information, your name and your landing in this case, right? Okay. It's attached to it, but they do know that this device here visited, say, Walmart or Amazon, for example. And is this the MAC address of this device is tied to that transaction or the operating system? I'm sure. Yeah, they don't have the MAC address. Cookies are used. Everyone knows what cookies are. Oh, they infamous cookies. Just text files, right? Okay. And their purpose of them is to hold stateful information specific to the site that you're on, right? Okay. And so they could, a lot of times it could be things that you use to personalize the site, whether you're filtering on various things for authentication, right? So if you log in, you'll get an encrypted key in that cookie that the site then knows when communicating with your machine, at least for that period of time, usually there's an expiration time on it, on that transaction or that session. We'll then allow you to get to your profile information instead of continually asking for your password. Like, who are you, every time you want to do something? So that's a session cookie we're talking about, right? Right. And these cookies could also be used for personalization, for tracking, right? And so one of the big things they do is you're assigned a visitor ID, right? It's just a visitor to the site. It's usually just a nondescript of a number of an American that's unique to your machine, right? And so repeat visits, you'll have that same ID. Oh, so it recognizes me when I come back. Yeah, I notice how sometimes, you know, if you're looking at a thing, maybe you don't even buy it, but then it'll somehow shows up when you're doing something else. And so is there people mining that sort of activity, I guess I want to call it, or that's cookie information for things that I did while I was shopping? There's probably hundreds, if not thousands of companies out there that provide very specific services that can be used by these online retailers, right? Okay. Now, whereas Amazon, I think, does a lot of their own stuff in-house, a lot of other companies like, say, a Walmart or a Home Depot or Hawaiian Airlines, they'll actually work with third parties to provide very specific services, right? So one of those might be there's these ad exchanges. Ad, like a double click or something. I've not seen that before. And so what they'll do is these companies will actually, the way it works is they'll actually have like, as typically, it's called a web beacon, and you'll put this on your site. A lot of times it could be like a one by one pixel that's invisible, and so when that page is accessed, right, that then calls off to the third party, radio or double click, which is known by Google, and transmits that maybe you were on the swim outlet looking at goggles. Okay. And so now double click knows that at least someone using this machine was looking at goggles on swim outlet. Okay. And so then maybe you go to another site, and you'll see a little image down there of your goggles. Okay. Because I didn't, does it know if I bought them or not? I was kind of wondering about that, you know? So it might keep selling me goggles until I make a purchase. That's right. Yeah. And so that, but that's not done necessarily by the retail themselves. You said perhaps Amazon's big enough to do some of that in-house. A lot of stuff they do. And is this, is this driven by like machine learning? Is it approaching an AI level? I think most companies out there are now starting to add artificial intelligence and machine learning to a lot of these algorithms. Yeah. Because it's interesting, you know, when you talk about that much data, right, obviously there's no group of people that can consume that much behavioral data, especially from a large retailer who maybe has, do you have any idea what, how many people shop online like daily or, you know, like- Yeah. I don't know that stat. Yeah. Okay. I knew one stat. Amazon might know. I'm out of stats now. Okay. But I think you could probably Google that. And I think you'll find that people, there's statistics on that. Like in North America, there's like 300, 400 million people. And it's astounding because it's growing every year, right? Yeah. So we certainly haven't peaked out by any means, right? Sure. And then on the business intelligence side of that, what you're really looking at are all those users and not just like what they bought, but all the things they did on the site while they were there. Now is that, my understanding is that these, the intelligence you're doing is really designed to help improve my experience when I'm using that product. It could be for a whole host of things. It can be for just, for example, it could just be to monitor site operations, right? Oh. So I'm sure we've all seen errors occur online, right, when we're clicking around and that's quite annoying. Oh, like the menu doesn't work or you try to go to an item. Okay, I got you. Right. Or even worse, you're in the checkout process and you get right up to the end and you go to submit payment and then you get, it spins for a while, right? And then what do you need to say, oh gosh, did it take my credit card or not? You don't know what to do. Okay. So logging all of this data is not just to track and to provide personalization, but it's also to improve the site experience, right, and to identify, maybe there's an issue with a very specific new release of a Chrome version of the browser, right? Oh. And so this can help with that. Oh, and the handling of it. But you're not necessarily a code guy yourself or have you done code work or do you typically I would say less the, well, the language that I typically use is called SQL. Okay. Sure is cool. So this is to basically query all of this data and make sense of it. Interesting. And what is the upstream reporting? So if I'm a business intelligence guy working in the retail space, are they, what's the top thing they need to know? Can you give us an idea of what sort of information they're trying to gather from the type of work that you do? I think the biggest thing is just taking a look at not just how the site performed yesterday or last month, but also how's it performing right now, right? So more, it used to be if you went back five, let's say five years ago, 10 years ago, a lot of reporting is, and a lot of companies still are this way, but a lot of the reporting was what happened yesterday, right? And now people are getting more into the what's happened five minutes ago. Wow. So that you can act much more quickly. Wow. Right? Go ahead. And so, but there's several key metrics that people look at on site. So one would be how many visits did we have in some period of time? And quite often for comparison purposes and to handle seasonality, you'll look at what, let me compare it last year this time. Oh. Right. Okay. So, on a 100% basis, and most companies have targets for expected growth in terms of visitors per day or visits per day, orders per day, that sort of thing, and you can track really in real time how you're doing on that. Wow. Right? And so that's at the highest level what you're looking at. Conversion rates a big deal, right? Is that when you're purchasing? So when someone comes there, they actually make a transaction. Which is just simply orders divided by visits, right? Orders divided by visits. Sure. That makes sense. And so that's a key metric. Do you know how people got there? Like do you track it? Because don't people sort of get paid to drive you to your site or stuff like that? So generally, every click that you make on a site, you can think of as being recorded. And in this, you'll have, there might be hundreds of different fields of information that may or may not be populated. Okay. Like your visitor ID, you'll have a sequence number for which visit is this, is this your first visitor, is it your 50th visit? There'll be a sequence for, is this your first page view within your, this today's visitor or is it your 15th? Oh. So you can do sequencing on that and look at how much time you spent on each page. But then it'll have IP address and geography. But one of the other things it'll have is it'll have your referrer. So where did you come from? Oh, okay. Did you come from Google? Did you come from someone's blog? Did you come from an affiliate site where you actually pay the affiliates to bring traffic to your site? I see. Right. And then a lot of times there's tracking codes in there so that you can give credit to various affiliates and then there's actually, they'll get money for that. Yeah. So there's people that are, I know like, like if you can create a good YouTube channel like Think Tech Hawaii then if we could, if we marketed stuff here, if we could, if we could get like, like revenue, right? If we have referred, got people to go to a site and buy stuff. That's really awesome. Right. So we're coming up on the middle of the break. This is fascinating stuff. I know not many of you work behind your browser. So thanks for sharing this. We're going to come back with a little bit more of, I think business intelligence. We're going to talk about this stuff behind the browser and so join us back in a few minutes. We're going to pay some bills. Thank you. Aloha. I'm Kili Iakina and I'm here every other week on Mondays at 2 o'clock p.m. on Think Tech Hawaii's Hawaii Together. In Hawaii Together we talk with some of the most fascinating people in the islands about working together, working together for a better economy, government and society. So I invite you into our conversation every other Monday at 2 p.m. on Think Tech Hawaii Broadcast Network. Join us for Hawaii Together. I'm Kili Iakina. Aloha. Good afternoon. My name is Howard Wigg. I am the proud host of Code Green, a program on Think Tech Hawaii. We show at 3 o'clock in the afternoon every other Monday. My guests are specialists from here and the mainland on energy efficiency, which means you do more for less electricity and you're generally safer and more comfortable while you're keeping dollars in your pocket. Hey, Aloha, and welcome back to Security Matters. I am here today. We're talking about the experience behind the browser. I'm here with Chris Schwartz. He is a business intelligence consultant. He's been at this quite a while, so we're trying to get some insight into what's behind that online shopping experience. So Chris, we were just now kicking around the idea that a site like Think Tech could, I don't think we do this here, but if we had advertising on here that drove people to click into a Amazon site or another retail site, that we could perhaps get revenue for that. Essentially. And is that a popular way that the internet works to drive people to things? Do you have to pay to subscribe to people to do that, or do you just pay them per transaction, or do you know much about how that works? Yeah, I think I don't work in that space under closely, but there's affiliate sites. I think one of the big commission junctions, a big one, where they will basically, they have all these different sites that I think that they manage and they'll actually drive traffic to all of their customer sites, basically, right? And so does that translate as well from like, I'm thinking like, you know how you're on the plane, for example, reading a magazine, and there's like a link to something you want to know about. And can that start that way as well? I could that magazine actually get some revenue for having, you know, do they know that that link sort of came out of that magazine ad and things like that? Yeah, I think it depends, but often that would be the case, right, especially if it's in print. Yeah, it's tracked well enough. So usually there's what's called a tracking code is in there. Okay. So if you do click that link, the tracking code gets recorded in your visit as well. And then you can tally up how many visits did we receive from this particular source? Yeah, so because that's how, you know, how else would you know that you're advertising, you know, bringing your value, right, or whatever? Sometimes you sometimes you're only paid if there's a purchase. And other times it might be you're paid just via clicks. But I think most prefer that a purchase occurs. I see. So, you know, because if they believe in their advertising, if they're doing a good job with their promotion of their material, then probably they get a higher return from you actually making a purchase. Do you know what much about is it per purchase or is it percentage? Is it all over the board? Yeah, I'm not as close to those, but typically you'll hear of like revenue sharing models where it might be on a percent basis or maybe there's probably a cap on that as well, of course. Yeah, I'm wondering like if you're an Adidas, right, or something and you sell your shoes at lots of places. So, if I'm on Adidas' website shopping for some shoes, I don't know if they sell direct or not, but then when I go to, I want to buy some shoes, how do they determine where do I go? Footlocker or, you know, I wonder how that works. I mean, so there must be some sharing of that. Like you would want to necessarily give me to your competitor to make a purchase if I sold similar things. What side are you on right now, though? I'm on LinkedIn, but so let's just say for example, let's say for example I was buying some tools and so I was on Lowe's and then I wanted to check that item and I maybe took the item from Lowe's and said and stuck it back in the browser and clicked it again and it came out and moved me to over to Home Depot, for example, would you think they share revenue as well for that hammer? That would potentially say you went to Google, right? There's a number of different types of search, but at a high level that's either there's SEM, which is which is paid search, right? Or there's SEO, which is more organic search, which is which is free. The SEO, which is free, is the preferred model. Most most companies will spend a lot of time improving their SEO and that's that search engine optimization. That's right. OK. Whereas because that's free, right? And that they do that through tagging on the site. So all the search engines have web crawlers that read these tags periodically. And then there could also be rankings from other sources as well. That'll bump you up. And then the paid search would be so if you had clicked your Adidas shoe in Google via the paid ad, which is typically above the organic search. Oh, you mean when the results come up? How I don't have you. I don't have advertising at times. Yeah, exactly. Yeah. Is that because people are lazy? Yeah, I don't know. But I think most I think Google knows, right? You generally click the first one you see. I see. OK. Yeah, because I'm guilty of that myself. You can't blame them right for putting that one first. But so they pay to do that. Yeah. In most cases to be first, even. And that's because it's typically not like the Adidas site. You know, it's like there you have to go down a few. Yeah. Or you'll end up on some other retail. It's like, really? I want to go look at like the features and benefits of the thing. Right. Interesting. OK. So how much? Well, let me ask you a question. Do you have an estimate of like how much organic people? You know, get what's the word I want to use? Not manipulated. But how much people arrive organically to a retailer versus paid? Or do you think it's I think it varies by site? Right. And so I think most retailers are going to have segmentation of all of their sources. So at a high level, there might be maybe they have eight to ten different called marketing channels. The most basic would be direct, which would just mean you directly typed in where you want to go into the URL. Gotcha. OK. So you don't like you don't have you don't have a refer. He didn't come from a search engine. OK. He didn't do a click through from an email. So I want to do this dot com. Right. OK. Right. Then then then you have search, right? And in search, you could break into the paid versus free by just your choices. Like so by right after I do my search, I get presented some options. Right. And then and so I guess Google, that's their their version of what they sell, right? They sell that search, right? Optimization to retailers. Right. Right. But there's right. But you can you can come from Google and it could be SEO or it could be SEM right, either one. Sure. And they might still get paid anyway. Right. You might come from an email, right? Let's say you subscribe to an email program for for a company. You do a click through from there. That's likely considered a channel for most companies. OK. Right. Because you want to be able to measure performance of your your CRM, you know, customer relationship management with email marketing. Coming like from your your newsletter, for example, or something like that. Right. What else? You might have affiliates, right? So that might be another channel. Socials, another channel. You might look at coming from or Pinterest, OK, Twitter. I see. Our is everything is everything that I do. Like, is that valuable to someone online? Oh, for sure. For sure. But I think every company, you know, spends time on all those channels. But some companies might spend more time on SEM. Other companies might spend more time on on email or on social, for example. Right. So I think it really depends. Is there any seasonality to that? Like, you obviously they have the Black Fridays in the mall. They have a do they have an equivalent in the in the retail world for that behavior? Certainly for retail, Black, the Black Friday holiday period is one of the biggest for most most companies in the store. Is that also online? It's it's it's both. OK. It's both. But it's it's a really big deal online. Yeah. OK. And so is is there a. Is there a change in the behavior of websites at that time that they're trying to do? Or because they know people are more permanently purchasing just a lot more deals. A lot more. More also for deals of people and they'll they'll they'll they'll a lot more marketing spend. Oh, so the marketing side. Yeah. And that's and so that's related to how many times maybe your your retail name gets presented to someone like if I search blue jeans, maybe Levi's pays the most. So I see 50 different blue jeans outlets or is that is that a reasonable characterization? Yeah, for sure. Oh, wow. Amazing. So is is the user very much in charge of what he's doing when he's browsing? Can I ask that question? Like, I think you can. But maybe it takes you to pay attention. A little willpower. Yeah, you really need to take. Yeah. So so the but the so we talked a little bit about the cookies. I want to I want to make sure I'm clear. So they're they're just collecting information. There's they're not driving stuff, not in a session. Like so if I'm here and I'm looking at some shoes and moving around in a site, that information is all outbound. It's recorded, but it's outbound. They're not they're not at that at that point like flipping a page to me. Like there's no there's no activity. They can't drive their browser and think of it like cash. It's just recording a piece of information to hold it on your machine. So it can be referred to later by the by the server, the site you're on. Right. And so some of this business intelligence that you do takes a look at that behavior of those people and tries to decide is the site working as well as it should with this. Or could we do you do comparative? Like if we put the title bar on the left versus a title bar on the right, do you learn from that? Absolutely. OK. Yeah, there's all kinds of A.B. testing that that occurs testing. Yeah. Is that site A site B? You could think of it like that. Is that like live back to your example, right? Should we put the the bar on the left? You put on the right. OK, but in the middle sort of thing, right? And so you can you you can randomly break that up with visitors that come in that day have, you know, roughly a third. Let's say we did left center and right and you could do roughly in thirds and then see in each of those groups, which one was used more frequently. And is that what is it? What's the goal? Do you want to keep people there? Do you want to make them purchase? Like if do you care if they come in? They're only there two seconds, but they buy or I mean, is the boggest the buying is the thing, the buying is the transaction. So what's the what's the hallmark of a good website? I think you want to make it as friction free as possible. Friction free. I'm going to make it easy for them to find what they're looking for. And that's this the personalization is part of that. OK, it's it's to show you things that the company or the site thinks you're going to be interested in, right? OK, which comes from that cookies or those those other affiliates that have tracked all that about me, right? Or just data that the company has in house on your previous visits. Oh, right. How many do you have any ideas? How many people go to buy to the same place? Do they trust their trust a piece of it? Like I know always have a great experience with with Best Buy. So always go to Best Buy versus I don't know where else to go. But yeah, I think so. I think trust is a big part. And I think ease of use on the site. Like the checkouts a big deal, right? That's where they are. So a lot of a lot of websites are moving to the single page checkout. I think, you know, so Amazon had had a patent on the it was at the one single click checkout or something. OK, and I think I saw something about that actually ended in 2017. So now companies can more easily start to do these things where you have just one page for checkout instead of having to click through five different pages, right? Because each one of those clicks is an opportunity for you to to balance to leave or or an error to occur or is that maybe you change your mind. Yeah, I was there. I was wondering if it's like buyer's remorse to people get to if it's a slow process, they have time to think, you know, do I really need this? Right. Is my wife really going to ask me what I bought this time? That's what I think about. She has all of our Amazon Prime. I don't even get to buy anything. I just I have to send it to her. Can I have this? Yeah, same same as my wife. That's amazing. Well, any any last things you want to share with us about the retail experience with the, you know, any last advice for the audience? No, I think most of this personalization is there to try to help you, right? Yes, it's first and foremost to to improve sales on the site. But I think generally a company is trying to show you things they think you'd be interested in, right? I don't think they're trying to to show you things that you don't really care about, because that would be a waste of screen real estate, right? It's all about monetization and and making money, making money. So the retail experience is good for you out there. It's, you know, they're collecting information about what you do, but it isn't personalized to you. We've learned that today. We learned there's a whole lot of interest in how you act online. So I hope that was beneficial for you next week. We may be dark. I have a change of command at the NSA that I'm supposed to attend. So I may not be here, but please join us again on Think Tech Hawaii anytime you want some great community programming and join us here because security matters. Thank you.