This episode of “The Edge of Innovation” is about Big Data. The Edge of Innovation is produced in partnership with SaviorLabs.

Transcript

Sections

Introduction to Big Data
Metadata and Targeted Ads
User Advantages of Understanding Big Data
Business Applications

Paul:       This is the Edge of Innovation, Hacking the Future of Business. I’m your host, Paul Parisi.

Jacob:     And I’m Jacob Young.

Paul:       On the Edge of Innovation, we talk about the intersection between technology and business, what’s going on in technology and what’s possible for business.

Introduction to Big Data

Jacob:    Today we’re going to be talking about big data, a category that seems rather big and large and has a lot of parts to it.

Paul:       Hence the word big.

Jacob:    Yeah.

Paul:        It’s like a big wheel.

Jacob:    So, Paul, I’ve heard this term. Can you kind of fill it out for me, because it’s been related to Google and Facebook, which seems rather intimidating. Can you help us understand what exactly big data is and what it’s used for?

Paul:      Well, it’s big and it’s data. That brings us to the close of our show for today.

Jacob:    Thank you for listening, everybody.

Paul:       Okay. So, big data has really happened in the past 10 or 15 years, where we realized that there are lots of little pieces of data out there. When you click on a website, you open that website and you type in your name, that’s all producing data. If I gather those all up, it starts to get really unwieldy, almost. The amount of clicks that you do, the articles that you read, the time at which you did that, what you spent at that store. All of these are attributes of big data.

Metadata and Targeted Ads

Before we get into that, there’s something that we need to talk about called metadata. Meta means information about the data. So, metadata is the fact that…Let’s say I go to Google. So, I open the website, and I type in, I want to find the best hat within 20 miles, a hat store.

So, a lot of things just happened. I did it at a certain time. That’s a piece of metadata. I typed in some words, and I clicked enter. And I have an IP address. So, Google makes a note of that that says this IP address just searched for the word “hats” and it did it at 1:15 in the afternoon, for example. All of those things are metadata.

Google then displays to me the answer to the query “hats,” and I then click on that. And I click on the first one that says, “red hats.” In a perfect world, somebody might conjecture that I like red as opposed to the green hats that were second.

So that whole concept of tracking all of those little bits of data make it very, very much big data. It’s lots of little bits. So, what does that mean?

If you were over my shoulder, you could have asked me the questions, “Why are you clicking on red hats? Do you like red?”

“No. It was just the first link, so I clicked on it.”

Okay. Well, that’s sort of a false positive. But if I find out that you went to the store and bought a red shirt with your credit card, now I have another data point, and I can start to correlate this. Oh, and you bought a car, and it’s a red car. I can now start to begin to make a judgement that you might like the color red.

So, now I find out that you’re shopping for boats, and I’m Google. I’m going to put a picture of a red boat in the right hand side for the ads. And why wouldn’t you? Because you know that I like red. Why would you show me green ones when you know I don’t like green, maybe?

It’s the same thing where you meet somebody and learn that this person really likes Star Trek, but doesn’t like Battlestar Galactica. Boy, am I geeky here. Or you go to somebody who doesn’t like fish. The number of people I can tell you… We live in New England. “I don’t like fish.” You try to invite them over to have swordfish because it tastes like steak, and they’re like, “Wait a minute. Is this fish?”

Jacob:    I was thinking recently about this whole idea of, “I like fish but I don’t like the taste of fish.”

Paul:       Fish are people too?

Jacob:    It’s like, “I like to eat fish, but I don’t like the taste.”

Paul:       You like the idea of eating fish.

Jacob:    Yeah. Or I like fish that doesn’t taste like fish.

Paul:       Then you should try swordfish. We did that with a friend, a very good friend. We had swordfish, and he’s like, “Is this fish?”

Well, it is. We knew that, and we were sort of pulling a fast one on him, but Google, if you don’t like fish, shouldn’t show you fish recipes, because that’s not a good thing. So, big data takes all of those verbal and nonverbal cues that we as humans do, and you might have seen somebody reading Sports Illustrated. They’re probably interested in sports. And aggregates them and helps marketing people deliver information that’s valuable to you.

So, I have a good friend who loves boats. And he’s like, “I don’t care that they know it. In fact, I’m glad. I want to see things that are relevant to boating on my Facebook feed, because I’m interested in boats.” So, that would give him the sense that, “This is cool. This is new in boating.” Or a new law has passed. I need to know this. That’s a benefit.

So, we’re in this very infantile beginning of big data and utilizing it. And then, you have to think about the actors involved in big data. What does Google do with it? They sell ads. That’s all that Google does is sell ads. They say that we can show your ad to people that love red boats. And you’re sitting there saying, “We make red boats, and I really need to sell some red boats. So, Google, if you can show this to somebody, I’ll pay you a dime.” If you can show this to somebody and they click on it, I’ll pay you a dollar. And that’s how Google makes its money. That’s all they do.

As Tim Cook said, “If you’re not paying for the product, you are the product.” They are selling the fact that you are interested in red boats. And they say they do it anonymously. They don’t tell who it is, because they don’t. And frankly, Red Boat Company doesn’t care who is looking at their ads until they want to connect to them. And then it’s I’ve clicked. I’m on the Red Boat Company page. I can then give my information, and then they can contact me, or I can contact them.

Everything we’ve talked about up till now is dealing with good actors or sort of passively not bad actors. And the real crisis comes in when we have all this history of data. What is a bad actor potentially going to do with it?

We have historical precedence for the suppression of certain views. If there was a military regime to take over and, as absurd as this sounds, anybody who likes red boats, we’re going to put them in prison. Well, now, that data is out there about me that I like red boats. How do I control that? It’s insidious. How can I control it? I can’t, because it’s in so many different places, and it’s not set up there.

Of course, the government says, “We would never do that.” Same thing with health care. The minute they find out your sibling or parent dealt with this, that’s big data. Can we correlate to them having that issue? So, we’re going to change them more for their health insurance rate because there was a problem with a sibling.

Jacob:    I imagine specifically with that health insurance situation, there’s big data associated with things…for example, if somebody had a genetic disposition to have a condition, there are big data elements that go into realizing that person has that condition, even before they’re diagnosed with it that would factor into how that insurance now interacts with that person.

Paul:      Right. Requiring us potentially to now make laws that say you can’t use that big data in setting your rates. Well, then what good is the big data, except for the fact that it might help you avoid the outcome.

So, all sorts of questions start to come about with the big data. In a good actor society, it’s all beneficial. It’s all good things. Because if you don’t like sports, don’t show me articles on sports. Don’t send me, “Do you want to subscribe to Sports Illustrated?” because it might cost them $0.50 to send you that card in the mail to say, “Why don’t you subscribe?” Well, that was wasted money on their part, and it was annoyance for you because you had to throw it out.

Big data helps us conquer that, but it also does very much so, expose us to who-knows-what.

Jacob:    This seems rather large. And it seems like the more you begin to poke on this, the big data category seems a bit monumental. What are the important things for small business owners? For entrepreneurs to understand about big data? And how can they use that?

Paul:      It’s vastly different. You can manipulate some of the big data engines out there. And if you go and search for… Let’s say you want to buy a new stereo system. And you search for it on certain websites, you will magically see stereo systems ads to appear in the next two days.

Now, what’s interesting about that to me is that they’re smart enough to do that, but they’re stupid enough to show you an ad for a stereo system, but you already bought one. So, obviously there’s some holes in the system. And frankly, it’s annoying. I just made the commitment. I spent that.

Jacob:    They should be trying to sell you CDs at this point.

Paul:       CDs or a warranty or something like that. “Hey, you just bought that.” But you can see how that loop is enclosed. So, that’s some of the advantage to it. But to get back to your question, what are some of the advantages to small business?

The big businesses are doing that, retargeting, saying, “Hey. You just browsed a stereo. Let me show you ads for stereos.” And they might even… You might have gone to the stereo shop, and you’ll see ads from the stereo shop. And that’s Google tying that together.

User Advantages of Understanding Big Data

When it gets really interesting is, we can go to Bob’s Stereo Shop. We see an ad competing with that. That would be an advantage to me. And there are some certain ways that people have observed is if you search for something, even like airline flights, you can get a better deal by doing it in a certain way.

So, if you search for this and then don’t look for it for a day, and then go back, your rates will be a little bit different. And nobody knows what those are. They’re very opportunistic. But they are manipulating the pricing of that.

You could actually do this, which was… Let’s say you go… In a perfect world, you go to amazon.com. You see that the price of something is $100. Or you call up somebody in California that you pick out of the phone book. “Go to amazon.com, and search for product x, y, z.” And they say, “It’s $100.” You then go to jet.com and search for product  x, y, z, and it says it’s $100. Then you go to amazon.com, and it’s $95.

Jacob:    Oh, wow. Right.

Paul:      They will respond in that way using big data, because they want to beat jet, and they’re willing to pay for it.

Jacob:    So, there’s even the possibility while you, of course are being monitored by big data, to manipulate it for your own advantage.

Paul:      Absolutely. Those are very hard to discover, those things. And usually you discover it in the midst of things. The other thing can be that let’s say you went to amazon first, saw it at $100. You go to jet; you see it at $100. You go back to amazon, they have no advantage of dropping the price, because they’re already said, “It’s $100.” That would look weird to you. So, they don’t drop the price, hoping you’ll buy through amazon. Which, they have advantages because you know who they are, etc.

Now, bringing it down to a smaller business. We recently had a client who said they wanted to add some text in the front of their webpage. And I really was struck by why didn’t they look at the – I didn’t use the term big data, but the analytics that they had. Because a homepage is basically like a hurdle to get over. It really doesn’t do too much, but it’s really the test on whether somebody is committed and really interested in what you’re saying, or is not interested. It’s the ultimate, walking by a store. You think of walking by a physical store, and you see in the window, and you see something. Why didn’t you stop? Well, the reason you stopped is because you saw something interesting there. And you went to the next step.

So, they didn’t ask that question. They just said, we want to put something new on the front of our website. They should have at least said, “What’s our current metrics on our website? We’re going to change this, whether it be from red to green, or put a new text up there that says, ‘today only,’ or whatever and see if that changes.” That’s a wise thing to do.

But more so, it might have been better to say, “Where are people ending up?” They come to our website. They engage. They make that commitment to come in. Let’s put the new information that’s really important there. So, that’s one very simple type of big data.

Another one. We have a client who has a service firm that does services for pregnant women and asked, “Where do your people come from? Are they coming from mobile devices or from desktops?” They really didn’t know. With Google Analytics… And by the way, Google Analytics is a brilliant product. It’s free. So, what’s the rule here? Who’s the product? Who’s the… You’re giving away, when you use Google Analytics on your site. You’re giving away information about the people that are coming to your website.

So, if you have an ice cream store, and you put Google Analytics on it, and they click on it, and they go in and click on an article and read that, Google now knows – you do too – that the person with this IP address, likely in this area, in this town, with this browser, that uses a Mac, came to this site and read about ice cream. So, now Google can now show you ice cream ads.

And so, that’s big data. Google is making…every dollar they make is based on big data.

So, all I think is that small businesses should be using that same thing. If they have a lot of people that are coming to their site, and they look at the people that convert, and they find out that everybody that converts is using a Mac, what does that mean? Why is that?

We’d have to take a sort of use case and deconstruct it to really understand what that meant. We could do that, even in a future show.

Business Applications

Jacob:    So, what are two or three things that small business owners and entrepreneurs can do to start using big data on their websites?

Paul:      I think the bottom line is they really need to become intimately familiar with their analytics. Google analytics is great and easy to put in. You do need to make that emotional judgement; you need to make that philosophical judgement of whether you want to give away your data. There are other analytic packages, which don’t give data to Google, so you could use those as well.

But the point is you need to look at it, and you need to say, “What am I trying to get people to do?” The big trend right now is one-page websites. And they’re great. And we can do some fancy JavaScript stuff to see what parts of the pages were viewed or displayed on the screen. We don’t know whether they were read. But it really changes our analytics information, the depth of the information from an analytics point of view.

So, whereas on an older-type website, if they clicked on more information, or what colors are available, to keep our color idea running, I’d know they were interested in colors. Whereas in the scrollable website, I could see that they scrolled past colors. I might even be able to see that they scrolled, stopped, five seconds later kept going. But it’s much harder.

So, becoming intimate with your analytics, understanding what people are doing, and then starting to think about how do I – I’ll use the bad word – manipulate them into doing what I want them to do.

That might be as my other friend that likes boats saying, “Give me the information about boats up sooner.” And I can do that by understanding where they ultimately get to.

Now, having said that, it’s really complicated to understand these things. So, there’s people like me out there, we do that. We help you understand what it is that you’re seeing. We also ask you why in the world do you want to do that.

One of the holy grails that hasn’t happened yet. There are people that are working night and day to try to figure out how to figure out to get a list of what you bought on your credit card. So, right now the credit card industry is based on 40-yr old, 50-yr old technology. And it has to do with how much was it and did you have the money, and did you get the transfer of the funds working. That’s clearing house really. They just never had the comprehension to say, “Wait a minute. If I could get the fact that you bought Twizzlers and beer and bread, and you do that every Friday, that’s going to give me opportunity to market to you.” And that is the holy grail, is to figure out what you spent the money on.

There’s two ways to do that. One is to get the store to give that to them. There’s a whole bunch of privacy issues there. AT&T Universal card, when you register a purchase, they will tell you if it’s cheaper for 90 days. Why are they doing that? Because they want a list of everything you buy.

Now, they aren’t effectively using that, because I just bought a new television. I said, let me do that. It was actually more of an experiment. So, I registered the product, put in the model number and all that stuff, and I got a beautiful email the next day. “We’re hunting for a better price for you.” It was $760 is what I bought it for. I entered that, let’s say on Tuesday. Wednesday I get the thing, “We’re hunting for it.”

I want to Amazon to look. It was $740. So, that’s $20, and they never came back and said, “Hey…” They didn’t catch it. Now, they may say, “We can’t scan amazon,” or whatever excuse they’re going to say, but they’re fundamentally doing that.

So, if they were to have done it, they would have refunded me $20. And now, after the 90 days is up, they’re like, “Sorry we didn’t find a better price for you. Do it again.” I don’t think it’s a scam, but I do think the incentive for them is certainly client retention. That’s a neat feature. But it’s also to profile me, and say, “Well, he bought a television. So, maybe in three to four years when technology has changed enough, we’ll start telling him about TVs.”

That’s heavy lifting to do all that. Do I send him a flyer? Do I send him an email? What do I do with that? But that’s the holy grail.

One client we’re working with, they are consultants. They have long-term lead nurturing. A client doesn’t come in and decide, “I’m going to buy from you today.” They’ve heard about that company. They come to the website. They want to feel good about that company. They want it to be reinforced that this is a competent company in this market. So, all of those things are there.

Well, it would be nice to know that they’re looking, to be able to measure that, to say, “Jack, I saw you were looking at our website.” Or even if you meet them at a conference, reinforce the things. “We’ve got a great article up there about your market,” knowing full well that he’s read it. But start a conversation that way. So, that kind of thing.

 

Jacob:    You mentioned one thing we can do for big data. Is there anything else?

Paul:      Sure. As a small business, or as a not-a-Google, what you can do… One of the things that allows you to use big data to your advantage is Facebook. Because you can target people in a certain town that are interested in a certain thing. And that’s why Facebook is worth what it’s worth. Because I now have a description of the person that they’ve provided to say, “I am interested in boating. In fact, I’m interested in red boats.” This is a little bit of a silly example, because that’s not one of the segments in there. Boats is, certainly.

But if I make red boats, I can say, “Anybody in Boston that likes red boats, show them my ad.”

Jacob:     Well, you could do that even to say anybody that lives in Boston within this zip code, that works on a Mac, that lives within x-radius of the Apple Store and blah, blah, blah. You target an ad down exactly to that person.

Paul:     Right. Now, this is called business to consumer, B2C. Now, LinkedIn is trying to be that for the B2B. But they haven’t really gotten it there yet. So, if I want to basically find people who are executives and they’re interested in buying life insurance for their employees, that’s not really an attribute out there. I certainly could go and send things to C-level executives in my town and have that show up in LinkedIn. But again, there’s not as much of that drive. People go to Facebook to see what their friends are doing. A CEO doesn’t go to LinkedIn necessarily every day to keep up on things.

So, that is, if you’re business to consumer, you can definitely target your ads with Facebook. It’s a wonderful platform to do that. You can also do the same thing with Google AdWords, where basically they will show ads based on the words that people type into the query. What you’re going to see over time is I want to be able to show my ads to people that are 18-24 and own their own car or have a lease. That kind of stuff will be coming for AdWords.

Jacob:    The Edge of Innovation is brought to you in partnership with SaviorLabs. SaviorLabs exists to help businesses mature and strategize for the future. Learn more about SaviorLabs at saviorlabs.com.

Thank you for listening to this episode of the Edge of Innovation: Hacking the Future of Business. For the show notes and more information about Paul, please visit paulparisi.com.

The Edge of Innovation is produced by Jacob Young, in conjunction with copious amount of coffee. Music on today’s episode is from bensound.com. Paul can be found on twitter at PDParisi and on LinkedIn at linkedin.com/pdparisi. This episode, like all our episodes, is transcribed and available at paulparisi.com. Thanks for listening, and we’ll see you next week.