The Haunted Frosty

Speaker A: Foreign. This is Marketing Over Coffee with Christopher Penn and John Wall.

Speaker B: Good morning. Welcome to Marketing Over Coffee. I'm John Wall.

Speaker C: I'm Christopher Penn

Speaker B: and ah, it's the uh, spooky Halloween edition. I've come to the point, you know, we've talked about April Fool's Day, how that's basically obsolete because the Internet has infinite lies for us every single day. And, and I'm kind of feeling that Halloween is the same way too. And that like, there's scary stories every single day. So. But I think there's a place for costumes, trick or treating and fun houses at least. Like that's legit. But yeah, I don't know. Anything going on in your Halloween celebration calendar?

Speaker C: No, it's all candy all the time. Uh, Janessa Drainville commented, uh, on LinkedIn the other day. She said, uh, uh, for the love of marketing, please enough with the no tricks, only treats emails. So I, being my usual self, sent out one saying all tricks, no tricks streets. Uh, and it was a.

Speaker B: I was gonna say that's an excellent list. I'll put a link to that in the show notes because you've got some interesting things in there too. Uh, you know, uh, tricking people into subscribing to your YouTube channel and the like. It's stuff that's definitely worth doing.

Speaker C: Exactly. And there's some, some other silly stuff. Our friend CC Chapman said that his favorite of the list was learn how to cheat at card games like second dealing and, and bottom dealing. So that when the inevitable holiday, you know, Uno competition or whatever rolls around at your family, you can cheat by giving yourself the worst possible hand so you can exit the game as quickly as possible,

Speaker B: get out earlier.

Speaker C: Exactly. Because nobody complains about cheating if you're cheating yourself,

Speaker B: Right? Yeah, exactly. And everybody loves to win. Well, yeah. And I can do a classic rant for me on holiday games too of, um, Monopoly. Monopoly was designed to prove to people how bad monopolies are. It's not a game for fun. At no point did anybody say that this was going to be entertaining. Hell, I've had branches of the family that get to straight fist fights over monopolies. So if you're just arguing and angry, you're actually doing all right with that. So throw away your copy of Monopoly and yeah, get yourself out of the card game, folks. You're winning today. You're just getting stuff you can put to work and save your holiday. So back from Macon, we were talking about that. So this is, you know, marketing AI's premier event, the Biggest thing going in the States. How did it go and what were your takeaways?

Speaker C: Oh, it was so much fun. Uh, so Katie did a three hour Claude workshop, Claude for Business, which was fantastic. She does a great job with it. I recorded it and um, we're going to turn it into a full course in the Trust Insights Academy. But everybody took and said it was fantastic. We have people asking like, hey, how can I bring this workshop specifically to my company? So there's a lot of follow ups. Uh, I did a build session, the first of its kind where people had to make something in the session and what I showed was how to use the built in Python libraries in every single AI system to do analytics better. Um, and what people came up with was really shocking to them to say, like, oh, I didn't know. Even regular, even like the dumbest version of Copilot has a tremendous amount of stuff built into it, like over 300 different libraries that it can use. And so people walked away at the session going, okay, I have a skill now, a technique that I can use to dramatically increase the accuracy and um, interest of what you can derive from your data. And I said in the session, if it fits in a spreadsheet you can use this set of techniques. And if you want to check it out, that session is recorded. It's available for free on the Trust Insights Academy. We'll have a link in the show notes, but it is completely free. You can, It's I think 31 minutes long, uh, after editing and stuff. But um, it was a fantastic conference. Again as with last year, the big thing that was missing, no mention of China, no uh, mention of Chinese models, like passing mentions of open weights models. But it really is the, the whole OpenAI, anthropic, Google meta, uh, sort of ecosystem which is really interesting. Um, we might talk about this in a bit, uh, with Anthropic's S1 prospectus for their IPO being announced.

Speaker B: Yeah, and this brings up something that we didn't have on the topic rundown, but I had come across this just yesterday and wanted to get your take on this. This idea of distilling models, I didn't realize that this is a thing. So OpenAI and Anthropic do their model. The Chinese models instantly grab it and use it to train their latest thing. It seems like it's just a fantastic way to automatically upgrade your product. Is that, and that's just endemic to how this stuff works? No.

Speaker C: So there's a lot of FUD on that. Uh, unsurprisingly put out by the Western technology companies, uh, because their IPO is like, if it was widely known just how good Chinese models are, the IPOs of the Western companies would be in the toilet. Because everyone's like, well, why would I invest in this company that's, you know, has committed for every dollar of revenue it has earned, has committed over the next 10 years, $113 in unbreakable contracts for compute costs. Right. This is not a winning situation here when you can get the same performance and the same quality out of an open weights model for 1/50 of the cost. Right. So there is, yes, Chinese companies do do what's called distillation, which is a, uh, fancy for, hey, I'm going to prompt your model, I'm going to record its responses. I'm going to reverse engineer how your model did that. I'm going to teach my model how to do that. Like, that's all distillation is. It's is, they can't see the inner workings. But just like if you were a chef of a restaurant and you wanted to understand a competing chef's work, you go to their restaurant, eat there, take notes during dinner and go, okay, I'm pretty sure I know how you did this.

Speaker B: Right. Well, but then the thing with that is they do that at AI scale. The ability for it to do millions of millions of queries and, and to be able to pull that data back out, you know, just as quickly. So that makes sense.

Speaker C: Here's the thing about this. If you look at what the Chinese model makers are doing, particularly Alibaba Minimax, um, and Deep Seq and what they're releasing, they're kind of past that now. Like they will do that to kind of do alignment and reality checking. But the math that these, the Chinese model makers are coming up with is in many cases better than what the Western labs are doing. The Western labs are kind of brute forcing things. Uh, and again, we don't know for sure because we, you know, they're all closed weights models. You can't see under the hood. When Deep, uh, SEQ released its inference framework this week or last week, everyone can go download it. Everyone can go look at and go, oh, that's how you're making your thing. And if you, this is of interest to you, you can download it and use it yourself. Now granted, you need a lot of compute to support it, but it's very interesting to me just how much FUD there is around open weights models and things because the US got into a math war with China. That's not going to end well.

Speaker B: Right, Right. You're going to drop the ball on that? Uh, okay. We can kick some more stuff around with that. We just have to take a second. We want to thank Framer for their support of marketing over coffee. If your team wants a website that looks and feels handcrafted but is still fast to ship, Framer is built exactly for that. You design on a visual canvas with responsive layouts, hosting and CMS built in. So the work is production ready. From day one, agents work alongside you to draft pages and polish sections. Then you review and publish what goes live. Just what we talked about. Human in the loop. Agents and humans work in tandem. Agents bring speed and scale. People bring taste, judgment and control. If you're a larger site or a corporate organization where you've got a bunch of checkpoints and people that you have to go through, don't get wrapped up in a pile of tickets and handoffs. Framer helps your team move faster. Teams can collaborate in real time, iterate on the same page and publish instantly. Agents help you go beyond the vibe coded site and a first draft to bring you ah, a production ready site faster than ever. Agents work alongside teams to help you build custom code components, create and manage CMS content, optimize SEO settings, audit for issues and ship everything all in one place. It's an enterprise level solution powered by agents that provide premium hosting, enterprise grade security and 99.99% uptime. Service level agreements that give the world's leading brands like Perplexity and Miro the confidence they need to build their websites and Framer. Learn how you can get more out of your site from a framer specialist or get started building for free today@framer.com MOC for 30% off a Framer Pro annual plan. That's framer.com MOC rules and restrictions may apply. Framer.com MOC and we thank them for their support of the show. Yeah, so a bunch of big numbers getting thrown around with a bunch of stuff. Salesforce acquiring Braze was big news that was actually talked about back in June at their event. But now it's all official. It's happening. Uh, 5.6 billion against about a billion per year. So you know, 5, 6x multiple which is not crazy at all. I should know more about Braze than I do too. I have to say that I know it's real time monitoring adaptation and it will help Salesforce make it more adept. There is some discussion. I've seen a lot of people talking about like do we really need more features and more things to figure out how to plug in? But that's just the reality of, of how that works. Is that on your radar at all? Anything you've heard about that or concerned about?

Speaker C: So I've used Braze, I've used their software, uh, actually on behalf of a client and it's yet another. Okay, well this is nice. This is sort of a weird blend between CRM and marketing to automation, but it lives in a weird little spot. I uh, mean Salesforce clearly is going to fold into their platform. That's not, not a huge surprise. Thing I always wonder about though with these things is does it actually provide lift? Right. Does it, does it do something that you can't do with existing technology? Like when it comes to things like, you know, real time engagement and high speed engagement. If your salespeople were not drunk all the time and actually check their inboxes, would you need that software?

Speaker B: Yeah. Right, right. Can it happen? That's definitely. Uh, and this week has been really hot with talking about measurement because we've had Rand, uh, fishkin at SparkToro, had an article talking about is GEO worth doing? Like what are the results of that? And we came to the same conclusion of this. Well, just because you can't measure it doesn't change the fact that you know, that is where all the leads are coming from. So like this measurement gap. But it seems like Braves might on the other side of that, that actually is not changing the outcomes. You know, I mean you're there and doing it but uh, yeah, we keep coming back to this thing of yeah, it's difficult to measure but that doesn't change reality and the fact that you have to adapt to what's going on and actually being used.

Speaker C: Yeah, exactly. And this is, you know, measurement has always a, uh, marketing measurement has always struggled with long term stuff and with stuff that does not have a direct click stream. Um, in fact uh, as we record this we're getting ready for marketing, perhaps B2B form. And one of the big topics that uh, the keynote speaker, Rory Sutherland will be talking about is when you cut intangible experiences like having a doorman to save operational costs, you lose intangibles like the fact that the doorman knows who people are in the building, um, and, and can create a higher level of satisfaction with the building experience that a machine can't do. Like a machine will not greet, you know, so and so and help her get her groceries, you know, to, into the elevator and stuff like that. And if you're. If your goal is tenant, um, uh, retention, firing the doorman, which will save you some operational salary costs, may cause you to lose money in the long run because now your tenants are disassembled. How do you measure that? Uh, it's very, very difficult to measure. The thing that always comes back to me is this. We now have AI large language models. That's literally language is built in the name. There is no excuse for anyone to not be collecting and processing the language of your customers now. So, for example, on all the forms that we publish for Trust Insights, we have a How did you hear about us? And it's a free text field. And it is trivial to have a language model go, okay, categorize, uh, whether this is AI or Google or social media and stuff. And so we don't have that measurement problem because we let people talk and then we actually listen to what they say.

Speaker B: Yeah, that's been game changer for us. Like, that's always been a problem is okay, you know, you have to force everything into a pick list or somebody has to manually go back and do that. And having the machines just grab that is. It just saves all the time in the world and gets you again where you want to be as far as actionable. Um, and yeah, this all bundles up under crappification, right? It's just the slow, gentle slide. I do have Eric Ries, the Lean Startup guy's new book, Incorruptible. I'm halfway through that and I'm interested to see if he can kind of solve that. I don't know. So far it's come down to, you know, you just have to have senior leadership that gets it and is willing to commit. You know, like, it's one thing for them to glad hand and say, yes, we want to have the doorman there because we love service. But do they have the intestinal fortitude not to lay them off at the quarter, you know, to make their budget or whatever or their bonus or whatever the, you know, problem of the week is. So, yeah, I don't know more on that. Another big one was Doordash buying Instacart for 17 billion. I mean, just this idea of having one monolith delivery company. Uh, obviously this administration probably will pay no interest to that. I think other administrations would start to be worrying about Monopoly and, you know, competition disappearing. It's also hard to challenge the fact that subscribers only pay for one service instead of two. There's a lot of people that use both those services, and now they pay for One. But, uh, yeah, I don't know, what's your thought on this? Really? Things have changed. I've even been amazed how Home Depot has picked up on this. I mean, I have random parts and tools to show up same day at my house and I live in a pretty rural area. It has been game changing. But again, I. Home Depot knows how much I spend every year with them, so I'm surprised the manager doesn't drive over here to, you know, with a burrito. But yeah, what are you thinking about as far as how this, uh, changes the shopping landscape?

Speaker C: Here's the thing. The consumer is under more pressure than ever to handle, you know, ever rising costs. You know, regardless of, of, you know, all the election and political stuff, the reality is things are a lot more expensive. I don't use any of the services and I, I literally will just take 15 minutes to get in my car and go get it myself. Because why, right, if I want to get food from a restaurant, you know, uh, I know the restaurant's my area, so I will just get in my car and go get it and then I don't have to pay delivery fees and all this stuff. I will literally go and just get my car and pick up things. So in terms of, you know, I think you're right to say this is part of the incredification of everything that, uh, we're get a hold of something, you acquire it, you consolidate it, and then you say, okay, yeah, we're. How can we cut our costs? I saw in the news that a private equity firm is looking at Wendy's, uh, in particular because they control a lot of valuable real estate. And I suspect they're probably going to pull the same thing they did with Red Lobster, where you sell off all the real estate and then you lease it back to your franchise until you drive the franchise out of business. So if you enjoy Wendy's, enjoy it now while you can

Speaker B: get yourself the, uh, one of those Frosties. It's too late. All right, there's a. There's, uh, our Halloween horror story to match the title, the Vanishing Wendy's Frosty. I've got coming up next Thursday at 1pm Eastern. I'm on with Quoted. They're doing a webinar talking about how to get your clients or management on podcasts. So I'll be telling all kinds of stories about the horrible pitches I get and, uh, what actually makes it through the gate. You know, what's good enough to get in there. So I'll have a link to that in the show notes. You can check that out. Uh, and also Zwift season is back, the cold has struck and I am biking indoors. I did my first shakedown ride. All the stuff still worked. The one thing that I missed, I ended up drinking like six months worth of dust that was wedged into my water bottle cap that I hadn't. That was, yeah, exactly that one. And a new battery for my heart rate monitor. But so yeah, if you're on Zwift and you like to race, ping me over there. We'll get that going. I did have in the video game front James Bond, First Light and Resident Evil Requiem both uh, games worth playing and checking out. My in house game tester field tested those and ranks them highly. I don't know how about anything you've got going as far as media, uh, and movie stuff that you've been checking out?

Speaker C: Uh, World of Warcraft Forever is coming out in uh, I think it's a couple of weeks. This is a reboot of the original game game. Uh, and they've said at least for now they don't plan any expansions of it. Uh, they instead are going to just keep adding content to the world and the, the philosophy changes. An interesting change from the they say the, for the retail game, the philosophy is you the player are the hero. Right. Um, you're the hero of the story. In forever they said their philosophy changes. The world is the hero and you're just in it. And so it's a very different perspective. So we'll see how it goes. Uh, I, I, it's included retail subscription. So if you have a retail subscription, you know, you can play it uh, without any additional cost. So we'll see how it goes.

Speaker B: Yeah, that's interesting. You know that idea of the world being the hero. I mean off the top of my head, the easy win with that is kind of if you have to make everybody the hero, that's very limiting as far as storytelling. Right. There's no sidekicks, there's or everybody has to be automated. Whereas if you switch to the model of it's more like real life, you're going to have people at all different levels. I guess the question is how many people are hooked on being the hero and are they still going to find it interesting if they're not? Um, but if there's ever a group of people that understand gameplay and how to make this work, it is this group. They're sitting on decades of data now, right? I mean it's just insane how far they've gone

Speaker C: and it's interesting because they're including it basically in the subscription. So you have classic, the classic reboots where you can replay the classic versions. You have the retail version which is, you know, whatever, I forget, uh, we're on version 12 point something or other. And now you have forever. So if somebody's not happy with not being the center of the story in forever, they can bounce back to retail. Uh, so it's, it's sort of an interesting product market fit idea of okay, we've got something. If you honest, if you want to relive nostalgia, there's classic. If you want to just be in an ever changing living world, there's forever. And if you want to be the hero of a story and have a campaign progression, there is the retail version. So it's, it's kind of interesting to see how Blizzard is tackling this. And obviously being owned by Microsoft, they have the compute resources to throw at it to allow for these different things to happen.

Speaker B: Also in the gaming front, watching gta, trying to make the Christmas deadline here, it looks like they're going to do that. But that's, man, that is one workplace I don't want to have anything to do with right now. There must be some folks pounding the uh, Pepto Bismo and Toms right now because they've got a lot going on. We'll see how that comes through. Uh, anything else? Of course. We've got props coming up. Next week I'm going to be talking to Amanda Natividad. She's got her new book, Zero Click Marketing. I'm going to get some signed copies out of that. I have to order that stuff. I am like right up against the line on logistics. This is coming up Monday, behind the eight Ball.

Speaker C: Are you interviewing her at the show?

Speaker B: Yeah, yeah, we're going to talk on uh, Tuesday after lunch I believe is what we've got scheduled, so.

Speaker C: Oh, nice.

Speaker B: Um, yeah, I've tried to line that up after so that folks can tag along and check it out because they've got a morning break in there that we can plug in. How about, are you going Tuesday morning or Tuesday afternoon?

Speaker C: Yeah, my session is Tuesday after lunch.

Speaker B: Oh, okay. So right after. So after I talk to her will be your session.

Speaker C: Yeah.

Speaker B: Um, and that's election day, so get your votes in

Speaker C: on the gear. Watch front. I ordered an Akaso V50 action cameras. Basically it's a knockoff DJI, uh, you know, GoPro style camera. Uh, I'll have it with me, uh, at Profs. I'm going to Test it out and see if it's any good, uh, for, like, not, like, you know, action. I'm not taking biking or climbing up a mountain with it, but, um, I've been wanting, uh, to try a camera out like this for like, vlogging and stuff while I travel. But I also don't want to spend 500 bucks on it. So, uh, we'll. We'll. By the time we're back from props, we'll have a review as to whether it's. Whether it's worth $139 or not.

Speaker B: And is it kind of a gimbal mount? Does it have its own stabilization or what's the.

Speaker C: It's got electronic stabilization, plus it has all the standard GoPro style connectors, so I can connect it to all my existing hardware.

Speaker B: Oh, that's interesting. Okay, well, yeah, we'll have some live shots on that and we'll see that in the field. All right. Yeah. If you're going to be at Profs, be sure to give us a yell so we can catch up with you there. But that's going to do it for this week. So until next week. Enjoy the coffee.

Speaker C: Enjoy the coffee.

Speaker A: You've been listening to Marketing over, uh, coffee. Christopher Penn blogs@christopherspenn.com Read more from John J. Wall at jw5150.com the marketing over coffee theme song is called Melo G by Funkmasters and you can find it at Musically from Mevio or follow the link in our show notes.