#063: The Trials and Tribulations of Tool Transitions with Nancy Koons

Change. It’s scary. It’s exhilarating. It’s a song by Churchill. Sometimes, be it due to your manager, due to a corporate acquisition, or due to a job change, you just wind up with a voice in your head belting out, “You want me to change, change, change!” In this episode, Nancy Koons from Team Demystified joins us to dive into our collective histories when it comes to switching analytics tools — where we stumbled, where we succeeded, and how we’ve come to approach the ever shifting landscape of analytics tools.

Linkable Things from This Episode


Episode Transcript


00:04 Announcer: Welcome to the Digital Analytics Power Hour. Tim, Michael, and the occasional guest discussing Digital Analytics issues of the day. Find them on Facebook, at facebook.com/analyticshour and their website analyticshour.io. And now, the Digital Analytics Power Hour.

00:28 Michael Helbling: Hi, everyone. Welcome to the Digital Analytics Power Hour. This is Episode 63. It’s today. You walk into work and you get great news. “Let’s change tools. Yay!” Whatever tool you’ve been using the last how ever many years, we’re gonna switch. Google to Adobe, or Adobe to Google, or Webtrends to Mixpanel, you know the drill. Change is good until change is bad and we wanna talk about it so much. But to make this more real, more interesting, more authentic, we wanted a guest who has lived in the real world and is also cool. Nancy Koons, noted industry speaker and consultant, formerly at Vail Resorts and now part of the Team Demystified Crew. Welcome to the show, Nancy.

01:21 Nancy Koons: Thank you very much.

01:22 MH: And also joining us on this trepidatious, touristic trip into the travails of tools, I did it wrong, there’s a better alliteration someone else will do, is Tim Wilson. Welcome, Tim.

01:42 Tim Wilson: Hey, Michael. Hey, Nancy.

01:44 MH: And I am Michael Helbling. Okay, let’s…

01:46 TW: I was just trying to offset Nancy’s total, proper, “Why, thank you for having me” with a completely unprofessional spaz.

01:54 MH: Oh, see, let’s not fall into these traps.

01:58 TW: Roles.

02:00 MH: Yeah, roles.

02:01 TW: I collaborated with Nancy for long enough that it’s just a rut I’m used to slipping into.

02:05 MH: That’s right. Okay.

02:06 TW: I’ll be inappropriate. She will save me.

02:08 MH: Sort of. I mean, that’s how it works. So Nancy, I think you and I really could get along on that level.

02:15 TW: Should I go ahead and drop off now?


02:16 TW: Let’s change the topic.


02:18 MH: Yeah.

02:19 TW: Therapy for those who had to collaborate with me.

02:22 MH: We’re switching tools, not switching hosts, switching tools, but, Nancy, we’re so glad to have you on the show. Let’s talk about switching tools. First up, what’s the first thing that you take somebody through or you consider when you’re going through a tool transition? I think that’s maybe the first place to start.

02:42 NK: Well, in thinking about this topic, one of the things that I thought about or realized is that, one of the times I made a tool transition was when I was in my same role at Vail Resorts. We were using WebSideStory and then Omniture, now Adobe acquired them. So we had to… I got to pick the new tool, but I also went through that evaluation process. The other time I’ve done a big tool transition was when I switched from Vail Resorts over to Team Demystified and I then jumped from Adobe over to supporting a client running GA. There have been times where I was at the wheel, controlling my destiny, and then there have been other times where the choice has been forced upon me. And I think it’s natural that for an analyst or a person going through that process to have a little bit of angst as they figure out, “What am I even dealing with here?”

03:40 TW: Were you more excited or scared? I have only dealt with… On the Adobe side, I dealt with a client that had Hitbox that then got pushed into Omniture. I can’t remember if that was Omniture or Adobe, where that was. It was a battle. They didn’t wanna move ’cause they were comfortable with where they were. When you’re moving from WebSideStory and making that transition, were you thinking, “Oh, I’m getting rid of… ” It seems like there are transitions where you’re… If it’s forced down your throat that you have to move, ’cause the tool’s being removed, it’s definitely scary. But if there’s the, “We’re so unhappy and displeased with the functionality or customer support or something that we’re now gonna paint an overly rosy picture of the replacement is gonna be.” In the Vail case, was that transition one when you were more looking forward to switching to a new tool or was that one where you were saying, “Well, we have to. What’s the one that’s gonna be the least painful to switch to?”

04:45 NK: It was definitely the latter.

04:46 TW: Is that a fair question?

04:47 NK: Yeah. It was the whole Adobe has purchased WebSideStory. They’re sunsetting the tool, so we have to pick something new. I definitely had stakeholders, though, that were pretty grumpy about it because they… I don’t even remember what they had been on prior to WebSideStory, but they had already been through one tool migration and they were accustomed to having such big differences in reports that when they heard that we had to switch again, there was a lot of eye-rolling, like, “Oh, man, are we gonna have to do this every two years. This is so painful. We are never going to get any insights out of our data.” I had to simply say, “Guys, this is not something we can control because of this acquisition and tried to look at it as, “Here’s our opportunity to go out, look at the different vendors, and pick what we think is going to be the best.” I tried to paint that rosy picture, but I definitely had people who were frustrated that that choice was forced upon us. It was probably easier for me, though, because I had not lived through the previous tool migration and data migration.

05:53 TW: But so you were having to be both the champion for whatever the new tool was, the selector of what the new tool was, and the manager of the transition and fear uncertainty and doubt of the people who were moving?

06:08 NK: Yeah. And the adopter of the new tool and the roller out of the new tool. Yeah, a little bit of everything.


06:18 TW: How did that go from a… And I managed the transition from SPSS NetGenesis. I managed the selection and roll out of Webtrends, so switching from a log file base to a tag based, and I remember being a little bit of, I needed to get a running start and understand how the tool worked with Webtrends for anybody like… I know, Michael, you were a heavy Webtrends user. The whole aggregated tables. We’d already implemented the tool before I was fully understanding how the analysis tables and the report tables worked, but I felt like I needed to stay two to four steps ahead of the business with their requests and it was really uncomfortable, because they… Somebody had seen something in a demo or somebody just assumed, because we were getting this new tool, that anything they wanted to know, they would be able to answer, and I wasn’t super, super comfortable with the tool yet. Did you have that sort of experience or were you able to ramp up ahead of things in a way that you felt like you were pretty confident?

07:25 NK: I was new enough to web analytics overall that I was definitely in the naïve state of… I think the first tool you learn, you think that’s the way every tool probably works, or at least that was my naïve belief when I first started out. So when I started looking at some of the other tools or options, I remember even asking Omniture, “Hey, if we go with you guys and you guys acquired WebSideStory, are you gonna be able to port in our WebSideStory data into SiteCatalyst?” Honest, naïve question. I was like, “Are you guys gonna be able to offer that bridge?” And they were like, “Oh, woah, no.” And now…

08:09 TW: I thought they would’ve said, “Oh, absolutely. Have you signed yet? We’ll give you the real answer afterwards.” Sure.

08:15 NK: Now, I get why that’s not really feasible, but I asked that question because I didn’t even know. I was just hoping that they might say, “Sure, yeah, we’re gonna figure out a way to bridge these two systems and so we’ll just port everything over and we’ll carry this historical view of your data for you.” It wasn’t really until we started to do some of those vendor evaluations, and at the time, there were more vendors to speak with. And the more we talked through it with salespeople, it became clear no one was going to promise that there was going to be a way to save our WebSideStory data on our behalf and that it was gonna be more of a clean cut over.

09:05 TW: Webtrends, that was like… I must have heard that a half dozens times. “You know, it’s all about the trends. That’s why it’s in our name.” ‘Cause we did an exhaustive… We were trying to, actually, match and we set up Webtrends so we could do it. We were actually trying to, at a hit level, match traffic. ‘Cause we had that concern, and it’s probably different concerns now. When I was going through this, it was 12, probably 12 years ago. And it was this whole page tag based thing, is it capturing the complete data, and we had this log file thing which was very tangible, and we knew there were challenges with that, but we said, “Well, we have a record of what’s being captured. Let’s actually compare at an IP address user agent string level,” and, in hindsight, that was a lot of work that was never gonna fully close the gap. Now, personally, professionally, it was enormously helpful, because I got sys admins drawing diagrams of how the internet worked, and even they were sometimes saying, “Yeah, I don’t know why we could be missing this data, but, hey, the internet is messy.” But I don’t know. Did you wind up trying to do the reconciliation in that transition? And Michael, I don’t know if you’ve gone through that as well anywhere.

10:33 MH: I’m never willing to admit having done any kind of data reconciliation between analytics tools.


10:42 NK: That’s a whole ‘nother podcast.

10:44 MH: Yeah. As far as anyone listening to this show is… Be concerned is that is not possible and you should never attempt it.

10:51 TW: I had a client that went from Adobe to POI… Directly to Adobe to POI through Tealium and we still had to do a reconciliation.

10:58 MH: No. The real answer, though, is absolutely. In fact, back in the day, I think it was much more common to go through and do these painfully fine-toothed like, “Okay, sessionization cuts off at midnight on this one, so any sessions in this tool that go past midnight become a session on the next day. And that’s how we drew up the sessions from here to here.” And like, “Oh, and this one’s not set to GMT over here.” So even all these different little ways of seeing how the sessions and things like that break down, you start trying to figure out the differences between the tools. And that’s one of the traps, if you will, of tool transitions, is getting too fine on the transition, in my view.

11:49 NK: Yeah. I think we ended up going with Omniture, and so, at the time, the nice thing about transitioning from WebSideStory to Omniture was that we were able to keep both tools running concurrently within our contract. Whereas, if we had gone with Coremetrics, it would’ve been start a Coremetrics contract, end your WebSideStory. We were able to do a sliding scale implementation and I remember, on the first site, we actually had them both running for a little period of time.

12:23 TW: Like two weeks? Couple of months? What’s the… Do you remember?

12:27 NK: Nah, like six months… It could’ve been like six months.

12:29 TW: Six months. Okay.

12:30 NK: I’m not remembering how close or how far off they were. I wanna say they were probably pretty close because I’m not remembering any type of horrific experience with trying to answer questions around the differences. But then when it came to rolling out SiteCatalyst on the rest of our sites, I was fortunate that we paired it with an entire platform relaunch of those sites, and so I was able to point to a very natural apples to oranges comparison of, “Well, we had WebSideStory on the old site, we had SiteCatalyst on the new site but remember we redesigned, we relaunched the whole platform.” I didn’t have anybody totally married to last month’s numbers when we switched because there was such a large switch, so that worked out in my favor.

13:22 NK: Also, prior to me coming on board, they had always had issues with WebSideStory was up, WebSideStory was down because they didn’t have anybody in an analytics function paying attention to the data. We had really spotty historical data when we were trying to do year over year comparison. I lucked out that I came on and I got to smooth out the data and babysit it better, and then all of our historic comparisons, everyone didn’t have a lot of faith in them because nobody was really paying attention. That worked out in my favor.


13:56 TW: I’ve done the comparisons a few times. I had a client that was going from Urchin to Google Analytics and I think that trying to run things in parallel long enough and focus on those higher level, the visits, the page views and say, “You know what? Let’s figure out if we have a constant delta and what that is, and that means we can even restate historical if we need to have those high level.” But once you start getting into channel detail or specific pages, that’s not gonna be feasible. I’ve done that a couple of times where it’s like we have to manage this transition. In some cases, it’s gonna be a year before we can say it is apples to apples. Of course, I love it if there’s been… I’ve had a client that said, “Oh, hell no. We’re going from Coremetrics to Adobe and our Coremetrics data is utter crap.” Not Coremetrics. That’s the implementation, not the tool. “So, no. We think it’s a waste of time to try to compare.” And it’s like, “Oh, thank God.” There are lots of ways to work with that.

15:06 NK: Yeah.

15:07 MH: Yeah, and going back to that, that’s the thing is a lot of times the reason people are switching is because it’s completely broken and there’s a new team in place and they’re bringing in a new set of tools. That does come with some risks too. Have you guys seen where maybe people are over-preferring the tools they know? So they walk into a new company and they’re gonna get analytics up and running like, “Well, the first thing we need to do regardless of anything else is get the tool I’m the most comfortable with installed.” ‘Cause I definitely seen that kind of activity.


15:45 TW: Oh, Nancy, have you run into that at all? Anywhere? Ever?

15:48 NK: Yeah, I would say I think that’s definitely… To me, that goes back to that, it’s a little bit of that natural inclination. I think analysts wanna feel proficient and competent in their work and so I think there’s a natural, you wanna stick with a tool you know because you’ve maybe spent a couple of years learning it, you finally know how all of the ins and outs of the data collection and reporting work. And so I think it’s natural to want to be able to apply that in a new job or a new role. And I have worked with teams where they hire someone who’s coming from a different tool.

16:26 NK: They need to come on board and learn the other big web analytics tool and it’s natural to ask a lot of questions and poke holes and ask, “Well, can I do this? ‘Cause I used to do that thing at my old job.” And it’s a big challenge, I think, for people to get over. In the case that I’m working with now, the analysts, they don’t have a choice. Their company has already made the decision on the tool. Most of them come in with having experience in the other big web analytics tool. They’ve now got to wrap their brain around, “Okay, what does this new tool do? Now, I’m completely unfamiliar with some of the terminology or how it works.” I think it is… It’s frustrating for them.

17:07 TW: And yet, we love to say… And I think we went around on this on a earlier, earlier episode. We love to say, “This specific tool doesn’t matter. When we’re going to hire, it doesn’t matter as long as they have some experience, or they have the curiosity or if they have some familiarity with web analytics, I can teach a tool.” Which, in the abstract, I completely agree with, but in the reality, having gone through that multiple times… And I’ve seen it happen with both analysts who come in… I’ve got one example where somebody really knew Omniture inside and out and it was like a year in, he was still fighting. Most of our clients were on Google Analytics and it got to where I’m like, “It is exhausting that every time we head down the path on a new tagging project I’m gonna have to listen to you tell me how you would do this in Adobe for 15 minutes. And I know I have to listen to it and it’s starting to annoy me ’cause you just haven’t learned how to do that translation.”

18:13 TW: And even with business users there have been… I’ve had cases where a business user comes in and says, “Well, gee, I could always get X in Adobe or I could always get Y in Google. I expect that exact same thing.” And that is like the new employee who’s doing a little bit, I think, of strutting their stuff, where you were heading, Nancy, I think, to say, you want to hit the ground running and make a good impression and if you can say, “Well, I’m hampered because you’re using a different tool and especially if I can convince you it’s an inferior tool, then now I’ve bought myself some time.” But I think you made the point that once you’ve done that, made that transition once, then all of a sudden, to go from tool number two to tool number three, now you’ve become a little more fluent in, what’s the sort of stuff that stays the same? What’s the sort of stuff that changes?

19:09 NK: Yeah, and I would say too, even with other people on Team Demystified, you always hear a little bit of some grumping when someone’s trying to figure out, “Can I do this thing that… How does Google handle this thing that I used to be able to do in Adobe?” Or vice versa, because you…

19:26 TW: How do I do merchandising? How do I do merchandising in Google Analytics?


19:30 NK: ‘Cause there’ve been a lot of people who have joined Team Demystified and a wealth of experience, but when you are doing that crossover… I hear very seasoned expert analysts saying, “Yeah, I miss that I knew how to do that and now I have to figure out how to try to… Whether I can get that same data or how that works.” Or even just being a little bit grumpy about some differences in the tools themselves, like people saying they wished Adobe supported Reg X, or they wish Google supported something like… I can’t think of a good example right now, but there’s always people… I think there are enough distinct differences that you will hear people grump about them here and there.

20:20 TW: I think we have to say, to be fair, we gotta get at Google, I wish Google had hit-based segments. There you go.

20:25 NK: There you go.

20:26 TW: Reg X in Adobe, hit-based segments in Google would be fantastic.

20:31 MH: I’ve also noticed, a lot of times, people wanna go after the tool they’re most comfortable with without necessarily considering all of the aspects of the tool that need to be in place for their organization. What are the administrative capabilities? What is the reporting interfaces for all of the people who will be the end users? And all those things. So if you’re gonna switch, switch with a holistic mindset or point of view. And something else I’ve observed, especially in the fast growth start up and in that side of the world, is a lot of companies, because they have such deep engineering talent, will even go from, “Hey, we got this analytics tool, but actually we wanna go build our own tool. We’re gonna build our own proprietary set of tools for this and build out our own analytics tool.” I’ve seen that switch happen a number of times as well.

21:24 TW: On the part of the not considering everything, that did remind me that I’ve realized how often there will somebody who says… They might’ve been working in a team that had an amazing, phenomenal, robust, mature Adobe analytics implementation with a kickass implementation and management team, and a fantastic group of analysts who knew Adobe inside and out, and so their perception was all of that was Adobe. And then they wanna come and say, “Adobe had all this stuff and we’re gonna switch to Adobe.” And they don’t actually have the infrastructure, the people, the knowledge, the internal. And I know… I’ve gone through cases where somebody says, “I really know Adobe and it’s this.” And it was like, “Oh, okay. Great, I’m kinda fluent in Adobe and Google so let me explain event tracking to you in Adobe terminology, or let me talk about custom dimensions, how that’s really just like an eVar, although you don’t have as much allocation and expiration control.”

22:26 TW: And then I realized that, “Oh no. You know what? They didn’t.” They were getting data out of Adobe, but there was somebody else who was way more knowledgeable about the tool who was actually managing it. That’s become my little… When somebody says, “Oh, I know this tool really really well.” Like, “Eh, really? What’s my way to check?” Because the fact is, if that was the tool of choice at your previous company, you, the business user, or you, the incoming analyst, then that’s great, but I’ve made the wrong assumption multiple times. That when they said, “I really knew this tool well.” That they actually knew that tool really well, and they didn’t. And then it’s led to where I’m trying to help and I’m not actually helping because they really didn’t know that other tool that well. Sorry, that just took me down the other cynical rant.

23:18 MH: But yeah, when you’re the analyst and you’re sitting there using those tools everyday, and you make that transition, it’s a big hill to climb. And we certainly talked about a lot of the negative steps, but have you seen positive things come from that? I know… I migrated from a Webtrends to an Adobe world, so I went… It was mostly positive, frankly, ’cause you get your hands on some better capabilities and it’s good. After you’ve died on the Webtrends mountain a little bit it’s like now you can ski down the other side.

23:58 TW: It’s an opportunity to reset and say, “Are we clear on… We’d Frankenstein stuff together in the old tool as requirements of all of this is our blank slate. So hey, it’s an opportunity to do a baseline requirements gathering. Hey, let’s figure out what is the truly key data, and that can all be couched, and we’re implementing this new tool. We have to make sure… We’re not necessarily gonna get everything perfectly implemented right out of the gate, and we’re not necessarily gonna get, exactly one-to-one, a mapping of every other report you had, so let’s look at the data you’re getting now that you can’t live without. Let’s talk about what you’ve never been able to get out of the old tool that you have always wished you could and been told you can’t.” And sometimes that’s a, “Well, yeah, ’cause the internet doesn’t work that way.” But sometimes it’s the “Oh, we can plan for that, like we can implement for that.” So I think there can be those upsides, being careful to not over-promise that, “Oh, everything you’re gonna want, you’re now gonna get, ’cause we have this fresh new tool.” Yeah, I definitely have seen that.

25:06 NK: Yeah, and I’d say from a positive… One of the reasons I think I’m well-suited for analytics is that I do like learning and I would say the chance to learn a new tool and some new tricks and ways to fully leverage it is interesting. So when I went from Adobe to Google Analytics, it’s like, “Okay.” I spent a few months scratching my head, like, “How does everything work in here? And what can I do, and what can’t I do?” And when you then figure out what GA is good for and how you can use it, and some things that it can do, that Adobe can’t, I found that fun and interesting to be able to speak to, “Well, here’s what… Now I can speak to what Google does well. And I can still speak to what Adobe does well, and they’re two different things.” I found that process interesting, just because it’s a part of our industry, is always evolving, changing, learning.

26:07 MH: Yeah, I just got a Tweet from Ben Gaines, he asked me to just mention that anything that Google can do, Adobe can also do.


26:16 MH: I don’t know where that came from, how that coincided just… He must be in my Alexa or something. I don’t know how that happens. No, I’m just kidding. I think that’s a great point, Nancy. And as analysts, that’s it, is that constant learning and getting in and finding new ways to do that. What I’ve also found is that has helped me solidify and deepen the, “Why am I going down this path?” kind of a thing, because when I change how I do it in the tools, it’s making me think about how the data and the meaning of the data as I pull it out and work with it. So I’ve felt like working across tools has helped me systematize, if you will, the analysis that I’ve done.

27:04 TW: Yeah, and I think it’s rare to be just working with just one tool, so I think that’s the… In that, a lot of times, you’re also gonna have… Maybe you have supplemental tools that you’re also running, Clicktale or Crazy Egg, so it’s web analytics, or you’re doing voice and customer stuff or you’re doing social media analytics, and in my experience, it’s rare for anybody to say, “We’re using the exact same toolset across the spectrum of what the analyst should be doing for the last five years.” So analysts, and I’ve run into this, analysts who are really not excited about saying, “Here’s a new tool that works a little different and differently and maybe is going to give me some supplemental data,” it’s a little scary. So to me, there’s a little bit of… Yeah, like Nancy, when you were switching from Adobe to saying, “Well, I haven’t done a whole lot with Google Analytics, but I’m 80% excited, 20% scared.” That’s what you want to see as an analyst. If you’re an analyst, saying… Actually funny story. That company that I managed them from NetGenesis to Webtrends has since gone through UNICA and is now on Adobe. So they really were on the every two to two and a half year transition cycle, and I do know there’s at least one analyst who’s been there for almost all of that. So she clearly enjoys doing that or, presumably, she would be gone at this point. But I think that’s actually a sign of a good analyst, somebody who’s excited about the possibilities.

28:46 NK: Yeah, I would say that first time that I had the change thrust upon me, it was just more confusing. It was like, “Wait, Omniture isn’t gonna work the same way as WebSideStory?” Again, just naïve on my part, but I thought, “Well, don’t these all work the same?” And I realized a couple weeks ago, I had somebody also who has built his career in Adobe Analytics ping me to say, “Well, is there really that much that’s different with Google Analytics? I assume that the data collection is the same. It’s just like the UI and the way you report, like the reporting interface is different.” I explained some things that are different and I said, “There is a little bit of a mental, mind shift that you do make when you switch between those two.” I just think, once you go through it, and you then recognize, “Oh, these are tools built by completely different groups of engineers, or quite different product roadmaps, once you just get it, that every tool is different, I think it just becomes that much easier to embrace a new tool and say, “Well, I bet it could do some things that I’m not doing today or I bet it can do some things faster or differently that could be an advantage. My mind opened up once I had gone through it the first time. And now, it’s completely normal-sounding to me that if I had a new tool thrust upon me tomorrow, it would be like, “Okay yeah, let’s figure out how this one works.” Not, “I wanna make it work exactly like my old tool.”

30:17 MH: Yeah, that’s great. No, I love this discussion. I think one of the things we always try to do is try to compress down and make a couple things real. So like if somebody was walking into that situation right now, what’s the one thing or one tip that you’d give somebody? And we’ve talked about them so we can maybe distill some of this conversation down, and I don’t know if anyone can go? Who’s got a tip?

30:43 TW: I would say my one thing is because I’ve watched it be such a hindrance to people, literally, in their success, in their role, and that is fight with every fiber in your being your belief that the other tool did something better because you can’t figure out how to do it the same way in the current tool. It’s what Nancy was just speaking to, but the longer you hold on to I need to keep translating into the language of that other platform, one, you’re not gonna really learn the new platform and you’re probably gonna implement some things in a lousy way, and you’ll struggle. So I think just accept that the tools that have decent market share right now are all pretty freaking solid, and the only people who will tell you one or the other is just unabashedly better, it’s the vendors themselves or the people who only know one tool and have struggled with the other ones. That’d be my one, is just accept that there aren’t really good and bad tools. They’re just different and it’s on you to make that transition.

31:54 MH: Nice. Nancy, you got a tip you wanna share?

31:57 NK: Yeah, on that same line, I think it really is just cleansing the palette or clearing your mind when it comes to… Again, I’m thinking more of watching people go from Adobe to Google Analytics, or vice versa. I’ve watched people transition both ways, and I went from Adobe to Google Analytics, and I think there’s always a little bit of clearing your mind and just being very open-minded, to say, “Tell me what this tool does or tell me what the advantages are of this tool.” And letting go of the things that you can’t necessarily do in the old tool, or I have to move away from. And it piggybacks on Tim’s point that the longer you stay bogged down and negative about the fact that you have to make a switch and you can’t do the old thing, you’re really only just bogging yourself down and taking time away from becoming productive in a new tool.

32:54 MH: I love it, and if I had to share a tip, I would share two tips, because they’re both important. One is…


33:00 MH: Think about your transition holistically. If you’re an analyst, you’re using the tool one way, push for a deeper understanding of the governance, the administration, the user base, all of the different things that go into a tool because that will save you transition pain and suffering to address those issues later, and then I would also say, and I think, Nancy, you were the one that talked about this, about running the old tool with the new tool for a period of time to give yourself across over a timeframe. And I think that’s a really good tip as well. Anyway, one of the things we like to do on the show is do a last call. It’s where we go around the horn and just share anything that’s been of note or interesting in the last little bit. Who’s got a last call they wanna share?

33:47 TW: Go, Nancy.

33:48 NK: I am a member of the DAA, and specifically, I’ve been participating in the Women in Analytics Group, and if you’re not familiar with it, I just wanted to give a shout out to them, because we’re gaining a lot of interest, and momentum, and some great discussions. If anyone’s interested in joining, please head to the DAA website. You can find out information for joining, and volunteering, and helping out.

34:13 TW: Awesome. Michael, you wanna do yours next?

34:16 MH: Yeah, I’ll do one.

34:16 TW: Hate to make you wait to be last all the time.

34:18 MH: Well, thanks, Tim. So nice of you. So, yeah, actually my last call is not analytics-related, but a few weeks ago I saw somebody Tweet this and I just really enjoyed reading it, which was Jeff Bezos’ letter to shareholders for this year, and he covered a lot of topics, but he was describing Amazon’s pursuit of excellence in this concept of a day one company versus a day two company, and around being customer-obsessed, being skeptical of proxies that get in the way of understanding customers. The adoption of external trends and how you need to be really aggressive at that, and then also how to make decisions more quickly, and I thought all of those things were just super key to anybody who’s operating a business in any context. And certainly, as analysts, I think all of those things are really helpful to think through. I enjoyed reading it. Just thought it was something really cool to internalize. I have a Bezos crush.

35:25 TW: Was it part of their… Like just an earnings thing? Or was it just random, or was it…

35:29 MH: I think it’s probably part of their filings each year. There’s an annual report or some kind. I don’t know. I used to be in that world of high finance, but not anymore.


35:44 MH: What’s your last call, Tim?

35:45 TW: I think you picked up a Georgian accent.

35:48 MH: Yeah, I declare.


35:51 TW: I am gonna go with a little bit of a shout out, or recommended, I think it’s a 39-minute video on reinforcement learning that was friend of the show, past guest, Matt Gershoff of Conductrics.

36:10 MH: Frequent mentionee.

36:12 TW: Frequent, most mentioned individual, on the Digital Analytics Power Hour. But Michael and I both got to see Matt stress for multiple days in Hungary earlier this year about how his presentation just wasn’t ready and he had to go work on it and create it. And then he proceeded to, in his delightful and very clear and coherent way, talk about reinforcement learning and how it changed his perspective on everything he did and what it was. He presented shortly after you did, I think, Michael, so there was actually a credit to you early in his session but the video is posted. It’s Reinforcement Learning: Sequential Decisions as a Unified View to Conversion Optimization, and it’s well worth a watch.

37:01 MH: Hear, hear. That was a very good talk. Well, if you’ve been listening and you’re thinking to yourself, “Hey, they didn’t talk about what I experienced in this tool transition,” we would love to hear from you. The best way is through the Measure Slack. We also have a Facebook page. We’ve got a Twitter account. We’ve got a website. We’d love to chat. We’d love to hear your feedback about the show and aspects of the topic that we’ve missed or that you really agreed with. In my case, if you agree with me, I definitely would love to hear that.


37:36 MH: No, I’m just kidding. No, but it’s really great, and being part of the analytics community, we thrive and gain a lot from interacting with our neighbors. Anyway, we’d love to hear from you. Nancy, it was so great to have you on the show. Thank you so much for coming on.

37:56 NK: My pleasure. Thanks for having me.

37:58 MH: And, obviously, for my co-host, Tim Wilson, no matter what tool you’re on, keep analyzing.

38:09 Announcer: Thanks for listening. And, don’t forget to join the conversation on Facebook, Twitter, or Measure Slack Group. We welcome your comments and questions. Visit us on the web at analyticshour.io, facebook.com/analyticshour or @analyticshour on Twitter.

38:28 Charles Barkley: Smart guys want to fit in, so they have made up a term called analytics. Analytics don’t work.

38:37 NK: Spending hours on the phone with a client trying to explain a solution that Tim devised right before he rolled off.

38:46 TW: Not right before.


38:48 TW: It was like last June. Free tip, you get a little system diagnostic along with your podcast appearance.


39:02 NK: I put my son on a sheep on Saturday, Michael. That’s the…

39:07 TW: You said what?

39:09 MH: Well, it’s a timely and topical… What’s a T word for discussion? Talk? A timely, topical talk about the trials and tribulations.

39:17 NK: The trials and tribulations…

39:17 TW: The trials and tribulations of tool transitions.

39:21 MH: Yes, exactly.

39:22 NK: Say that three times fast now.

39:23 MH: We’re gonna get there. Keep working on it. This is just a thinly-veiled career counseling session, Nancy.


39:31 MH: Yeah.

39:34 NK: Tim, can you hear Michael?

39:36 TW: No.

39:37 NK: Michael, can you hear Tim?

39:39 MH: No.

39:40 NK: Okay, I can hear you both. You guys can’t hear each other.

39:45 TW: Son of a bitch, I can’t hear Michael. Can you hear me?

39:49 NK: Is he talking?

39:49 TW: Yeah.

39:50 NK: I can’t hear… If Michael…

39:52 MH: Oh. Now, you can hear me. Sorry. I was…


39:55 TW: Now, you’re just fucking with us!

39:58 MH: Yeah. I was on mute. I thought I had unmuted.

40:01 NK: I thought he was just sitting there.

40:01 MH: Well, most of the time that’s what I do, just sit here, just collect the royalty checks. One of the things we try to do on the show is have my dog interrupt while I’m trying to do the last call.

40:17 TW: Rock, flag, and tool transitions.


2 Responses

  1. Victor says:

    I just went through a reconciliation but with Adobe Analytics and Looker/Redshift. The business transitioned to Looker/Redshift from Adobe Analytics and reconciling the numbers was challenging, namely because Analytics has segments and evars which aren’t available in Looker/Redshift. Eventually we got to a point where the business was satisfied with the differences but it took nearly a year.

    Based on the above experience and from evaluating analytics platforms, my tip for transitioning would be to start with the business reporting. What are the critical business questions that need to be answered? Then work backwards to figure how to get the answers in the new tool.

Leave a Reply

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

Have an Idea for an Upcoming Episode?

Recent Episodes

#243: Being Data-Driven: a Statistical Process Control Perspective with Cedric Chin

#243: Being Data-Driven: a Statistical Process Control Perspective with Cedric Chin

https://media.blubrry.com/the_digital_analytics_power/traffic.libsyn.com/analyticshour/APH_-_Episode_243_-_Being_Data-Driven__a_Statistical_Process_Control_Perspective_with_Cedric_Chin.mp3Podcast: Download | EmbedSubscribe: RSSTweetShareShareEmail0 Shares