#029: (Reflections on) The History of Digital Analytics with Jim Sterne

February 2, 2016

Philosopher, poet, and essayist George Santayana wrote, “Those who cannot remember the past are condemned to repeat it.” We thought we’d have him on to reflect about the history of digital analytics…but he died in 1952. Ambrose Bierce wrote The Devil’s Dictionary, which we think is brilliant, so we thought we would have him on…but he died in 1842! Lucky for us, we landed the best of both worlds with very-much-alive philosopher, poet, essayist, DAA founder and chairman, and eMetrics founder Jim Sterne.

People, places, and things mentioned in this episode officially ran a full, certifiable gamut:

Show Transcript

The following is a straight-up machine translation. It has not been human-reviewed or human-corrected. We apologize on behalf of the machines for any text that winds up being incorrect, nonsensical, or offensive. We have asked the machine to do better, but it simply responds with, “I’m sorry, Dave. I’m afraid I can’t do that.”

[00:01:05] Hi everyone. Welcome to the digital analytics power hour.

[00:01:09] This is Episode 29. You know every nation has founding fathers and digital analytics nation. That’s a term I just made up is no exception. Who are these brave men and women from our past and how can we find out more about their short but rich history of this most amazing of disciplines. It sounds like a job for our guest tonight. Jim Stern you probably know him best from his cameo appearance in episode 26 but actually Jim has led the charge for digital analytics in digital marketing for almost 25 years. He started the most recognizable Digital Analytics Conference metrics. He cofounded the digital analytics Association and has written three books on the topic of Digital Analytics and he got Chris decided to start drinking scotch. So sit down pour a drink as smooth as his voice and relax. This episode of The Digital Analytics Power Hour is about the history of digital analytics with none other than the Godfather himself. Welcome Jim.

[00:02:17] I’m a little overwhelmed and a bit for CLEP thank you so much for that introduction of overwhelming clamped.

[00:02:25] Let me also bring in our other two Cohoes to Wilson senior partner analytics demystified. I am giving a running list of vocabulary words and to Google. During this episode. And of course our neighbors to the north.

[00:02:42] Jim Kane CEO of napkin and Babbage systems. How most of you know that I am Michael Roebling and I lead a small company in Georgia called Search discovery.

[00:02:56] So I had to go to my the first dictionary I looked and did not have a definition for Klemm because that was the devil’s dictionary.

[00:03:03] But I did. I did.

[00:03:05] I did find it on the interweb. So I’m all caught up now.

[00:03:09] Carry on. So I think most people would want to start off right away. Jim Stern with the question What were you thinking when you invented web analytics.

[00:03:23] So here’s where the name godfather really comes in handy because I was not there at the conception. I had nothing to do with the birth but I am devoted to making sure that this industry grows up and lives a long healthy life. So godfather is pretty appropriate. Just like the Internet.

[00:03:47] This is something I tripped over rather than had anything to do with inventing this does kind of make me feel like we should have a dramatic reading from the devastated dictionary for the definition of algorithm could you deliver that from your book a selection.

[00:04:01] Are you kidding I travel with the book in hand constantly. Algorithm is defined as regularly recurring remarks from the former U.S. vice president who invented the Internet.

[00:04:12] See how I brought that together.

[00:04:14] I feel like I studied up for our guest use the definition in the book that anybody over 40 years old gets even though they don’t get any of the others. My dad like that one. And he’s 89.

[00:04:29] Well and since we’re kind of regular anti millennial it’s OK if the other people who are under 40 don’t get it right.

[00:04:35] If somebody comes along who gets all of the definitions and understands all the jokes it will freak me out.

[00:04:43] Did I tell you that actually my mother was in town and I made a comment and said Oh you’ve heard me talk about this gay wrote this book because she’s an English major so she certainly gets the Ambrose Bierce reference. However I had to leave the room after about the fourth grilling she was trying to get me to explain the humor. One of the definitions because she not only had like never heard of the word and then did any of the jokes behind it. I was like I’m so sorry. This was a poor choice on my part.

[00:05:13] Tim I am so sorry I apologize for putting a rift between you and your mother but for your listeners benefit. I would like to introduce Ambrose Bierce to those who do not know who he is. This is the guy who wrote The Devil’s Dictionary in 1906 and it is full of stuff it’s a big book and it contains really choice gems including bore a person who talks when you wish him to listen.

[00:05:40] Love a fairy insanity. Curable by marriage politeness. The most acceptable hypocrisy and success.

[00:05:49] The one unpardonable sin against one’s fellows. I take no credit for those but I have inspirer. Did he have a definition for web analytics or let the doubters did that predate the history.

[00:06:01] Just just a little bit in 1996 I didn’t get that far in the book.

[00:06:06] So when exactly did this I should know this.

[00:06:10] What do we date the start of web analytics to twenty five ninety five and that was long flight 93 was the browser was mosaic and by 95 we had commercial available stuff and open source although they didn’t know to call it that. So there was sawmill to do logfile analysis web trends came along in 1995 that genesis came along in 1995. And that’s that’s where it began.

[00:06:39] Analog I think came out that year. Turn back Mike.

[00:06:42] My cousin was part of the original Mosaic team that is fun fun fun fact web analytics is 10 years younger that back to the future.

[00:06:52] I thought we’d move beyond back to the future at least rather than Facebook. So that’s all of that.

[00:06:58] I was standing on my toilet trying to hang a clock and I fell and hit my head and that’s when I came up with the idea for the web trends.

[00:07:05] It was a dark brown reference.

[00:07:08] Well done. Yeah. So. So there is a really interesting thing to talk about really briefly and it gets onto a couple of different tangents but we’ll see where it takes us. Web analytics did come from first and foremost the log files coming out of websites and people started looking at those. And so that was primarily a technical activity. And so the first tools were being used by technologists and Aidid organizations. When did people start to really make that corner of Hey we could use this in the context of our business right because that’s certainly where we are today. But like let’s talk about that. What were some of those first uses that you know you’ve seen Jim Stearn and others.

[00:07:55] Well so that would be 2002 because that’s when the first Imitrex summit was 50 years ago. There was a shouting argument in 2002 the metrics in Santa Barbara between two guys wearing Birkenstock sandals and ponytails about whether logfile analysis or javascript was better than to tell us what was on which side or Evans was one of the technical founders of a company called clickstream in the UK that did packet sniffing and the other ponytail wearer was Eric Peters. And they got into a heated argument that extended all the way into the lobby bar hence the beginning of the lobby bar and the result was they ended up agreeing completely that both of them were necessary.

[00:08:48] So it’s just that the inheritors of the ponytail bar debate.

[00:08:53] That’s right. But it does seem like that transition and I count myself back to about 2001 when I inherited that Genesis implementation. And so in a sense I count myself lucky in that between migrating from net Genesis to web trends I got a a deep technical education on what goes into a logfile and what’s in the header and what can be recorded. But there’s part of me that while the data was always messy like with the net Genesis stuff it took forever to run because we were we were basically once a month cranking out a ton of reports that took increasingly long to crank out.

[00:09:33] And then there were the static report and then they put MicroStrategy on front of it and said oh now you can can slice and dice and you really couldn’t not in an effective way. So we were battling it on the one hand I feel like back then I was battling the shoot. I don’t have that data or I don’t trust that data or Gomez was doing a sales pitch and they ping this page and they fucked up the data. So fighting kind of data quality. Ever since then but on the flip side I feel like the old guy who says yeah but when it was just when digital was really just our site and we were just discovering AdWords as a way to drive you know pay to drive traffic to the site. Things did seem simpler although I’d say it was still incredibly guilty of you know just counting counting visits and page views. I feel like the data has gotten more complex and we haven’t kept up with getting the data consistently clean.

[00:10:29] Like that’s a never ending battle but the behavior of the consumer and the number of digital channels that all fall under when we transition from Web site analysis to digital analytics that feels like just such a much more involved and complex scope and I don’t know that I can live through that transition and was oblivious to it through the entire change stuff just kept getting more complicated.

[00:10:57] I remember when the WEAA changed its name to the DA and I was like Come on guys you just trying to get cute and then I started to think about it more and more and it was like that’s a good step towards not marginalizing the web analysts.

[00:11:10] Yeah it’s almost like we had a brief period there what it must have felt like for advertisers when there was only three television stations. You know it’s like we just have a website to measure. That’s all we’ve got to worry about.

[00:11:24] And when the WEAA started the first meeting of what would become the board of directors was what do we call this thing. And I was adamant that it should be customer analytics and I was shouted down because no no no we want to draw attention to the fact that this is a new dataset that nobody’s ever seen before. It’s the Web site. It’s web data. It’s it’s behavioral data stuff. You’ve never been able to analyze before it’s not CRM it’s not database management systems in doing direct mail analysis. This is unique. So it’s the web analytics association. So I lost that battle then six years later the board had the discussion again. Shouldn’t we change our name because web data is such a narrow thing and we’re measuring e-mail and search and campaigns and social media and oh by the way when they run an ad on TV they ask us to measure the impact that it’s having on Twitter. So clearly we’re beyond web analytics we should be the digital analytics association and I was adamant that we’d really needed to be the customer data analytics kind of you know the customer relationship with the company online the digital experience. No no no that’s too narrow. We might want to take over the world someday. So that’s twice that I’ve lost in the battle of what we should call this thing.

[00:12:49] It surprised me as I’ve run into the more companies that I work with that when we talk about customer analytics and I know that I absolutely get the need to have be truly customer centric and there’s been some great writing on what is customer centric even mean but regularly run into organizations that have you know customer analytics lives in the organization and they are the ones who are cranking through the customer data warehouse the CRM wherever the customer data lives. And it’s not even occurring to anyone that that world should be blending and merging with the traditional web analytics world. The funny thing is in my when I got into web analytics I was very quickly in an organization where all of that was in one B.I department and I just didn’t we didn’t know any better. It just seemed like that was all the data we should have are our data warehouse our customer data our ERP all that should be kind of in one. Our market research should be in one centralized group. But wait. We’re talking about the history. Maybe this is more of the future and I don’t know what you guys see as well. I certainly still seem to run into places where they say we have no digital analytics. Yeah we’ve got a whole customer analytics team we’ve got 20 20 customer analysts and it’s not but iPod was not going on that hey maybe that group should also be the experts on the on the digital beta.

[00:14:22] I have maintain my sanity by dividing all of this up into four really clear areas B.I has traditionally been about internal business processes. Let’s manage shopfloor control let’s manage supply chain let’s manage efficiency of our employees. Then we’ve got customer analysis which has been the mailing database direct mail database which then sort of became customer relationship management and sales force automation. And then there was market research. Let’s go out and interview and survey a gazillion people and produce reports that are general across the whole thing.

[00:15:04] And then came along this weird little thing that was behavioral data. Nobody had ever seen mouse clicks before determining intent based on what you type into a search engine.

[00:15:16] This was so unique and so new B people didn’t know what to do with it CRM people said it won’t fit my database I can’t handle time series data flows that big. So a unique group a unique industry was created digital analytics. How do we measure the ongoing touchpoint relationship between customer and company so that we can make mobile better we can make the website better we can measure whether or not our campaigns are working and whether or not our search keywords are correct and this is so different from the structured data mainframe mindset that it had to be a unique different group of people.

[00:15:58] It’s funny.

[00:15:59] I told this story before my the way I got into what was a B.I role although it probably fit more under the customer analytics definition was almost exactly market research lived in the advertising Web analytics lived in Markham and it kind of moved me through a couple of things from our community to Marcom B.I what we call B.I. Maybe it was more customer analysis although we sort of touched all sorts of stuff and I was having to logically it just made sense to several of us to move the behavioral to web analytics into the B.I or in this case the customer analysis group. And after having that conversation of saying hey does this make sense. And they said sure you know if you’ll help kind of hold our hands and help us understand what the data is we’d love to get our hands on that data it totally makes sense to bring his behavioral stuff together with the customer stuff and then as I was leaving that meeting I said all this kind of sucks. Is the one part of my job. I was a market manager basically when Mark manager. This is the most fun I have with my job it just doesn’t fit in me. I don’t have the bandwidth for it. And they said oh what’s keep talking so I’m going with it. And it was inside of a year after that that we said why is the market research department off in a totally different group than they were kind of orphaned in the group they were in.

[00:17:25] So it wasn’t like we were doing any sort of a land grab it was just a little two person department and we found ourselves talking to them all the time and said Hey wouldn’t it make sense for you guys to be with us as well. And I continue to be amazed that this is it sounds almost like a humble brag. I mean we literally were just oblivious. It made so much logical sense. And now with as many companies as I’ve seen how they’re organized. One the ones that don’t have a behavioral digital analytics staff or they don’t have market research where they don’t have any of those groups or when they do there’s not a clear direction that these groups could come together in maybe just a little bit with what Gary Angel talked about on the last episode that that’s just kind of organizational evolution that we just kind of gravitate to to silos and that seems like for a for something that’s now been around since nineteen ninety five that we should think that Apple would have gone on by now that the light bulbs gone off before people are going on rather it’s to go back to Jim’s statements earlier about why was it the DA or why was it the W.A..

[00:18:37] If you call it Web analytics or digital analytics you’re building a very valuable house but you’re building it on land. Nobody else want it. There are at the enterprise level preexisting teams of people that do customer analytics then just like you said in that role and there was there was a team of people who do behavioral analytics and they happen to get along with you that if you go and you rename the D.A. the customer analytics association you’re shooting a flare in the air you’re trying to eat someone’s lunch that becomes a thing.

[00:19:06] One is that is that the DMA would they say that they are the customer analytics association now understand that the DEA DMA started out as the Direct Mail Association. Then they changed to direct marketing. Now they want to change to digital marketing.

[00:19:20] It’s it’s humorous it is fortunate with their first letters of the evolution and I’m actually I’m not lobbying for keeping any name what it is.

[00:19:33] I actually think it’s important that if the analysts don’t start to say actually influence or direct key pieces of the business are just web analytics data we’re going to be marginalized to the point where someone eats our lunch and then the next thing you know we all have to show up to a bullshit meeting.

[00:19:55] Well I mean it’s so interesting that these silos happen in organizations. But then as you get good you break these silos back down again. And I wonder if let’s say the silos didn’t exist would we master this dataset. Where would we still be approaching it kind of the wrong way of doing it in the ways that the other groups would try to do it.

[00:20:20] I’m going to throw an orange flag on the field. This is where I have a problem with the democratization and that is the only way you can put all the data together in one place and have it work is if you sprinkle machine learning on top of it because every dataset requires a subject matter expert. And if you’re an in market research and doing surveys if you’re an expert in television ratings if you’re an expert in sales force automation tools each of those are really unique realms and you can’t just throw all that data together and expect it to make sense to an individual. So I don’t think that if there were one place in the company where all the data feeds came that it would magically have value. I think every data stream needs a steward really understands the deep darkness of how trustworthy the data is how reliable it is how fresh it is.

[00:21:19] We used the word curator a lot but so said. Totally agree.

[00:21:24] I agree with that as well. I think maybe the non sequitur. For me there was the necessity for machine learning. But maybe I just didn’t understand that. You wanna expand on that a little bit.

[00:21:38] So the thing that machine learning is really good at is finding correlations. So there is a correlation between the sales of ice cream and the number of people who drown in a month. And for the machine to go Oh look there’s a correlation is useful to the human the human can look at that and say Well of course because the temperature goes up more ice cream is sold and more people go swimming but there’s no causality here so thank you but no thank you. And then the machine says hey there’s a correlation between when it rains and people buy umbrellas and the human says thank you Captain Obvious. That’s true but it’s not useful. Try again. The machine says you know when this particular kind of person who exhibits this particular kind of behavior. Who came to us from this particular source and responded to that particular email is 37 percent more likely to purchase if we send him a message that has this benefit statement.

[00:22:35] I say thank you so that I’m with all of that well but I’m skeptical to one further to agree on that because there is still a need to have variation in activity. I think that’s the other place where things start to fall down.

[00:22:55] What does that mean. Well so are very very simple simple examples social media Facebook back in the what. What’s the best time for us to post on Facebook. And this is kind of a gross simplified example of where if the machine isn’t fed it can’t do anything with it that when it’s like well we want to know what the best time of day to post on Facebook is. They say well OK we’re going to look at all the posts you’ve done for the last six months. It turns out that you’ve only posted between 8:00 and 10:00 a.m. every single Facebook post because that’s the process that the agency or the company set up that somebody thought. This is the best time to post this. That’s when they post. So you can’t find a correlation between oh if we post after 10pm we get higher engagement because there’s no data there. So if you take your example of within this type of email what that means I have to I have to be sending a range of types of e-mails to a range of types of people. And all too often we get in the What’s the simplest path what’s the one email we’re going to send it the one time. Maybe we’ll do some subject line testing. We’re going to be in this rhythm and we say we’ve got all this data. I guess you have all this data but it’s all largely the same because you haven’t tried anything different.

[00:24:13] So you need to kind of back up and have a little bit of a design of experiments approach and saying well how do we want to mix things up so that we’re generating some data so the machine has something to play with.

[00:24:24] And this is why Skynet will never eat your lunch and take your job because Skynet is really good at stuff but it’s really dumb and other stuff.

[00:24:32] You know computers are really fast but really dumb and humans are really smart that really slow and you’ve got to put them both together.

[00:24:39] Well next time I’m eating some ice cream by a pool and a robot slaps it out of my hand I’ll know what happened to save your life.

[00:24:49] Just saved your life. Wow. So that took a anyways right where we were with.

[00:24:56] So question for me. You met a lot of companies that do measurement. I’m sure a lot of them have brought you in. And everybody on this call today is making money off companies that know what winning looks like and measurement that they need someone to help them do it and then they’re all missing different pieces of that equation. Can you give an example of a company where you walked in and you were preparing because you know like it’s not your first rodeo I’m sure when you go into a company you go when I ask this question they will say no and I will help them fix it. Like you know the dance steps. Can you kind of share an anecdote about an organization you walked into. We were like whoa shit that you don’t need me let me write this down.

[00:25:33] Yeah there were two of those. The first one was Hewlett-Packard. So when I wrote customer service on the Internet in 1997 I interviewed a guy at Compaq Computer and he was getting it. I mean he just totally had it wired. But I was talking to them the day that it was announced that Hewlett-Packard was going to hire them. So a year later he calls me up and says hey we need a workshop at Hewlett-Packard because the Compaq Computer team became the head of all analytics for HP and I’m running it and we need your help went. You’re right. So I go to HP. We spend the first half of the day where they’re telling me what they’re doing and I am just blown away.

[00:26:18] They are doing conversion analysis and source analysis. And how is my dollar being spent and where is it being wasted. Perfectly it’s just beautiful I’m ready to read another book. And we go out to lunch and I turn to Seth Romano one of the founding board members of the WEAA and head of analytics for HP and I say Seth I’m I’m out.

[00:26:43] I am beyond you guys are just like totally in the groove. I can’t there’s nothing I can do for you. He says that’s fine you’ll be fine. No really I can’t. I cannot send you an invoice. I shouldn’t even be here. You guys are awesome. I’m writing another book about you said well just you know c’mon back in like let’s reconvene and we’ll talk some more.

[00:27:03] And I got nothing. So I pull out of the bottom of my bag. The last question that proves to the audience that I have no idea what I’m talking about. It is the consultants last ditch effort. I say so guys tell me what’s the hard part. And it was like they picked up the garbage can and dumped it on the table. Well this doesn’t work and this technology is broken. And this executive won’t give us budget and these people don’t believe what we’re saying and these people won’t do this. And it was like oh guys guy settle down. Yes I can help you. The second time was in Boston that the last Metrick summit I went out to dinner with Julie and Stefan Himel and Josh Averitt who is the postmaster at Twitter who had given a keynote speech in Boston and we said so you know here all the problems that we’re all experienced with our clients and you know they’ve got data silos and we’ve got legacy systems and we’ve got people who don’t get it. And we’re trying to convince upper management and we’re trying to get more budget and he’s looking at us like we’re from Mars. He says I’m sorry I cannot relate at Twitter. We have one data set and we have one goal and we all work off the same data and we all believe in data. And a year ago the CEO got up at a company all hands meeting and said If anybody wants to run a 1 percent test they don’t need permission. They just go ahead and run the test.

[00:28:40] We all went oh we can just like hypothesize out the Kazue and test everything we want and then we realized that if you ran a 1 percent test on Twitter and it didn’t work you just pissed off six million people for a whole day. Your job is in trouble so be careful what you’re testing and then they ran all these tests and they would go Wow look we found this and we found this and we found this and then they discovered that they were up against the wall of I found this really cool correlation. And the first question was always well really what experiments did you run. What does it mean. So June Estefania and I are looking at this guy going well this is nirvana.

[00:29:18] This is a giant corporation that has one data set. There is one source of truth. They believe in data. They love testing. They’re doing it right and it’s still really hard.

[00:29:30] And then just left two months later and went to smart post right.

[00:29:35] He went to spark post where he has been doing some further research and then doing research about what Google and Microsoft are up to. And he will be keynoting again in San Francisco at the metric summit in April. So we’ll get a bigger story.

[00:29:51] It sounds like in both cases you walked in and there was a good story or there was a story of we’ve got it all figured out. But it sounds like is you probed farther. There really wasn’t wasn’t really nirvana.

[00:30:08] There is no end game. There is there is not a solution to this problem. It is data which is messy and difficult to understand and requires a curator for each data set. And there is politics and there is legacy and there are people who are not very bright who have too much control over the process in every company at every level.

[00:30:35] I like how the two companies like one of them was an old school enterprise that had a winning team but kind of organizational entropy and the other one had organization at the speed of light. And people were just getting like ice cream headaches.

[00:30:48] You don’t have an opposite problems on both sides. Yeah.

[00:30:54] Yeah I’d like to turn our attention toward the venerable hippo for a minute because they entered our digital analytics world and caused a lot of turmoil and yet the three of us find ourselves defending them quite often. When when did that really start entering the lingua franca if you will of the analytic community. I’m not sure of the timing.

[00:31:23] As soon as it became about money and the culture of this stuff goes way back 1993 is when mosaic came out and the only people that had websites were called webmasters. It was Dilbert in the bowels of the I.T. department and he built a Web site out of a server that he downloaded somewhere put it on an old 386 machine under his desk and hooked it up to ISDN line and eventually somebody in the company realized that they had a website and that the brand was being represented by Dilbert and the marketing department took it over and said we’ll Dilbert we’re announced a new webmaster and Wilbur had never been in marketing and marketing had never seen technology before and it was a mashup of people who had never talked to each other before and the web master was responsible for all of the creative and display because people said we want this picture. So we can put up a PTF. But if you want something with buttons I’ve got to do all this stuff. Same thing happened with analytics. It was the webmaster who discovered there were log files and so he said Oh there’s no look at all this data. There’s got to be a pony in here somewhere without any clue that there was a department without any clue. Had never had statistics classes. He’s just looking at these static reports from that Genesis and going wow I got money for this product and now I’m getting these reports and I don’t know what they need. And that’s the way it has always been until eventually the website was actually making enough money that somebody upstairs said oh this is important.

[00:33:03] It shouldn’t just be something we’re doing for fun. We should invest in this and therefore we should manage it and control it. And my wife likes blue so all the buttons should be blue.

[00:33:14] And that’s how the hippo came to be because of some of what can I throw one more idea in because they know we’re getting close to time.

[00:33:21] I’m getting close to drunk throw another idea and then every other big marketing group has a lobby like beyond the lobby bar.

[00:33:33] The lobby group. There it is. That’s a podcast title. There you go. The IAB spends a lot of money in Washington that is on the web analytics association.

[00:33:44] We spend our money literally at the lobby bar right there.

[00:33:53] And so when people look at our community they confuse us with people that do dirt bag things with online data.

[00:34:04] I was reading something about Mr Sanders and Ms. Clinton and they have access to this democratic you know member database and weird things happened and it immediately starts turning into analytics and analysis we get painted with some pretty shitty brushes and analysts tend to be some of the more conservative people about what you should capture and how you should maintain and do things with it.

[00:34:28] Are we at the point where it could be in your opinion and name change.

[00:34:32] It could be you know everybody is in the game as to put a few dollars in the hat so we can start to lobby or position ourselves differently than some of these these you know advertising organizations. You know is there a big next step for us.

[00:34:46] As a community. Good answer thank you. You’re welcome. Glad we’re.

[00:34:53] So when the W.J. started it was a considered decision to be inclusive of practitioners US consultants US and vendors.

[00:35:10] Now if you can figure out a lobby platform that would benefit all three of those you are a better person than I am.

[00:35:20] So we just said look if we’re going to be a lobby organization we need to go to each corporation and ask for fifty thousand dollars a year to represent them in Washington D.C. and have a very clear platform.

[00:35:34] B They have a clear platform there. They have a purpose. They represent a segment we don’t we represent an area of inquiry that is inclusive. So from the very get go we said we’re not going to do lobbying. It’s not who we are.

[00:35:54] That’s interesting. And I love that you just said that because that really helps position the DA as your organization that it is. But it does make one wonder in the future is there a separate organization that will need to come up to you know content for the interests of actual measurement and doing it responsibly in ways that businesses can use it to do the things to make the customer experience ideal.

[00:36:22] I don’t see that as really possible because you are advocating for and against collecting data. So we did have Eric Peterson did come up with and posted the standards of ethics you know.

[00:36:40] Do you agree to this code of conduct. Yeah the code of ethics you ran around like that I signed it I think the first week. I absolutely I do. I never signed it. I’m telling because I am my just told you I don’t know why I didn’t sign it because I certainly agree with it.

[00:37:02] Michael can neither confirm nor deny that Dr. credit card numbers had it right back they say confirm or deny.

[00:37:13] I have such a schizophrenic relationship with privacy that I really have trouble discussing all of this stuff because on the one hand I do not trust corporations and I do not trust governments at all with my data. On the other hand I will gladly tattoo a barcode on my forehead.

[00:37:38] If you’ll get me on the airplane faster or get me the products I want to my house or what I need before I need it you know be the artificial intelligence all led to believe could happen where’s my flying car or you know Google on your phone predicting what you want to know next which is totally happening every time I get in my car Google says how many minutes it will take me to get the office.

[00:38:06] I’m not going into the office right now.

[00:38:10] How could he be so stupid. They’re critique. Godless yeah. Reiffel wait till they start driving your car. You get in your car you just show up at the office be like. This is not where I was going guys.

[00:38:24] Self driving cars cannot happen soon enough. So I’m not excited about. I can’t wait till I can just sort of be reading the devil’s data dictionary which is commuting to the office. I will never need to get my license.

[00:38:37] I get that whole phase of adulthood entirely and steps that are made until I can finish this bottle of Jamison’s without fearing for my life I get pulled over.

[00:38:47] Have you been drinking. Hell yes. OK. But then texting. Thank you very much. The car said nothing but 220 volt the pure the good the goods. Right. And yeah if you have electricity. All right.

[00:39:05] So just that just because it’s time for a break in the rhythm I’m going to read all of the A’s from my wife or not. Not that many.

[00:39:16] So the first one the first entry in the Devil’s data dictionary you have already heard its algorithm regularly recurring remarks from a former U.S. vice president who invented the Internet. This is followed by analysis when a calculation requires more than 10 fingers analytics same analysis but garnished with red pepper rings micro greens and Chickaree.

[00:39:45] Analysis paralysis. Thursday API API is a gateway drug that leads to pulling more data than rational from systems gathering data faster than sensible for reasons more aspirational than comprehensible with feeling. I like that one assumption is the part that’s always wrong and the A’s finish up with average obscuring the interesting or valuable through homogenisation. In other words on average Switzerland is flat now saying you skipped one.

[00:40:25] A It is the most appropriate for this entire podcast. Oh that Fraser Mr. brevity would you please define artificial intelligence artificial intelligence was included.

[00:40:38] You must have already been artificially intelligent because I did not read that out loud. Artificial intelligence is the fourth cocktail I swear maybe I had the fourth.

[00:40:49] I think we went straight through to assumption that now that’s an assumption. We’ll have to listen. So that is the title of this podcast is called The Fourth cocktail summit interactive. I’m like Hey guys I’m keynoting in the morning. Maybe it’s time because I and I say Hey Jim I hate him I got invited to you.

[00:41:11] I’m going to be speaking after lunch on the last day of the show. How did you get a key note.

[00:41:19] All right well it’s that time of the show where we go around and we reflect on maybe what we’ve learned or share an observation. Who wants to go first. Now go first.

[00:41:33] Yes a little I have been a little depressing. You’re welcome. I feel like we talked about the history that could not help but morph into a discussion of the present and that that also led to a little bit of the future. And there’s a lot of stuff that maybe stops just short of intractable from challenges organizational the data the complexity the data the aspirational beliefs and what automation you know can and will do.

[00:42:15] Tim I just like you to change the word intractable to job security job security.

[00:42:20] Actually what I was the key security is looking up the I can continue to be marginally above mediocre and still be gainfully employed.

[00:42:32] You two are from Lake Wobegon yes.

[00:42:36] So I think that’s kind of my I mean I’ve enjoyed the discussion it was it wasn’t picked up a few kind of historical anecdotes. But I I would say one of the last things we covered the why the D.A. is unlikely to ever be an effective lobby organization. I don’t think I’ve heard it put that clearly as to why you don’t have common interests records by design. The DA has diverse interests represented and it would be hard for them to unify and design but by design. So that’s kind of fascinating but also mildly depressing. So thank you. I’m going to go and shoot myself now. No not job security never never. I mean I would not do that because I’m going to remind myself that I have job security. There you go.

[00:43:23] All right. I’m going to do this thing where I jump right in front of you. Jim

[00:43:26] Kane I was having fun until Wilson got all Darkin shit. I know I did not take that from this conversation.

[00:43:36] I was delighted to share a brief anecdote about my first metrics conference. I will share a couple of details in the hopes that some of our listeners can use some of these tips to themselves come to any metric some day. I went to my boss and I said there are two conferences I want to attend this year. The metrics in Santa Barbara and this search conference over here and my boss looked at me and he’s like I don’t know. I don’t know if we can do two conferences in a year. I said okay I’ll just do the one then the metrics Santa Barbara. So I was fortunate enough to attend the last metrics in Santa Barbara in 2006 and it was life changing experience. And I owe you Jim Stern a debt of gratitude for fostering a community so open and friendly and willing to share their knowledge and experience in a context like the lobby bar. It has enriched me as a professional for years and years and years. So this is my school he just so why I asked. But it’s true. You know if you think about the work of this industry and the people of this industry where they have intersected over the years in the most meaningful ways has been in the context of the metrics conference. So it’s something I’m really you know it’s something we’re always trying to recreate together right. And so anyway I’m going on too long but I’m very glad to do that you were able to to be our guest and join us on the analytics power hour even if it’s mushy.

[00:45:30] Jim Kane is you know nervous about his feelings.

[00:45:33] Just let him out just now. I’m a good Irish boy attacked him into a ball and buried him so deep. No kidding kidding aside.

[00:45:44] When I started so ten years ago and I hadn’t started napkin I was still selling that started it is an analysis and I sat on the standards committee and so I very clearly now understand why it can’t be a lobby organization.

[00:45:57] That was not a fun club. No I was tough. But separately when I and Tim Tim and I in particular have talked about imposter syndrome before when I started getting going. One of the first things I got to do I started napkin and it was me and a bulldog in a living room with a skype phone. It’s hot. There’s a smart dog. But there’s little pause just couldn’t work the Axelle so I had them put down but that I don’t know.

[00:46:23] That’s right.

[00:46:25] But you know Jim let me speak of the metric Strano and that gave me a lot of confidence let me meet some people. It helped make some of impostures syndrome go away. There certainly wouldn’t be this podcast if there wasn’t any metrics that tell you that because this is literally what we’re trying to do is it’s pretty cool to have you on the show.

[00:46:41] Really appreciate your comments. It’s kind of why I do this for a living and will continue to do it because people end up saying stuff like that.

[00:46:51] So thank you Tim. Please don’t shoot yourself Jim. Don’t shoot your dog.

[00:46:57] Well it has certainly been a pleasure. And of course if you are listening out there and you want to get in on this conversation I think you may have noticed one thing as we’ve talked on this show today and that is Jim Stern is a really approachable guy. And so if you ever see him at any metrics conference lobby bar you could probably sit down and have a chat and buy him a drink. You may be surprised some person from Europe probably named Renee will come over and give you a chocolate know. And if you’re listening and you’ve got comments we would love to hear them. You can always reach us on our Facebook page or the major slack which you can also find from our Facebook page. And of course we’re on Twitter as well. I don’t know whose bright idea that was but apparently we might be running 1 percent tests on you on that. So just be careful what a great nostalgic show. My spirits are lifted. Apparently.

[00:48:04] I live opposite effect zero sum. Yeah.

[00:48:09] It’s it’s great. You get a chance to sort of peel back the layers and think about where we’ve come from and how that points us to where we might be going. And you know as you as you are listening how you can play a part in that obviously there’s a couple of things that we talked about tonight that probably should be of interest to everyone. The metrics conference certainly worth attending there are multiple locations in the U.S. and abroad. So definitely look for that near you this year. The Devil’s data dictionary I think aptly displayed this show is noticeable in that it should be a part of every good analyst desk reference. You should have that easy to hand. So definitely go and get a copy of that and of course the DA which is not a lobbying organization but as an organization for you the practitioner consultant or vendor so become a member of that as well. So Jim Stern thank you once again for being our guest we had a great time. And as always for my Cohoes Tim Wilson and Jim Kane keep analyzing.

[00:49:26] Thanks for listening and trying to get to join the conversation on Facebook or Twitter. We welcome your comments and questions. Facebook dot com slash atomistic now. All that and only now on Twitter.

[00:49:41] Smart guys want to fit in. They made up a term called Linny. Don’t work.

[00:49:50] With your indulgence I’d like to read a little poem I call insight insight from the devil’s data dictionary it’s not having the answer. It’s dreaming up the question. It’s pivoting like a dancer then testing with regression. It’s not about the facts. It’s making the connection. It’s that which then attracts you to create a new projection. It’s not about the figures it’s about the correlation and big data just gets bigger with more data integration. It’s not the data warehouse it’s the cross association it’s not the beans that count but considering causation it’s asking why and asking how and looking for the info that will open doors and disavow hip shooting from the hippo.

[00:50:34] Big data isn’t magic. Big data is not divine. Big data becomes tragic when art is left behind. The human head contains much more than a hundred billion neurons. Their thoughts and schemes and stuff that dreams are built of and are built on human beings can look at a cloud and see a face resolving and fantasize how that relates to the problem they’re solving. Feed your head the Dormouse Said don’t die from dried up facts. Imagination speeds creation and for that you must relax let your brain recuperate and give it room to breathe. Let the sweat of your brow evaporate. Stop the grinding of your teeth. Go for a walk. Have a beer. Go swimming in the ocean. Light a fire in doge desire. Wallow in emotion as one grows older. Einstein said one sees the great futility of imposing your will on the chaos with brute force sent hostility but if you can be patient there may come that moment when your mind is on vacation. The answer bows and says Here I am.

 

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

The Digital Analytics Power Hour © 2019