We love analytics and big data but what we love more is our ability to stick to our favorite terms. There are many cliche words we regularly use and sometimes we don’t realize what these words mean and how practical they are. In this podcast, we want to have fun with few of the most popular big data cliche words that are used frequently in the business world. Join us and lets have some fun.
A Quick Preview of the Lifestyle Analytics podcast (show notes):
- Jeremy’s talks about Q1 results and cliche terms [00:44]
- Top over-used buzzwords on Linkedin [02:16]
- What is the number one overused big data term? [03:57]
- What is the favorite term used for too much analysis but no action? [07:00]
- If universe is filled with dark matter then what is big data filled with? [09:06]
- What does vegas and marketing department has in common [11:05]
- What is the Pareto principle and why it is the most overused term in the analytics world [14:30]
- How are decisions made inside the organization [17:46]
- A cliche word that motivates us to do optimize results [19:47]
- What are some of the most popular Big Data buzzwords coined by industry experts [21:43]
Listen to the Lifestyle Analytics: How Data & Analytics Revolution are Changing our Lives
- GREAT JOB!October 26, 2015 by MattMcWilliams from United States
WOW…Analytics Today Podcast is flat out awesome. Good production quality. Easy to listen. Very impressed Jeremy & Sameer. Keep bringing it.
Resources discussed in this podcast:
00:15 Jeremy Roberts: Thank you for joining Analytics Today, a podcast series that focuses on big data in analytics and latest trends in the digital world. I am your co-host Jeremy Roberts and with me is my co-host Sameer Khan. Hey, Sameer.
00:27 Sameer Khan: Hey, Jeremy, how's it going?
00:28 JR: Good. Getting excited about summer coming up.
00:32 SK: Oh, that's exciting, yeah. It's fascinating the last couple of days we had pounds and pounds of rain in Houston so everything kind of shut down.
00:40 JR: Yes, same here in San Antonio. Yeah.
00:42 SK: Oh, you had the same problem? Okay.
00:44 JR: Yeah. But I think what's even more exciting about summer coming up and all that is this is that moment where Q1 results within a company finally comes out, right? And either you're panicking or you're not. And all of a sudden you get brought into these big meetings. And the funniest thing about these meetings is they're talking about things that we gotta do. And it's the cliche terms. It's like they're speaking a completely different language. Today, today's topic is analytics and big data cliche terms. And I think this is gonna be a fun one.
01:23 SK: That'll be fun, yeah. I love cliche terms and we all use it and we just make fun of it. Yeah, this is gonna be interesting.
01:32 JR: I mean, there's even cliche lingo that you can do. Or I've been on calls or I think you and I have been in meetings before where we're waiting for our VPs to really use all these cliche terms and you take a little count of how many times they say the same phrase over and over. Remember when cloud came out. Cloud was the big thing, so everything was cloud this, cloud that.
01:53 SK: Yeah. No, I remember it and I think it's become a standard part of the industry. Once you decide to use a specific cliche term then you just get stick to it and people start kind of annoying you with that term, because they know that you're gonna be using that term pretty frequently.
02:16 JR: Here's an example. I was looking online and I saw this article that says, "The top over-used buzz words on LinkedIn profiles"... I know this doesn't have to do with analytics and data but this is on LinkedIn now. They were looking, this is back in, this was in 2010, a little bit old, but it said, "In the United States, Canada and Australia, the most over-used buzz word were people saying they had 'extensive experience'." It's such a broad term. Like, "I have extensive experience, great." Brazil, India and Spain, their big terms was they kept on saying that they were 'dynamic'. "I'm dynamic, right?" In the UK, this is an interesting one. It's not 'extensive experience', it's not 'dynamic'. But UK, the word there is 'motivated'. That makes me think, are you unmotivated? And you have to tell people that you are motivated, 'cause that's weird. And then the other one is France, Germany and the Netherlands, the most over-used buzz word on profiles is the word 'innovative'. Does that mean most people are not innovative?
03:31 SK: And also innovative in what ways? That's pretty cliche would you say because there could be a lot of thing you could be innovative in. That's fun, that's fun. Why don't we get started, what have we got?
03:45 JR: Okay. Here's the first one. This is a funny one.
03:50 SK: Going round one. Analytics and big data cliche terms. What's the first one? Tell me.
03:56 JR: 'Data is the new oil.'
03:57 SK: Nice. I have heard that term so many time. 'Data is the new oil.' What does that necessarily mean? Just to give context, people, everybody knows how oil is important in the modern revolution, right? Everything today, the industries and the automobiles, the jets and everything. Pretty much the economy runs on oil. We live in the economy, oil economy essentially. With data being populated everywhere with digital technology as we talked about it in our Lifestyle Analytics podcast number 10. That how technology is taking everything by storm and because of technology what's underlying factor that's changing the world is the data production. It is, that's the reason why, because data is being produced so fast that it can be used in many different ways. It can provide a lot of different insights. I think that's what the origin of this term, 'data is the new oil'. What are your thoughts?
05:00 JR: I see that. I see it as a currency, exactly as you're saying the word. Data needs to be the currency to be able to drive the success of the business. And you and I are what we'd call analytics and data junkies to where we feel like if you don't have the right data, you can't make the right decisions. But, I also see it as the oil within an engine, to where you can't effectively run business without data because all you area doing is just doing a whole bunch of 'best practices' that are from some random industry from how long ago. And you're just guessing, or guesstimating. And that's never one thing. Imagine you're an analytics professional, you're trying to look at historical data to forecast out your upcoming year and new product launch. And you're just gonna say, "Well we don't really have the data, so we're just kinda guess what happened based on how you felt."
05:57 SK: And also the revolution standpoint, how everything is gonna change. Previously, it was all about, "Produce more oil, produce more oil, fill the economy, and drive the economic engine." Now, it's... We have limited resources with the population of the world is growing, how do we make more of those limited resources we're gonna have? Like either we're gonna and find a new planet or we're gonna find ways to capitalize on what we already got. So I was doing some research, and I was seeing who was the first person that actually coined that term. So there was a guy called Michael Palmer, as per Forbes site, and so he coined the term, 'Data and just like crude' back in 2006. So I definitely wanna give him credit there. But that's funny, since then, it's been repeatedly used across big, small medias within the organization and outside of it.
06:51 JR: I can't wait to see when they move more into wind, energy and solar energy, what the phrase is gonna turn into. But we'll see.
06:58 SK: Yeah, awesome.
07:00 JR: And I think what the last point of this, I think it goes with this idea is, data is the new oil. The idea of the oil economy... It's good to have more oil, right? But in a lot of companies, it's not good to have too much data because as you see, most people only use less than 1% of the data that they have, or effectively use. So I think that goes great into our second term that we wanna talk about today, is 'analysis paralysis'.
07:27 SK: I can relate to that very well. That's a good one. Yeah, and 'analysis paralysis' is an interesting one because that kind of like you said, the more data you have, the better chances are of you performing more and more analysis. But I think the question is, how much more can you do? There has to be a point where you have to just stop and think about it, and look at your previous analysis, and assumptions, and variables that you have as a part of that whole analytics package, and draw conclusions on it, which we call as key takeaways. Have insights and key takeaways, and if there's something missing, articulate out what's missing, but don't just spend a lot of time pulling data.
08:13 JR: Pulling data.
08:14 SK: And driving conclusion, and adding segments, and segments after segments, and not necessarily focusing on the output, like the garbage in, garbage out. It's kind of very similar.
08:27 JR: I used to work as an analyst. And I was like, "Well, hey, you know, go ahead and run a report for me. And let's see why conversion's down, and traffic is spiking, and where the pain points are, and all the stuff." She comes back three hours later with a 30-slide PowerPoint of charts. I'm like, "This is too much data, man." I mean, top three, tell me your top three. That's all I need to know. Yeah, three things. That's it.
08:53 SK: Yeah, I know myself. I'm being very analytical. Sometimes we got into the tendency of doing a lot of analysis. But then we forget about what we're actually trying to do as a result of the analysis. So that's a very interesting one.
09:06 JR: Yes. Let's get to this as to our third term. This is a newer one. I haven't heard of this one as much, but 'dark data'.
09:15 SK: So 'dark data' essentially, it's a term that is being popularized because remember we always talked about this Gartner study. So when the Gartner did a study to identify what percentage of the data that is being actually used by the organization, and it's very little. I can remember we talked about like it's less than 5%, I assume. So then we take that data, and we flipped the coin, and say, "What is the percent of the data that is actually not being used?" Which, essentially, a part of that is dark data. So dark data is... And a lot of people think of dark data in different ways.
09:48 SK: So first is data that's been produced but not used, so it's either sitting in some type of data warehouse or it's sitting in somebody's computer. Like I remember in our good old days, we were looking for this particular financial data. And it happened to be sitting in someone's computer. So and we were like, "Okay, how do we get this data back in the data warehouse?" We can't because it was sitting in that person's computer. So that's what dark data is. Dark data essentially, it's just like the dark matter. I don't know how much you follow the space and the universe.
10:22 JR: Oh, I do.
10:23 SK: You're nerdy, like me. Yeah, I know.
10:25 JR: Yeah, come on.
10:26 SK: So yeah, so just like the dark matter, this is what the dark data is. It's like nobody knows where the data is or nobody knows where the data is hiding, nobody knows what the data includes. Nobody is touching the data and using it, so that's what we call dark data.
10:44 JR: Interesting. It's Star Wars-esque, kind of cool. So okay, next one. This is a good one. This sounds like somebody who is an analytics professional who wrote the book on how to cheat Vegas in blackjack with 'double down', the term.
11:05 SK: 'Double down'. I hear that so freaking awesome times that sometimes it just makes me think like, "That is definitely a cliché term. Double down, what are your thoughts on that?
11:22 JR: I don't really know. Sorry, about that one. That's just... [chuckle] That's just a term that's terrible.
11:28 SK: "We have to double down on our numbers this quarter." How many times have you heard that?
11:31 JR: I mean, come on. It's such a stupid term. I'm sorry, but it's just it doesn't work for me because when somebody comes to me and says, "Let's double down." My thought on that is, "Double downing on what?" So you're effectively telling me that what we have to do is double down. Okay. What does that mean? Are you double downing on revenue or are you double downing on conversion rate? What does it mean? It's such a broad, terrible term to where people feel like they can get away with just saying it. It means nothing.
12:05 SK: It is, yes. Especially... And that's very cliché for marketing, I would say, even though it kind of keys into the data. So double down to me is, how can we do more of this, given that we have more of Y? So how can we do more of X with what we got, and Y, and produce Z results? So that kind of thought process, right? It's very centric around, how can we double our production? That could be one area I can think of. How can we double our revenue? Yeah.
12:41 JR: I remember an old story that I told you before. And the quick version of that story was: I was told at my old company to double down on the highest performing channels that produced the highest conversion rate. And they said, "We're gonna take money away from the lowest performing channels, based on ROI, and we're gonna put it towards the highest. We're gonna double down on the highest ones." And what ended up happening... Well, they didn't...
13:06 SK: Just like that.
13:11 JR: Yeah. Okay. And what ended up happening is, they realized that that lower converting channel was essential to the buyers' journey, as far as building brand awareness in the discover process. And it basically killed our entire conversion, because of this terrible term, 'double down'. It must have been something to where I'm sitting in the room, and somebody says, "Double down," and I'm sitting there thinking, "You gotta be kidding me. They just said 'double down.'"
13:39 SK: Well, because at the end, it doesn't mean anything, right? Yeah, it doesn't.
13:42 JR: It doesn't mean anything!
13:43 SK: It's a pretty broad term, doubling down on things. And practically speaking, in all the different organizations that I have personally worked, double down does not really work. [chuckle] It does not.
13:56 JR: I'm gonna asterisk this one and say this is the most terrible term ever.
14:01 SK: Okay, there it is.
14:02 JR: For analytics, 'double down'. I'm calling this one.
14:03 SK: So for all the lists we have, this is being crowned as the terriblest term ever. [chuckle]
14:11 JR: Because all it does is get you in trouble. It forces you to do something that you don't wanna do because you feel like, yeah, you're in Vegas, and maybe you had a few too many free drinks, so you decided to double down. Well, that's okay, because you're expected to lose that money. Not in business. I don't just expect to lose a million dollars. [chuckle]
14:29 SK: Yeah, no. Absolutely.
14:30 JR: Not really cool. Alright, so let's go to the next one. 'The 80/20 rule.' Oh, wow. Okay, so this one, for me, is all about assumptions. With 80/20. It's basically saying, "20% of the things you're doing are driving 80% of the revenue... "
14:50 SK: The results.
14:51 JR: Yeah, the revenue, or results, and so on. This, for me, is a cop-out term. I've used this before in my conversations with management, just to really... I shouldn't have used it. I kind of slap my own hand here. But it's one of those things to where you kind of use it just to say, "It's an 80/20," and you don't really wanna have to go into an explanation. But I think a lot of people use it as a default because they don't wanna really explain the detail. They just say, "No, no, no, 80/20." All we're gonna do is look at the top 20% of performing channels, performing attributes, and say, "This is exactly what's driving 80% of the revenue." That's bad, as an analyst.
15:31 SK: It is. Actually, the origination, I was doing some research on this, the origination of the 80/20 rule, which basically came from the Pareto Principle. Yeah, so that's where the origination is. And basically what that Pareto Principle states is, for many events, roughly 80% of the effects come from 20% of the causes. So that's how that cause and effect thing, coming to picture... So not all causes produces an effect, right? There is a 20% of the causes that will produce 80% of the effect. So that's what the thought process is there. But I completely agree with you. The first time that this was actually incorporated into management and marketing and business was close to by the University of Lausanne in 1896. So that's a very interesting thing, from an observation perspective. And there have been a lot of proven cases, if you think about it. Let's look at the American economy, right? The 20% of the American workforce, maybe not 20%, maybe 10%, that actually drives 80% of the economy, right? And you can apply that to your personal life, like 20% of the action that you take drive 80% of the results. So there is some correlation.
16:48 SK: But what's gimmicky about this 80/20 rule, to be honest, is that it's not really 80/20. It's just a placeholder for a lot of different things, like it could be 10/90, it could be 70/30, it could be 5/95. There could be a lot of different percentages. So what you're saying is basically, the 80/20 rule is sort of gimmicky, because it's not necessarily tied to the reality and practicality.
17:14 JR: It's a way for you to explain something without having to actually dig in and talk about it, or spend time on it. It's almost, like I said, a cop-out. You're saying, you know, you go to management, and you say, "We're gonna use the 80/20 rule on this one," and they get it. And they say, "Okay, good." And it's your default, a cop-out.
17:31 SK: It is, yeah. It's kind of like you were trying to cut corners. You don't necessarily have all of the data and all of the insights very well thought through. You're not looking to details, and you're just making a general perception about it, which, yeah...
17:46 JR: Not cool. Okay, so these last two... These are kinda funny. [chuckle] This next one here is more along the lines of trying to CYA, because you want people to know that you're using data rather than just kind of making it up out of thin air. But the term is, 'data-driven decision'.
18:09 SK: The best for the last.
18:11 JR: So, it's almost like, "Okay. Well, what other kind of decision would you be making if it's not a data-driven decision? Or you just... [chuckle]
18:21 SK: Yeah. And it's interesting to say that, because especially if you live in the organization and we start talking about kind of in the... With the funnel flow, which started with the 'data being the new oil'. We talked about 'analysis paralysis', 'dark data', so ultimately, the decisions today in most organization is to say, is being driven by data. I don't think people... Very rarely it's gonna happen that people are gonna make a decision... I don't think that doesn't happen, that still happens. A lot of people think because they like a particular person or they like what they're doing. The like has a lot of value today, so it doesn't matter what they do...
19:02 JR: It does, yeah.
19:02 SK: People are still gonna not look at the data but they're just gonna follow their likeness to take a decision. So this kind of is, takes that approach where we're just gonna say, "No, make your decision based on data." But for people like us who are daily involved in business and in analytics and digital pieces, we all know how data is important. I think the entire world is getting more and more data-centric. So then it becomes pretty cliche terms if someone comes to you and say, "Make data-driven decisions!" [chuckle] Which is pretty obvious that with the economy that we live in and the structure that we are a part of, we are definitely gonna use data to make decisions.
19:47 JR: I think that goes into actually the last term of the day is, 'Always be testing'. So there's a good and a bad to that. So the good to that is, yes, we need to continually test because things change, seasonality, buying patterns, macro economy, micro economy, all of these different things affect your numbers and you should always be testing to make sure that you're effectively doing the right thing and not just doing the same thing that you've been doing 'that has worked'. But the bad part of that is you could over test. It's almost like the analysis paralysis to where some people just get into this mode of always testing. And they never actually do the work, they just become test monkeys, you know what I mean? There are some people within a business who, all they wanna do is just test and test and test all day long.
20:39 SK: Yeah. And that sometimes again leads to something that is not worthwhile. Because if you have data that's telling you a good story and you can use that story to draw your conclusion and essentially make a data-driven decision, you don't have to always be testing. You don't necessarily have to apply the laws of that, you still have to segment and test it. As long as you know that you have done your homework, you have done the research, you have enough information to make a decision, there is no need to test further. And it exactly just leads to the analysis paralysis which is what 'always be testing' methodology is. And it doesn't matter if you are testing between two different data segments, it doesn't matter if you're testing between two different creative segments or two different data models, if one model is better, just go with that, in most cases if it's better than the other.
21:38 JR: Double down, man.
21:39 SK: Double down. [laughter]
21:40 JR: You're gonna double down?
21:40 SK: Double down and gobble.
21:43 SK: That's hilarious. So I do have a couple of notes before we finish... At least a couple more than I wanna point which is very interesting. So Tim O'Reilly in his 'Web 2.0', I believe that's a book, he coined the term, 'Data is the next Intel Inside'. Which is pretty interesting, this Intel Inside ad that always, we are used to seeing. What he said that data is actually the next Intel Inside which kind of feeds in to the next one which is, 'The rest is built on top', which is coined by Daniel Keys Moran. So he... Both of these are tied together, because if data's inside, then everything is built on top of the data. The infrastructure, the information, the technology, everything is built on top of data, like your cellphone, a simple example, smartphones. It's primarily running on data that's going between all the different capacitor, transistors and essentially providing all the information that you need to work without worrying about what's actually driving force behind it, which is the data.
22:46 JR: Good one.
22:46 SK: Great. So this is kind of a short one, but we wanted to have some fun with this analytics and big data cliche terms. So hopefully our listeners enjoyed this short and sweet podcast from us and leave us feedback where, again you know where to find us at datacrackle.com or jeremyroberts.com. Leave us feedback, write us a comment, we're always looking for it and thank you so much to our listeners for following us. You look at our show notes on the websites.
23:16 JR: Yep. And I'll see you guys in the air. Thanks, man.