#235: 2023 Year in Review with Josh Crowhurst

For those who celebrate or acknowledge it, Christmas is now in the rearview mirror. Father Time has a beard that reaches down to his toes, and he’s ready to hand over the clock to an absolutely adorable little Baby Time when 2024 rolls in. That means it’s time for our annual set of reflections on the analytics and data science industry. Somehow, the authoring of this description of the show was completely unaided by an LLM, although the show did include quite a bit of discussion around generative AI. It also included the announcement of a local LLM based on all of our podcast episodes to date (updated with each new episode going forward!), which you can try out here (or by clicking the little icon in the bottom right of this page, perhaps?)! The discussion was wide-ranging beyond AI: Google Analytics 4, Marketing Mix Modelling (MMM), the technical/engineering side of analytics versus the softer skills of creative analytical thought and engaging with stakeholders, and more, as well as a look ahead to 2024!

Links to Items Mentioned in the Show

The complete ChatGPT response to Tim’s prompt—inspired by Julie’s final thought: “Rewrite the serenity prayer to specifically apply to analysts:”

Grant me the serenity to accept the data I cannot change,
Courage to interpret the insights I can,
And the wisdom to distinguish between statistical noise and meaningful patterns.

Living one dataset at a time,
Enjoying each correlation, accepting each limitation,
Taking the time to understand the story hidden within the numbers,
Trusting that with diligence and analysis, the truth will emerge.

Embracing the challenges of uncertainty and ambiguity,
Seeking to uncover the unknown with every regression and model,
Granting myself the patience to iterate and refine,
And the humility to acknowledge that even the best analysis is an ongoing journey.

May I be an instrument of insightful discovery,
Where there are anomalies, let me investigate,
Where there is confusion, let me bring clarity,
And where there is potential, let me find opportunity.

In the ever-evolving landscape of data and trends,
May I navigate with skill and precision,
Always learning, always adapting,
And, above all, may I contribute to a world better understood through analysis.

 

Photo by Greta Farnedi on Unsplash

Episode Transcript

0:00:05.8 Announcer: Welcome to the Analytics Power Hour. Analytics topics covered conversationally and sometimes with explicit language.

0:00:14.3 Michael Helbling: Hi everyone. It’s the Analytics Power Hour, and this is episode 235. Hey, analytics Power Hour listeners. It is that time, we couldn’t be more thankful for you listening to the show this year and being awesome at analytics, maybe as a result, or probably already awesome. We probably had very little to do with it, but what a year it has been. We have fully transitioned the show to an AI, so this is not even a human talking at this point. We also added two new co-hosts this year, which has been a huge success. But yeah, this is it. The year in review show when we brought everybody. So let’s introduce everyone. We have Josh Crowhurst. Welcome, Josh.

0:00:57.5 Josh Crowhurst: Hey, good to be here.

0:01:00.6 MH: It’s nice to have you on the show for the Year in Review still, we also have Val Kroll. Welcome, Val.

0:01:08.4 Val Kroll: Thank you very much. Very ready to be reviewing.

0:01:11.2 MH: Excellent. Tim Wilson, as always.

0:01:17.7 Tim Wilson: Hey ya. And it’s good to see Josh, Josh was, that didn’t announce that, didn’t, didn’t reflect on that, that Josh was on some shows earlier in the year, not just making his annual appearance, which was nice.

0:01:27.3 MH: That’s true. And Julie Hoyer, welcome.

0:01:31.7 Julie Hoyer: Hey everyone. Hi.

0:01:33.1 MH: And Moe Kiss.

0:01:36.2 Moe Kiss: Hey, how’s it going?

0:01:37.7 MH: Hey, Moe, how you going?

0:01:39.8 MK: Good.

[laughter]

0:01:40.7 TW: It’s like you just ran in from out of the room coming in hot.

0:01:43.4 MH: And I am Michael. Helbling. Well, we have our plates full and the time will fly by, so let’s dig right into it. This year we wanted to look back on 2023 at sort of major themes that shaped our industry. And in quintessential fashion, Tim put together a handy dandy r-apps to rank our voting of topics. So while you know, we won’t actually stick to that order, at least he tried. And what do you think came up as number one? I think I will let somebody else. You say.

0:02:15.7 TW: UTM parameter structure.

0:02:19.5 MH: UTM. That’s right. [laughter],

0:02:22.0 VK: Bounce rate.

0:02:23.5 MH: Bounce rate is back [laughter]

0:02:26.8 JC: What’s a good bounce rate?

0:02:28.8 S1: That’s right. What’s a good bounce rate? We’re gonna spend the hour talking about what’s a good bounce rate. Oh, wow. I’m not gonna say anymore about that, but no, I think it’s no surprise that probably the most voted item for 2023 was generative AI, AGI, whatever you wanna call it. And certainly the last couple of months gave us no shortage of drama in that space in terms of open AI, which we could talk about if we want to, but yeah, why did we vote that number one, I have no idea. Maybe the AI voted it number one. Oh my gosh, we don’t even wanna talk about this, but the computer’s making us. [laughter]

0:03:07.7 TW: I mean, it’s crazy how much that is… Like, that is the zeitgeist right now, right? And, it feels like it’s, there’s the really cool stuff that’s happening with it, and then there’s the everyone and their cousin jumping on the train and defining things that… Yeah. What you’re doing is, machine learning and you’re referring to that as AI, or what you’re doing is statistics and you’re referring to it as AI. But…

0:03:40.8 MK: I’m not sure. Yeah, I probably like saw a different thing, which was that, I guess the reason I think of it as being so popular is I actually genuinely can’t think of like the last time in my life where something is so dramatically changing my ways of working even individually. And then obviously like working at a tech company, there’s the effect on our own product and our own ways of working. But like me personally, like it’s actually changing how I do my job or like how I, spitball team name ideas or other stupid shit, or come up with like lyrics to a song for my baby. [laughter] Just all these fun use cases that are just like, I don’t know, it’s like actually, like you can notice your life change because of it. And I just don’t remember the last time that happened with a technology.

0:04:31.7 MH: I mean the internet kinda did that for [laughter], some of us.

0:04:35.0 MK: The last time I remember it.

[laughter]

0:04:37.1 MH: Oh, that you remember it?

0:04:38.5 MH: Yeah. [laughter] Ah, back in my day, we didn’t have no internet.

0:04:46.3 TW: Can I ask, like, it’s funny, like the AI stuff has come on and AI’s obviously driven by data or data if you’re in Australia and that it spent this period where I feel like there’s times where it’s like, oh, this, this is analytics. Which to me, the more I’ve thought about it, like it’s, it is a very cool and powerful and useful thing and it is a useful tool for an analyst in many, many different ways. But it’s been interesting to me that it’s like everybody’s looking for the easy button. And I think we’re still in a mode of where there are analysts and there are stakeholders saying, oh, can I just ask the AI what my KPIs should be? Can I just ask the API to find me an insight? Like things that I think it’s uniquely not equipped to do well.

0:05:42.5 TW: And it’s just, it’s kind of disheartening to see some aspects of the industry think, oh, now we’ve solved it. Like we don’t have to have the curiosity or the technical chaps or the stakeholder relationship management or the thoughtful creative thinking going on. ‘Cause we have the AI and it’s like, actually no. Like there’s a lot of really cool stuff with it. But then I look at the core of what I see analysts doing, and I’m like, no, the AI’s augmenting aspects of that, but it’s just not replacing what I think is the kinda most, the hardest and most interesting aspects of analytics.

0:06:27.5 JC: Yeah. I think like from my side, I’m seeing this like, at least in my organization, there’s a lot of excitement around, okay, we’re gonna do self-serve analytics with AI. Like, you’re just gonna put your query in, you’re gonna type what you’re looking for, the AI is gonna interpret that, and then it’s go, it’s gonna go and it’s gonna actually, do the underlying analysis and come up with the insight. Right. And that makes me nervous… [laughter] And it also makes me think a little bit about, yeah, yeah. I mean, I’m getting these like, like, certain people in my organization are saying, oh, this is like, this is, this should be our big focus. We should be making this something that we’re building for our stakeholders in like in the next year. And I’m like, hang on. I don’t think we’re there yet. And also, even if that were possible, I think that it’s still a bit dangerous. Yeah. It makes me nervous. [laughter]

0:07:25.0 JH: It’s interesting you say that exact or like example because I was just catching up on some newsletters and one from Benn Stancil, he was talking about AI and he was actually talking about an example, like people say, oh, just use it to do all of your SQL queries. And he was like, but if you’re an analyst, you know, when someone says a generic question, all of a sudden you’re saying, well, do you need your to date? What’s your time period? What’s the definition of the metric you’re actually asking me to pull? Oh, did that change? Is the ETL broken? Like all these things that an analyst would have to know.

0:07:57.7 TW: Don’t forget to exclude this thing ’cause we always have to exclude this thing.

0:08:00.7 JH: Right? Or like different GL locations, right? And so you’d, even if you were prompted by AI then to say, like, ask you those specific things, those are just like inherent trip ups that analysts run into. And then if those change over time, it was interesting he was saying, so the people that got into AI to do cool advanced stuff just become the keeper of like the black box and making sure the data’s healthy. It’s like this full circle thing. It was really interesting.

0:08:24.0 MK: But so as in like, ’cause what I take away from that, like I’m kind of smooshing together two thoughts from like Tim and then Julie and Josh is like, I still think of it as like AI’s there to do the simple stuff. And the thing that’s actually hardest about it is to do the prompting well to get the right output. And that’s why I, it actually kind of doesn’t scare me because even as what you’re describing of like, okay, I’m gonna use AI to write my SQL query, it’s like I still need to do the same thinking. The difference is maybe I don’t need to have the syntax perfect, or maybe I can just Write in plain text, exclude this bucket of users because it’s not accurate. Or internal users, for example. But like, you still have to do the thinking and you still need to give your thinking to the machine.

0:09:13.1 TW: Yeah.

0:09:13.2 JC: Yeah. 100%.

0:09:13.2 TW: But Josh weren’t you saying it going to the, if the business asks, the business asks the AI, that’s the terrifying.

0:09:23.7 JC: That’s what makes me nervous. Right?

0:09:23.7 VK: So what you were describing Moe, about how it’s there for the simple tasks, that’s exactly where I am kind of landing on it right now with the current applications. We at further, our technology practice lead ran a gen AI robo mind battle as he called it, where we put together these like random teams of people to try to develop solutions that we could bolt onto things that we do at search discovery. And the intention was to make a client facing, but the majority of the teams actually ended up presenting solutions that were increasing operational efficiencies or would turn inward to help assist one part of the process that we felt was like, especially arduous or like not something that was like a good use of the analyst’s time, but it still required exactly what you said, like that thoughtful process upfront. It was really just like an accelerant. And so I thought that that was interesting that everyone’s like, yeah, solutions. And then it was like, yeah, to help the analysts, [laughter] and it was all mostly internal things.

0:10:22.2 TW: Well, and that’s like if you ask the, like the business, how many times have analysts have we had somebody say, I have a quick question, or I have a simple question. And so when we say, oh, the AI can do the simple stuff, well, it’s pretty common for somebody to think it’s a simple question and then it’s the human that responds. And I mean it’s easy to sound like a gatekeeper, but it’s just if one out of 10 of those where they just say it’s a simple question and then they just get the answer and they run and broadcast it, and then it’s out there and you’ve got data distrust.

0:10:57.9 MK: Okay. Just let’s just take a step back, right? Like you can have stupid people now who can look at a graph and completely misinterpret it. So the only difference now is that the person is like potentially asking the question of AI, visualizing it, whatever it is, getting an answer. And like, it is still about fundamentally a person misinterpreting because they’ve asked the wrong thing or they haven’t given the right inputs or whatever it is, right? So I just don’t, as the person who started this conversation by being like, it’s completely changed everything. I’m now like, it hasn’t actually changed anything. It’s still about teaching people how to ask the right question and what those caveats are and how to think through a problem. It’s just using a different tool which might be a bit better or like make things a bit easier.

0:11:44.3 TW: Yeah. I’m just gonna get preachy and I would stop.

0:11:46.9 MK: I feel like Tim is getting angry and angrier.

0:11:51.6 TW: Well, I would stop short as referring to the people being stupid and mis… Like that that is kind of the, we talked a lot about like relationship and skill stuff, right?

0:12:00.0 MK: Yeah. Fair.

0:12:00.5 TW: So just that…

0:12:02.7 MK: Fair.

0:12:02.9 TW: That caveat. I mean, I don’t know. This is, there are a million angles to go on this.

0:12:09.1 MH: Alright, it’s time to step away from the show for a quick word about Piwik Pro. Tim, tell us about it.

0:12:14.0 TW: Well, Piwik Pro is easy to implement, easy to use and reminiscent of Google’s universal analytics in a lot of ways.

0:12:20.4 MH: I love that it’s got basic data views for less technical users, but it keeps advanced features like segmentation, custom reporting and calculated metrics for power users.

0:12:30.1 TW: We’re running Piwik Pro’s free plan on the podcast website, but they also have a paid plan that adds scale and some additional features.

0:12:35.0 MH: That’s right. So head over to piwik.pro and check them out for yourself. Get started with their free plan. That’s piwik.pro. Alright, let’s get back to the show.

0:12:46.3 MH: I think the key thing is that we’re generative, right? When I think about analytics and AI, I always remember that it’s a generative AI, which means it’s built to make stuff up. And so depending on how you ask, it may give you things you do or you don’t want and the way you ask. And like, I think Moe you talked about this like prompt engineering, kind of giving it the guardrails to pursue the answer that you’re looking for. That’s kinda the skillset. I think this isn’t an original idea with me. I think I heard this from Mike Taylor, but he sort of said, I think prompt engineering will be like the Excel skill going into the future.

0:13:25.0 MH: So like if you’re good at pivot tables today, tomorrow you’ll be good at prompt engineering. That’ll sort of be the skill to have in the future. But it, I think that’s, to me, when I think about it, the applications, there’s so many of them, but I find that it’s mostly for me so far been an efficiency, like you said, Val, like, in a kind of an efficiency play so far. I haven’t used it to analyze data yet.

0:13:50.8 TW: But can we just, I mean I think there that we do have a semantic kind of challenge in that AI has been around, people have been doing things that, that count as AI for years. Generative AI.

0:14:00.5 MH: No, I’m pretty sure it was this year. Tim was the first time… [laughter]

0:14:06.5 TW: Well, but I think like.

0:14:08.1 MH: You’re right, you’re right.

0:14:09.1 TW: I keep trying to do say like the generative AI, I think is what’s super, super unique. But then that somehow caused a feedback loop to people that are like, we’re using AI and what they’re doing is not generative AI, it’s fine. Like they’re getting a, they’re getting a boost from it. I think it gets combined and then you have AGI, right? The artificial general intelligence, which I’m a little fuzzier on, right? But that’s actually I think a different world and that’s, it’s always the danger when something catches such fire that people are kinda just throwing around willy-nilly. Like I, if you go to Joe on the street and say, what do you know about AI? There’s a pretty good chance it’s gonna be chat GPT, which is generative AI.

0:14:53.5 TW: But then you’ve got, let me automatically generate your data catalog and update it. I don’t know if that’s generative AI, but I mean, it’s fun. I’m like, Moe back to your point, like the yeah, it is fundamentally, there’s not a week goes by that I’m not trying something out and it’s like, well that’s wild. And I could not have comprehended doing that a year ago. And that’s both with kind of the code stuff as well as the like image and text stuff.

0:15:27.5 MH: Pictures from far away, right on your TV at your house. [laughter]

0:15:33.1 TW: Yeah, [laughter]

0:15:35.6 VK: I tried to use Chat GTP to help write some of the social media posts for the Power Hour and it is terrible trying to be funny. Oh my God. I was like, this is not helpful at all. [laughter],

0:15:51.0 JC: Maybe that’s a non engineering issue.

0:15:54.7 VK: Josh, are you saying that this is a me problem? [laughter]

0:16:00.5 JC: That’s like a bowel problem. But I was just thinking like, oh man, yeah, that was something that I really like. Anytime I’ve tried to do that, I totally sucked at it. I would totally go to chat GPT to try it. And that’s discouraging that it didn’t work for you. [laughter]

0:16:10.4 VK: Made me feel better about what I was putting together. [laughter]

0:16:13.9 TW: But Michael, you’ve used it for some of kind of the re-exing or the re right? I mean you have used it successfully with anthropic and some other.

0:16:25.4 MH: Yeah. But anything I’ve actually used, like for a tweet or something like that, I’ve had to basically rewrite what the AI gives me. And even, even when we did a show about AI this year, Moe immediately remarked how acerbic and dry the intro was and not funny it was. And so like right away a human can tell that there’s this sort of anodyne nature to the way it produces text, which I think is sort of like a little marker that it’s, it does struggle I think with humor. But again, like I think it will improve as it gets better. Maybe that’s why they got rid of Sam Altman for a couple of days, you know, maybe he finally got it to tell jokes or something. I don’t know. [laughter]

0:17:07.8 TW: That’s what the board’s issue really was.

0:17:09.1 MH: Yeah, that’s, it’s like, it’s too smart. I could tell dad jokes now.

0:17:13.8 TW: Is is now a good time to maybe wrap this topic by introducing the local language model little thing that our listeners can play with for the podcast?

0:17:23.7 MH: Yeah. Yeah. Do you wanna talk about it Tim?

0:17:28.0 TW: Sure. So I mean, if you’ve followed LLMs at all and then people talk about having local LLMs and as it happens, we have a local large language model based on all of our past episodes. So this was a company called Astral, try astral.com, that they kind of are geared towards taking things, organizations, people that have a decent corpus, a decent set of content and saying what happens if you build an LLM off of that? So if you go to, you can go to the show notes or probably on our homepage at some point, but the URL of bitly/aph-astral, A-S-T-R-A-L, you basically get a chatbot as well as some kind of prompts and a feed. It is a rapidly developing tool, but we would kind of love for you to play around with it.

0:18:22.6 TW: Astral would love for you to play around with it. Maybe the one request is, it’d be awesome if you thumbs up or thumbs down the responses you get. Like, we have as a group, have gotten some amazing answers and have gotten some absolutely terrible answers. And we’ve asked it some stupid questions and gotten some silly answers. So it’s kind of fun. And we’ve been kind of chatting with their, actually Michael and I just earlier today, were talking with their head of product about sort of where they’re seeing it go.

0:18:56.0 TW: But, it has kind of like been a nice short cut for us in that we didn’t have to actually do any of the technical work and we are definitely finding some things that are weird but it’s also been the idea of if we have all of this content with the guests we’ve had, the hosts we’ve had shifting over the years it’s kind of wild, if you ask a question like What are the key skills for an analyst and get the response from it and know that that’s coming just from the podcast it’s a little, I guess…

0:19:27.2 MK: Terrifying?

0:19:31.1 TW: Terrifying and this is their idea is that for people or organizations that are generating a lot of content, they can actually have for their audience a way to kind of ask them questions ’cause I certainly don’t remember everything we’ve talked about. So if you wanna try it out play around with it, it’s out there it requires you can with two clicks with your Google account, you can create an account but it’s new. Even from between now and when this show releases there will be enhancements to the platform but…

0:20:09.5 MH: I will have a link on the show notes page on the website so, yeah we’d love for you to try it out and ask it questions. Send us the biggest screw ups it sends you or whatever. I don’t know.

0:20:17.3 TW: Yeah feel free to tweet and tag us or throw it on LinkedIn if you find something funny or terrible or great.

0:20:29.5 MH: Yeah speaking of… Oh wait no I won’t start it that way.

0:20:30.7 TW: Like you’re terrible?

0:20:31.9 MH: It’s interesting.

0:20:36.8 TW: Yeah.

0:20:44.0 MH: Oh, what I didn’t say that. Another big thing that happened for a lot of analytics users out there was a big change to Google Analytics, so the sunsetting of Universal Analytics to GA4 and so this year was the year that happened and not qualifying that with a good or a bad, although, it’s been an interesting journey I know there’s lots of people who don’t use Google Analytics or don’t use it as their primary analytics tool so it’s not really about that, but it’s something I think affects a pretty broad swath of the industry and our audience. So I thought it was an interesting topic for us to maybe chat about briefly. I don’t know what do we think… How to go… If you were given a little post-mortem to Google on the launch of their new too?

0:21:28.8 TW: Are there other examples of where a platform has just fundamentally changed its paradigm, like if Power BI said…

0:21:42.1 MH: That’s a good question.

0:21:42.4 TW: We’re gonna move away from… I remember software programs when they moved from a DOS interface to a Windows interface… Sorry I’m just gonna… I’m gonna kind of keep going with the age things that was right around when the…

0:21:56.1 MK: Yeah I was like no do you remember that one?

0:21:58.0 TW: Do you remember that one though?

0:22:02.1 MH: When you went from MS DOS to…

0:22:03.4 MK: Oh shit…

0:22:03.7 MH: Windows 3.1 It was a big change for all of us…

0:22:04.5 TW: OS/2?

0:22:04.9 MH: OS/2 now you’re going way…

0:22:07.3 TW: What I mean is that it was an interesting… It’s Google, so Google can do whatever Google wants and that’s been kind of the complaint people were kind of stuck with it, but they fundamentally said we’re gonna change the data model, we’re gonna change the data collection, we’re gonna change the interface to access the data or they kind of changed, we’re gonna change the fundamental metrics that people have gotten used to and we’re just gonna…

0:22:38.2 MK: I don’t even think Google have ever done anything like this. I know they could do it and they did do it, but I don’t think there are a lot of examples of Google doing stuff like this. They tend to keep stuff like more similar.

0:22:50.8 TW: We’re gonna make the main search page be wrapped with banners, we’re gonna be inspired by baidu.com and make yeah…

0:23:01.3 MH: I remember when Universal Analytics first came out and people were talking about it. I remember being at a Google event and someone from Google saying, “Yeah we still get urchin.js calls from way back when GA was Urchin,” and so they still accepted them, so yeah this is definitely a big departure and it’s fascinating, I try to be somewhat sanguine about changes like that and even keel. But I think for me the biggest challenge was that we did a fair amount of work this year on that platform and notable to me was how buggy it was and how much time was spent trying to figure out is something wrong or is this just a weird data bug problem that we’re having at this, on this particular moment or day and it just took a lot of cycles and I feel like it really affected a lot of people… When you’re doing your job or trying to do your job well it stinks when the product you’re using isn’t holding up to your standards, that’s probably the way I’d like to say it.

0:24:04.1 MK: Do you think though potentially it’s also something that just in kind of a shitty way would have really affected smaller businesses so much more because bigger businesses they can afford to throw a bunch of money. They can afford to have a one-year migration plan, all that sort of stuff, but for little businesses this would have been so hard.

0:24:25.6 JC: Yeah and I think the other part of that is just businesses that are using it more from a business user perspective and not having a full-time analytics team ’cause people are used to those easy-to-use reports in the interface and the new version is much more technical so if you don’t have a full-time analytics team, that’s gonna suck.

0:24:48.6 TW: I think what made me so disheartening is because it basically got lots of people… It was kind of like GDPR in that it kind of gave free license to so much of our industry to focus on the tool and the migration and how are we gonna backfill what you had and replace it and conversations with business partners being… Well let me explain why there’s no bounce rate but now it’s an engagement rate and to me the big risk is because they’re such a big player in the market, all the stuff we were just talking about with generative AI and analysts needing to have conversations with the business and now the analyst has to have conversations about why this metric is no longer available or this metric has changed or why we can’t do anything ’cause we’re busy updating this technology platform.

0:25:48.5 TW: And it’s like that to me that feels like it’s kind of insidiously just pushed us back. I mean still the Measure Slack has so much discussion going on around how do I wire this stuff up to make it work and do what I’m expecting, it’s so much like debugging of the platform as opposed to elevating the conversation and the questions.

0:26:24.3 MH: Yeah.

0:26:24.4 JH: Yeah, it feels like it was a big step back for a lot of organizations for their trust in their data which just takes us back to your point Tim, like to square one. For analysts or data practitioners, they take all this time to implement the new thing and now they’re having to work again with their business partners to say these are the new numbers they don’t match the old numbers but you should trust these new numbers so… Yeah, I think that was a big setback.

0:26:48.1 TW: Which did plug why one platform Piwik PRO switching to another tool for people that wanna keep the old. Sorry I couldn’t resist putting the ad boy on for them.

0:27:06.7 TW: You were gonna say Val… Oh I’m sorry.

0:27:10.6 VK: No no. It’s a good call out. I was just gonna say that selfishly in my realm where I focus, I work with clients a lot in the research experimentation and optimization space, the sunset of optimize, obviously not the same thing but related was a really interesting moment for us because the archetype of the client who doesn’t pay for a testing tool because Optimize came free was very different than who we typically had worked with and so they loved having the ability to… When the time was right or someone suggested meeting, having the ability to run that content experiment but not necessarily investing in the full-blown program, but now they feel like they’re at a loss and so now they’re thinking about well, I have to spend 15 grand for VWO or EBTC some of those other smaller players and now they’re thinking, How do I justify that investment and who should own this program and so it kind of became a moment to start a bigger conversation and it was really interesting. Some of the organizations that were like, “Well if we’re gonna do this, we’re just gonna do this, like we’re gonna go Optimize-ly, you should tell us what job reqs we need to hire up for,”

0:28:21.9 JH: That’s surprising.

0:28:22.2 VK: It was just like it took so many different directions I wasn’t expecting, but it allowed us to talk to a lot of different client types or prospects and it opened my eyes to a lot of different aspects of this that I hadn’t considered about making a case for in a really long time so it was interesting. It caused a lot of interesting conversations. I think I just said interesting six times.

[laughter]

0:28:48.4 MH: So to what extent do you all think that this may be representative of maybe just a Google struggling a little bit on execution across the board like we just had a big conversation about generative AI. None of us brought up Google Bard but they’re out there right. So that’s what I mean, is Google just becoming like another IBM or Oracle type slow player and not the pace setter that they once were?

0:29:17.9 TW: I feel like Google Analytics specifically because it was always not a lost leader but it was like here’s how we sell more. I feel like they have taken their eye off the ball and knowing enough people who are or were on that team, I do get the sense that Google was like we’re investing in this Google analytics thing and the money comes in from another channel where they’re making bank on search but it was kind of weird. They understood that dynamic for a while and then it does seem like… I think specifically on the Google Analytics that there was a failure in sort of execution and roll-out, although, granted every passing year their footprint gets bigger and bigger and bigger, it’s… Take a massive 100-year-old company, it’s a lot harder for them to shift and Google’s now been around, Google Analytics has been around for a long time so I do think it becomes more challenging the more entrenched that you are but I don’t know, I’m not gonna give up on their… I’m not gonna say that their AI execution is struggling not that they haven’t had some black eyes.

0:30:40.0 MH: All I know is Google Analytics doing just fine until Krista Sieden left and then… So, maybe Google you gotta look into that…

0:30:57.9 TW: The views expressed on this podcast are…

0:30:58.0 MH: Yeah. Yeah. Oh everybody knows already…

0:31:00.1 MK: Personal in nature.

0:31:01.3 MH: Yeah. Alright. Well let’s shift to maybe something a little more positive how about…

0:31:12.5 TW: Well I was… We could stay negative a little bit.

0:31:12.5 MK: Are you gonna direct us to something positive?

0:31:15.2 MH: No. I was waiting for one of you to jump in and say, Oh yeah this…

0:31:21.1 TW: I’m not the one to do that I can jump to another kind of negative but…

0:31:24.9 VK: Most of the ones on this list.

0:31:27.2 MH: Of course you can Tim.

0:31:37.1 MH: No this year was a big year I think for marketing mix modeling I think that was something that kind of people were starting to get a bigger grasp around, some cool technologies came out, people are getting more focus on it. It’s starting to slowly become maybe even like something people are willing to pursue instead of multi-touch attribution. Well…

0:31:56.0 MK: I do feel like this year there was maybe a bit more attraction with marketers as in like… I feel like the data community was kind of already there but I feel like this year is a year that like… And this is me personally again but I feel like I haven’t had to explain incrementality as many times this year as I have in previous years or like…

0:32:19.7 MH: That’s important…

0:32:25.0 MK: The concept of MMMs and how they work with experimentation and well I mean this is obviously hyper-personalized now but I do feel like there was quite a bit of progress there and there’s also just a lot of vendors, a lot more vendors in the market now.

0:32:31.4 JC: Yeah and also just the availability of some of these open source… Meta has their Robyn and they also have an Open Source, I just learned about GeoLift for geo experiments, like another package they put out for that. They had a session the other day, I think they called it their Open Source Marketing Science Summit something like that. And I was on that one and that was kind of cool just to see what’s coming out like Python API for Robyn, for someone that run into this GeoLift package and then immediately going to start exploring these just personally in my own work. So I think that’s kind of cool that they’ve done that and I may put a link to that one in the show notes ’cause that was kind of an interesting session.

0:33:15.7 TW: I think. One I remember… I remember Moe, you and I offline talking about how do you explain incrementality quickly and simply and clearly a year or two ago so I whole heartedly endorse that and I think Josh… And that’s if we look at the community from the MMM, big hats off to Jim Genolio for all the work he’s done to try to really support and Michael Kaminsky, who was a previous guest on the show not talking about MMM, but there does feel like there’s a community coalescing around that, there is a slack team for MMM and those discussions are fascinating ’cause there are people who are really smart and really knowledgeable and some of the questions I can barely even understand.

0:34:07.5 TW: And then you also see the questions coming in where it’s like… Okay now people think this is their silver bullet and MMM is gonna do everything for them so it’s like there’s an adoption curve for all of these… I remember a few years ago when I was at what the company now known as Further Search Discovery and a new Head of Data Science came in who… She was like Look it’s MMM, like that’s all it. And I’m like no no no, it’s art randomized controlled trials. It’s all that. And I feel like she was maybe over-indexing a little to the MMM I was over-indexing to the experimentation and now, I don’t know how much is me personally versus the industry, the MMM conversation seems to be talking a lot about experimentation complementing MMM and sure you can keep your multi-touch attribution… But we had was it John Williams who’s the Lift Lab guy, Moe, that we had on…

0:35:19.1 MK: John Wallace.

0:35:19.2 TW: John Wallace… That was kind of in that vein. And I was like at the time I was like this is wild. And some of this is, I think Canvas is doing some of this…

0:35:31.7 TW: Now it feels like there are lots of people saying sure I see where these fit in together and what’s crazy is none of that is this user level tracking bullshit, where you just get sucked into the minutia of how can you deceive the user so that you can actually track and target them in a way that they don’t want you to do but you’re staying, you’re telling yourself that you’re okay so I do think that’s really exciting.

0:36:05.1 VK: I think the loss of some of that data, like coming to terms with it is kind of maybe part of the reason that this is coming back up and maybe Moe to your point why it’s coming more in the realm of the marketers even ’cause it’s like just coming to grips with that reality especially as we near the next cliff.

0:36:25.9 MH: Yeah. Alright, well there’s too many topics and too short of time, so we’re gonna change our focus a little bit, and yeah we didn’t even talk about things like privacy which is obviously all with us, always with us. And change is happening all the time, but Julie what do we got to look forward to in 2024? What’s top of mind for you?

0:36:49.6 JH: Oh boy. I feel like going into next year, obviously, we’re still gonna have a generative AI stuff, I feel like so many people have that on their 2024 strategies plannings, I hear it all over. They wanna figure out how to use it in their day-to-day, so I wouldn’t be shocked there and then I do feel like there is gonna be a lot going on with Generative AI specifically for the elections, I know people have been talking about next year is gonna be crazy with that and people are talking about how do you identify what’s been created by Generative AI and what hasn’t been… So I think that’ll be… I’ve heard people talk about that quite a bit.

0:37:27.0 MH: Trust nothing.

0:37:29.3 JH: Yeah.

0:37:32.9 TW: Do your own research.

[laughter]

0:37:33.6 TW: What?

0:37:36.9 MH: Stuff too?

0:37:38.3 TW: I mean, is it with the generative AI, like that sort of sea change and needing data and needing data, especially I think as organizations, there’s a heightened awareness of the risk. Like you want to use your data with generative AI perhaps. And does that bring kind of a increased need for sort of data engineering and the infrastructure talent that both understands how these things work and then also can in a safely and traceable way pipe the data where it needs to be to use it? Like are there people out there, like I get periodically people emailing, do you know anybody who can with these sorts of ETL and architecture skills and I’m like… In central Ohio for 10 to 20 hours a week and I’m like hmm, yeah, no, I don’t. [laughter] Is there gonna be a talent issue on the infrastructure side?

0:38:46.2 MH: I mean, are we still talking about AI or are we talking about just generally in analytics?

0:38:50.6 TW: Probably both.

0:38:53.8 MH: Both? And it’s interesting. So like from an AI perspective, I totally agree with you, Julie, 2024 is gonna be still at the very top of that Gartner hype cycle inflated expectations type activity. Just everyone wants to put their hands on it. And the way I heard somebody explain it is like every c-level executive is telling their org do something with AI. And the entire org is like well, what do we do? And so there’s gonna be so much figuring out and there’s such expectation that it’s gonna solve major things or do major things. And it does have cool applications all over the place. And it is interesting, like I’m not versed enough on it to know well, what will transcend whatever in 2024? But I agree, Tim, like there’s a lot of trending towards sort of private LLMs at the company level or things like that where you’re kind of making your own little AI that does something specific for your organization and things like that. So that’s kind of interesting. I also think there’s been court cases already about the content that AI consumes and how is that copyrighted or not copyrighted. And it’ll be very interesting to see how those legal battles kind of find their way through to sort of indicate sort of like well, what is AI allowed to consume and then regurgitate and what rights do copyright holders have in that process.

0:40:19.6 JC: I Just heard an interesting interview with Sam Altman actually where he is talking about exactly that. Right? And his point was, well they shouldn’t just be regurgitating. It’s taking that information and to some extent, I mean the word he used was learning from it and he said, well, if people can read content and learn and synthesize and generate their own content off the back of that, then why can’t LLMs? I thought that was an interesting perspective.

0:40:49.0 MH: Oh. So he’s equating it to a person now? Is he? [laughter] No, I’m sorry.

0:40:57.7 JH: Well, one of the things that you said, Michael, too, about the talent part with generative AI, I do think it’s gonna be interesting which companies are gonna be really good at re-skilling their people instead of just hiring in. Because to your point, I don’t think there’s a bunch of people out there in the job market just ready to be hired to step in there. And I think it would be more beneficial to the organization keeping people who know their business and their data intimately and like have worked with it for years to actually give them space and time to re-skill into the generative AI space. But that’s always tricky because what if they’re really good at their current job? You know, it’s like how do you carve out space for them to do that and still stay profitable and don’t let those other responsibilities fall? So I think that’s gonna be an interesting hurdle.

0:41:42.3 VK: I’m excited for the job descriptions that are like 10 plus years experience, Generative AI.

[laughter]

0:41:48.3 JC: Yeah. Oh. Did you see the kill switch engineer job description? Just while we’re on that open AI kill switch engineer?

0:41:57.4 JC: No? Too terminally online for this group, but there was a… It’s like a meme of open AI posted a kill switch engineer and it’s just like, oh we just need someone to stand by the servers and you know, throw some water on them if this thing starts going haywire… 500k per year.

0:42:19.6 MH: Year. That’s an important job.

0:42:19.7 TW: Oh my god.

0:42:20.4 MH: Willing to work weekends. [laughter]

0:42:24.1 TW: I’m sorry, I cannot do that for you. [laughter]

0:42:28.6 MH: So it does tie into another topic though, Tim, that you kind of surfaced, which is around talent, which is certainly there’s a ton of talent that’s leaning into AI and we’re all trying to explore that and figure out where it works and where it doesn’t. And skilling into that space. But I also know that because of things like the rollout of GA4 and other things like media mix modeling and marketing mix modeling, there’s now this need more than ever for analytics engineering talent in organizations that haven’t historically had that. And I see a ton of need for that going into 2024 as well.

0:43:02.4 MK: Really.

0:43:03.4 MH: Yeah, there’s not, it.

0:43:05.4 MK: I just, I feel like they’ve always been there. We’re just calling it something different and like I do think that we potentially need to have more of those skills and that’s because people are doing stuff in-house more. And like maybe less dependent on tools that do like the ETL component for them, or like they want things to be highly customized and build things themselves. But like I feel like I have been the bearer of that skill gap for a while. Like I don’t feel like there’s enough people in that field, but I don’t feel like it’s…

0:43:38.0 MH: The only difference. So I would say Moe is, I’ve worked with a lot of companies this year that have stood up a data warehouse for the very first time because of GA4.

0:43:45.1 MK: Really?

0:43:48.5 MH: Yes.

0:43:49.5 MK: Wow.

0:43:49.6 MH: So a marketing data warehouse is a brand new thing in certain places and certainly in the medium sized business sector where a lot of my clients are, that’s not something they historically had done for marketing analytics or digital analytics. And now it’s a necessity. And so that’s where I see this gap emerging is sure, there’s plenty of analytics engineers and we’re always in demand for those, but there is a whole new part of the economy opening up to this need that hasn’t really needed it before. And it’ll be very interesting, at the same time, I don’t know if I recollect a time when I’ve had more friends not having jobs in the analytics industry in terms of just layoffs or in between work or whatever. And so I’m just… It’s an interesting weird time ’cause it’s sort of like okay, so is it sort of like a… Like you said Julie, is it sort of like a skill hopping issue of sort of like we need to move from sort of being digital analysts to being data analysts and analytics engineers with the right skill sets or… So it’s just… It’s very interesting.

0:44:51.9 TW: When I think of… I mean I know some super talented, thoughtful, creative experienced analysts who have been caught up in layoffs for whatever reasons. And to me it’s super disheartening. Like what scares me is if we say, oh, the problem is we need more data engineers. Like there’s part of me that says, good lord, if we just scrapped all the data, I’m not gonna worry about the marketing and their wants to creepily do retargeting and targeting of their audiences. Throw all that out from an analytics perspective. Like what if we had no data outside of revenue at an aggregate level and said instead of hiring that all that infrastructure, we’re just gonna come up with really good questions and we’re gonna do it purely experimentally or something. That’s what’s… What scares me is that this becomes this death spiral of chasing the data completeness and the data quality and the multiple data sources. And I do like fundamentally at my core think that that is scary ’cause you can suck that… Doing that will expand whatever investment you can put into analytics. That can be… It can be filled with that. And I think there’s an opportunity to say, what if you stop and say it’s good enough and you focus on the business.

0:46:29.0 MK: So do you think Tim, like just trying to understand like exactly the problem. Is it like we’re getting too complex and like ultimately our practices and like how we’re… I don’t know, modeling data and all that stuff is just like we’re focusing on complexity instead of whether we’re asking the right questions and that if we focused on whether we asked the right question, you would actually realize you would probably get a good enough answer with some less complicated practices. Is that the TLDR?

0:47:06.2 TW: Yeah. I mean Matt Gershoff and I’m sure he did not coin this, because Matt will never take credit for coining anything, but he was like the Just in case versus Just in data collection. And I’m like that’s another like brilliant phrase I’ve picked up from him that we index to the, just in case like we’re better add that data source ’cause we’re gonna need to be able to do year over year comparisons of clicks on the Global Nav or you know, whatever it is. And it just spirals. I mean look at the MarTech landscape explosion. Well, I mean everyone, even the categories have… The number of categories that’s exploded and you say, well realistically you’re gonna have lines connecting every one of those generates data and you want to connect them. And it feels wrong to me because I watched the discussions be around which data is correct.

0:48:06.0 TW: Can we have this other data? Can we connect this ID who completed a lead to the ultimate conversion? I mean, I was having a discussion with another industry person and I’m like oh my God, like you could… It’s a good idea. You had a good idea and then poorly executed an experiment and now you’re in the minutiae of the weeds of trying to exclude bots and do this and do that when instead you could have just said, what if we designed an experiment well, took a big swing and then all of that goes away. So there’s like simpler answers that aren’t more complex data.

0:48:43.9 JH: Do you think the complexity is gonna help the pendulum swing back? Because we even talked on a few episodes this year about like better defining the problem. And I’ve heard that a little bit more with some groups I’ve worked with this year. And I do wonder because everyone’s kind of realizing, oh it’s so complex. People have been more… Or their eyes have been open more because of privacy and other things going on that like collection of data has never been perfect. It isn’t going to be perfect moving forward. So do you think that will help push people actually to kinda like focus back in on like I need to be asking a better question. So then when you dive into the complexity, you’ve like better structured your guardrails around like what you actually need to care about looking at because you know, the like problem you’re actually trying to solve and the question is more clear.

0:49:31.3 MH: Got it. [laughter] I hope so, but I…

0:49:34.8 JC: Well the thing I was gonna say earlier, so you talked about that there’s some clients that you’ve worked with that have stood up for the data warehouse for the first time. It’ll be really interesting to see how they leapfrog past some of the pain points that a lot of other companies have been experiencing along the way. ‘Cause I just learned, and I might be late to the party, but I just learned what data mesh is, that whole like decentralized data architecture model. And I was like well, shit, just when we thought that we understood the problem [laughter], I think that that’s gonna be another thing that adds potentially complexity. Even though I think it’s intended to be more simple. I don’t know. That’s another layer of consideration.

0:50:16.8 TW: But the whole modern data stack, I mean there’s another like Benn Stancil shout out. Like I feel like he is, whether he’s intended to or not, he’s leading the charge with kinda ringing his hands about, Oh my God, we keep having the new solution and the new solution. Oh, you just add one more, well seven more pieces of technology and just stitch ’em together and then wish away the cost of that complexity and then it’ll all be good. I mean, data lakes from when data lakes like exploded.

0:50:46.4 MK: But isn’t this like…

0:50:47.8 TW: I mean there have been some…

0:50:48.9 MK: Tim. I just feel like I mean it’s all a bit negative. Like isn’t this what we strive to do? We strive to build something better to like fix the problem, to build technology to make things easier. And like yes, I agree that it increases complexity or I don’t know, but like I feel like the intent is good and the world and the technology that we’re existing in is like…

0:51:15.6 TW: Yeah, I’m not knocking the intent.

0:51:15.7 MK: Getting more complicated. So therefore solutions are getting more complicated.

0:51:20.9 TW: Well. You’re the one who called stakeholders stupid. Who…

0:51:24.3 MK: Also, I want to retract that. Can we please retract that? I’m mortified I said that because if someone said that to me, I would be like oh, oh, oh, oh. We do not call people stupid. So I’m very sorry for saying that.

0:51:36.3 TW: Oh, it would’ve come outta my mouth and this could have easily, it could have been switched, it could have come out of my mouth and you could’ve called me on it. So, but the key is that there is a human nature instinct to say, what can I add to this to make it better?

0:51:52.1 MK: Yes.

0:51:53.1 TW: And I think that is a massive challenge to say, what could I remove to make it simpler? And now not only if I realized that I’ve gotta convince somebody else that they don’t need that data, they don’t need that integration. They don’t… I had a friend who spent like three years at a major financial institution doing nothing but data cataloging of terms when he said, this is a waste of time. Nobody uses it. We are making a data dictionary for the purpose of having a data dictionary that no one uses. He was like this is insane and I can’t get them to stop. So I think it’s just recognizing what we’re being drawn towards. Sorry, Michael.

0:52:37.9 MH: Yeah but…

0:52:38.3 TW: I had to get that out. [laughter]

0:52:41.9 MH: No, I think maybe what we’re seeing is that the… To your point Moe the increase in complexity is splitting what we have typically considered digital analytics or analytics. It’s splitting it out into its component parts. There’s some technology associated with this that’s now moving back over into the technology function. And there are analysts and analysts do things that look at how do we add value? How do we hypothesize and create a valuable outcome? And so like to your point Tim, it’s like that what you just said, it’s sort of like well, let’s get it down to its brass tacks. So how do we drive something of value through this information that we have? And maybe one of the places that AI can help us is an analyst could sit down and ask an AI, “Hey, describe this data set to me. Tell me everything I need to know about it,” and ramp up their data acquisition skills.

0:53:33.1 MH: What Viyaleta Apgar talked about in getting ready to expose and do analysis in better and better ways. But there’s a lot of people in our industry who are actually more like technologists or developers or data engineers or analytics engineers and they actually are more technical and that’s fun too. It’s just a different profession or emerging into a different type of profession. And I think maybe that’s what’s happening is we’re seeing that slowly split apart and it’s certainly gonna be underappreciated or under misunderstood at a sort of corporate level ’cause those things are slow to change, but it is exciting.

0:54:14.1 TW: Can I remount my soapbox one more time? Like if you ask who gets paid with the…

0:54:18.9 MH: Oh geez.

0:54:21.3 VK: Oh geez.

0:54:21.4 TW: Increased technology…

0:54:21.4 MH: Is there any chance that you wouldn’t necessarily do it? Or is it for sure guaranteed.

0:54:27.3 TW: I’m producing this episode.

0:54:30.2 MH: Okay. Well then go ahead. [laughter]

0:54:31.5 TW: I mean if you think about who gets paid when you add complexity, a lot of times that’s adding technology and the technology vendors get paid when you add the technology, not when value gets realized down the road. Consultants…

0:54:44.7 MH: That’s right.

0:54:45.0 TW: Often get paid to implement the complexity, support the complexity. So again, well intended, I’m not knocking them, they believe in all of that, but there are just the forces that are driving an additive mentality that it is really hard to push back against that and say, let’s… We shouldn’t be doing that. So just the incentive structure is unfortunate.

0:55:13.1 MH: Well, I’m excited to see what you do about that in 2024 Tim [laughter], wink, wink. Alright, we have gotta start to wrap up. This has been so great talking to all of you. I like how passionate we are about these topics, about what our industry is going through. I don’t know, any final words, wisdom, let’s go around, we’re not doing a last call, but maybe just any encouragement to…

0:55:40.6 JH: It’s just not called the last call, but it’s still a last call.

0:55:43.1 MH: It’s not, it’s a word of encouragement to our listeners. Julie.

[laughter]

0:55:49.6 JH: Panic.

0:55:50.7 TW: The word of encouragement is for every listener to find their little community of like five other people that they can periodically sit and have conversations about the space like we have. So I’m gonna throw out more my like positive thankfulness of having this podcast go for another year where I get to rant and actually get smacked around by Moe and others.

0:56:19.8 MK: Hey.

0:56:20.8 TW: Justifiably.

0:56:21.5 MH: And Frankly Tim…

0:56:21.6 TW: Well deserved.

0:56:21.7 MH: And frankly Tim, we’re sick of your shit. [laughter] That’s right. No [laughter]

0:56:29.3 MH: Oh man.

0:56:29.4 TW: I tried to start it off.

0:56:33.2 MH: No, you did. I’m waiting for somebody else to go now.

0:56:34.0 JH: Too good it’s too high of a bar.

0:56:38.4 MH: Oh geez.

0:56:39.5 MK: I think my thing and we actually didn’t touch on it today, that just like I keep coming back to is we keep talking about the technology and the problems and the working with stakeholders and stuff and the bit that I just am already thinking about for next year is like how do we keep developing the soft skills so that we can do all the stuff we need to do and fit in with AI or whatever the technology is. It’s still just the bit, yeah that’s top of my mind is the how do we communicate better and it doesn’t matter what the technology we’re using is it’s still just like… It’s all about that stuff. I don’t know if that was positive or just pensive but…

0:57:28.6 MH: No I like it…

0:57:28.7 VK: I don’t know this will be my small piece of advice. At least this is what I’m trying to remind myself going into next year so maybe it’s helpful. An oldie but a goodie, control what you can control. I think especially like we were just talking about in the face of complexity. Sometimes just focusing in on your sphere of influence, the direct people you work with like what good can you do that day in that space can be kind of helpful and calming when things feel crazy and big and too much to move on your own.

0:57:56.7 TW: ChatGPT to write an…

0:58:00.6 JH: That’s a good one.

0:58:01.2 TW: Analyst-oriented serenity prayer right now.

0:58:09.8 JH: Oh send it on over.

0:58:11.0 VK: I think in the spirit of things to keep in mind for 2024 because generative AI for me still just feels like an answer begging for a question. And this kind of ties still to what we were just talking about with chasing technology that if we really stay rooted in the business problems and the opportunities, you really can’t go wrong right? So just making sure that you’re constantly checking in and asking yourself that, I think will help kind of be some of your bumpers on your lane to make sure that you’re not just chasing the next cool, shiny thing.

0:58:51.4 TW: Oh my God. Grant me the serenity to accept the data I cannot change, courage to interpret the insights I can and the wisdom to distinguish between statistical noise and meaningful patterns. Then it goes on but I’ll stop there…

0:59:04.0 JC: Who said ChatGPT can’t be funny.

0:59:04.5 JH: Tim that could be your first tattoo. Just get that. That feels right.

0:59:11.0 MK: Oh man.

0:59:11.7 MH: With the Analytics Power Hour pioneering the first 12-step program for analysts, that’s perfect.

0:59:17.3 MH: Oh my goodness.

0:59:18.4 TW: It’s got five full stanzas, so yeah that’s great. Josh you got anything.

0:59:23.5 JC: I don’t know. Just try to be useful. That’s my mantra for 2024. Just try to be useful.

0:59:35.6 MK: I love that.

0:59:37.1 MH: That’s good.

0:59:38.1 JH: Simple sweet. I like it.

0:59:39.2 JC: Where my head’s at.

0:59:40.6 TW: It’s kind of channeling like Cassie Kozyrkov right? She’s…

0:59:43.1 MK: I was thinking of her when you said that Josh, I was like that makes me think of Cassie.

0:59:48.7 JC: Oh nice.

0:59:53.4 MH: It’s interesting. I think for me, I think the thing I’m trying to do more of in 2024 is just take my mistakes and learn from them, like always, but it’s been a fun year and I look forward to next year in terms of what things will happen. And I think it’s… Somebody said and I’m clinging to this desperately that it’s a sign of intelligence to learn from your mistakes. So that’s what I’m trying to do. And so yeah that’s the encouragement I’d have for somebody else out there. If you’re making mistakes too then just remember that if you’re learning and adapting to them it’s a sign of intelligence. So you’re super smart.

1:00:30.4 MH: All right. Well, 2024 is going to be a great year and robots will be able to do our jobs for us. We’ll be able to sit back and just pluck insights from the lowest hanging fruit. Yeah, that one’s specially for Jim Kane out there. And yeah I don’t know what else to say but obviously no show would be complete without a huge thanks to Josh. And you’re right here Josh, yay. So thank you.

1:01:00.5 TW: Josh you are so useful with every episode.

1:01:05.9 MH: Yeah, yeah I’m just saying we don’t get to thank you in person that often on the show, so thank you for everything you do for the show.

1:01:06.0 JC: Well, thank you guys.

1:01:06.1 MH: Julie and Val, what a privilege to have you two join us this year as co-hosts. Thank you so much for all the work and time you’ve put in to help push and pull the show forward into new and exciting areas. This has been a breath of fresh air and amazing to have you both join the show this year. So thank you.

1:01:26.6 VK: No one’s told me that they regretted it yet so…

1:01:30.3 MH: Yeah all the surveys…

1:01:34.2 JH: Maybe in 2024…

1:01:34.3 MH: All the surveys we’ve put out have been pretty positive.

1:01:35.6 TW: 2024 year in review.

1:01:39.9 JH: Yeah get the hook.

1:01:44.5 MH: And Tim and Moe, what can I say but you two are awesome. And I love working with you on this show. It’s a delight even though we might not always get along. No but we typically always come back together and that’s what sort of makes this podcast happen. So I loved looking at the issues that face our industry together. And I look forward to next year. Please do reach out to us. If you try out our AI tool astral.com/@analyticspowerhour, I think, we’ll have it in the show notes. But go try it out. See what you think of it.

1:02:12.2 TW: Try astral.com or bit.ly/aph-astral@digitalanalytics… Oh shit I don’t know.

1:02:19.6 MH: There you go. Yeah, that’s right. I don’t know the link, so we’ll get that figured out and put it in the show notes and that’s where you can find it. And yeah anything else you want to say to us please do so. Measure Slack or LinkedIn whatever you want to say. We appreciate you, appreciate you listening. All right onward and upward into 2024. And I know I speak for all my co-hosts Julie, Tim, Moe, Josh, Val, when I say, whatever 2024 throws your way keep analyzing.

1:02:56.8 Announcer: Thanks for listening. Let’s keep the conversation going with your comments, suggestions and questions on Twitter at @analyticshour, on the web at analyticshour.io. Our LinkedIn group and the Measure Chat Slack group. Music for the podcast by Josh Crowhurst.

1:03:17.5 Charles Barkley: Smart guys wanted to fit in, so they made up a term called analytics, analytics don’t work.

1:03:23.9 Kamala Harris: I Love Venn diagrams. It’s just something about those three circles and the analysis about where there is the intersection right?

1:03:30.5 MH: Just happened. I know. [laughter]

1:03:32.3 MK: I appreciate that. I love that.

1:03:34.5 MH: No as I was reading that I was like I don’t know. It’s like we’re just going to jump into it.

1:03:39.3 TW: You’re right.

1:03:42.4 MH: It’ll happen. It’s dare Tim all over again.

1:03:43.6 TW: There’s like God how long did does it take to calm me down. ‘Cause I’m working on it.

1:03:48.9 MH: No. Somehow Tim, Val asked in a way that doesn’t make me upset. So I don’t know what she’s doing differently. [laughter]

1:03:57.0 JC: Soft skills, soft skills.

1:04:05.0 TW: [laughter] Man.

1:04:07.2 MH: All right, good that you are recording that.

1:04:07.3 TW: Wow, glad I got that on you.

1:04:12.2 VK: In 2015 I had to become an official registered US importer/exporter because of all of the shit that I bought from Australia.

1:04:22.4 MH: [laughter] What?

1:04:24.7 VK: My husband was like I’m sorry, what is happening?

1:04:29.8 MK: What did you buy?

1:04:30.9 VK: Do you know that brand BlackMilk, that did the crazy leggings and everything. It was the start of that trend. Yeah well, I was obsessed.

1:04:43.1 MK: Okay, also do you think I’m a crazy legging person? My clothes are white, cream, grey, black, navy. That is the most color I own.

1:04:50.0 MH: That’s awesome.

1:04:54.0 VK: I’ll show you some pictures, well, you’ll appreciate…

1:04:54.1 MH: You’re showing us right now, oh wow.

1:04:55.1 VK: It’s insidious.

1:04:56.2 TW: What was the company that they had on Seinfeld. This Import/Export?

1:05:00.7 VK: Vandalay.

1:05:01.9 MH: Vandalay Industries.

1:05:04.9 TW: Vandalay Industries.

1:05:05.4 VK: I was so tempted to include that in my paperwork, you have no idea.

1:05:13.4 TW: Oh, Valdalay Industries Import/Export. That’s such a random… How did you do it Val? That’s so random.

1:05:23.7 JH: Did they just stop you like, ma’am you’re buying too many leggings from Australia like here’s some extra paperwork.

1:05:28.7 VK: Well, every like cloth had to go through customs and so once it reached a certain…

1:05:29.2 MH: You must be a retailer, nope just for me. Yeah. Did they give you a wholesale price and you’d just be like yeah I’m a business. I’m your BlackMilk US retailer now.

1:05:41.2 VK: I never thought of that. But I only need one size in all of them.

1:05:47.1 MH: That’s right. Yeah. I have a very specific target audience.

1:05:50.3 TW: That’s right. I sort of…

1:05:51.9 VK: I know my consumers.

1:05:53.5 MK: Oh my god, I can so imagine you in this.

1:06:05.5 MH: Okay, perfect.

1:06:06.8 TW: And we’re doing, no last calls.

1:06:07.9 MH: No last calls that’s right. We’re gonna scream at each other to the very end and then… Unless…

1:06:12.5 MH: Okay. No calls. Alright, we are singing the Christmas song though at the end everybody you’ve practiced that.

1:06:23.2 JC: Fuck you.

1:06:24.3 MH: I’m sorry.

1:06:28.5 VK: I was like, which one?

1:06:30.3 MH: I was mostly…

1:06:31.5 JH: I guess I was like I’m game.

1:06:34.3 MH: I was mostly trying to get a reaction out of Moe and she was perfectly straight faced, like yeah I totally expected that.

1:06:42.4 MK: No. I was like fuck you, that’s never happening.

1:06:49.8 MH: I know, okay well, it’s the same thing. Alright, let’s go. Rock flag and be useful.

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

#239: Non-Technical Backgrounds in the Modern Analytical World with Kirsten Lum

#239: Non-Technical Backgrounds in the Modern Analytical World with Kirsten Lum

https://media.blubrry.com/the_digital_analytics_power/traffic.libsyn.com/analyticshour/APH_-_Episode_239_-_Non-Technical_Backgrounds_in_the_Modern_Analytical_World_with_Kirsten_Lum.mp3Podcast: Download | EmbedSubscribe: Google Podcasts | RSSTweetShareShareEmail0 Shares