1 00:00:00,000 --> 00:00:01,180 - Welcome to Five Minutes Product. 2 00:00:01,180 --> 00:00:03,240 Today is May 23rd, 2025. 3 00:00:03,240 --> 00:00:04,800 We're digging into some sources today 4 00:00:04,800 --> 00:00:06,660 that really got us thinking about 5 00:00:06,660 --> 00:00:08,800 where product and marketing are actually heading. 6 00:00:08,800 --> 00:00:09,880 - Yeah, there's a lot of talk, right? 7 00:00:09,880 --> 00:00:13,120 Big shifts with AI, but are teams really leveraging it 8 00:00:13,120 --> 00:00:14,960 or is it more like tweaking around the edges? 9 00:00:14,960 --> 00:00:17,080 - Exactly, and then there's this whole other thing, 10 00:00:17,080 --> 00:00:19,400 vibe marketing, what even is that? 11 00:00:19,400 --> 00:00:20,240 - We'll get into it. 12 00:00:20,240 --> 00:00:21,880 - Okay, but before we dive deep, 13 00:00:21,880 --> 00:00:23,840 here's a little question for you listening. 14 00:00:23,840 --> 00:00:27,480 Before he shifted focus towards media and AI, 15 00:00:27,480 --> 00:00:30,720 what was Dorkesh Patel's main field of study? 16 00:00:30,720 --> 00:00:33,240 Keep that rattling around as we explore. 17 00:00:33,240 --> 00:00:34,720 - So the big picture first, 18 00:00:34,720 --> 00:00:37,240 it's easy to get caught up in the hype, right? 19 00:00:37,240 --> 00:00:40,480 CEOs making grand pronouncements, these AI manifestos. 20 00:00:40,480 --> 00:00:41,880 - Oh yeah, you see those everywhere. 21 00:00:41,880 --> 00:00:44,080 - But what our sources are pointing out is this. 22 00:00:44,080 --> 00:00:46,480 Well, this gap, teams on the ground. 23 00:00:46,540 --> 00:00:49,660 They are using AI, sure, but often it's pretty incremental. 24 00:00:49,660 --> 00:00:52,020 - Incremental how, like small fixes. 25 00:00:52,020 --> 00:00:54,420 - Exactly, little improvements to the workflows 26 00:00:54,420 --> 00:00:57,180 they already have, optimizing things here and there. 27 00:00:57,180 --> 00:00:59,500 - Okay, so just making the old way slightly better, 28 00:00:59,500 --> 00:01:02,140 not really inventing a new way with AI. 29 00:01:02,140 --> 00:01:04,220 - That's the core issue highlighted. 30 00:01:04,220 --> 00:01:06,700 The sources argue the real opportunity, 31 00:01:06,700 --> 00:01:08,980 the big step change innovation, 32 00:01:08,980 --> 00:01:12,140 comes from truly reimagining processes, 33 00:01:12,140 --> 00:01:14,620 building stuff that's AI native from the start. 34 00:01:14,620 --> 00:01:16,420 - Replacing the old stuff entirely. 35 00:01:16,420 --> 00:01:19,860 - Right, that's apparently where the massive value unlock is. 36 00:01:19,860 --> 00:01:22,180 - It makes sense, it's just probably harder, 37 00:01:22,180 --> 00:01:24,900 feels riskier than just adding a bit of AI here and there. 38 00:01:24,900 --> 00:01:27,260 - It definitely requires more boldness 39 00:01:27,260 --> 00:01:29,340 to kind of discard what's familiar. 40 00:01:29,340 --> 00:01:33,300 Incremental feels safer, less disruptive internally, maybe. 41 00:01:33,300 --> 00:01:36,470 but the sources suggest it just won't deliver 42 00:01:36,470 --> 00:01:40,290 those exponential gains AI really promises. 43 00:01:40,290 --> 00:01:41,910 - And speaking of bold re-imagining, 44 00:01:41,910 --> 00:01:43,890 there's actually a great example on the sources, 45 00:01:43,890 --> 00:01:46,970 really concrete, yeah, the story about a team, 46 00:01:46,970 --> 00:01:49,410 they were really frustrated with their user activation rates 47 00:01:49,410 --> 00:01:51,130 just stuck, flatline. 48 00:01:51,130 --> 00:01:51,970 - Been there. 49 00:01:51,970 --> 00:01:53,450 - So they looked at their normal onboarding, 50 00:01:53,450 --> 00:01:55,210 the usual flow, and basically decided, 51 00:01:55,210 --> 00:01:58,490 this isn't just inefficient, it's the wrong approach. 52 00:01:58,490 --> 00:01:59,690 - Interesting, so what did they do? 53 00:01:59,690 --> 00:02:01,590 Not just A/B test button colors. 54 00:02:01,590 --> 00:02:03,350 - No, no, they went bigger. 55 00:02:03,350 --> 00:02:05,410 They replaced the whole traditional flow 56 00:02:05,410 --> 00:02:08,210 with an AI-powered user onboarding agent. 57 00:02:08,210 --> 00:02:10,930 - An AI agent for onboarding, how did that work? 58 00:02:10,930 --> 00:02:13,090 - Well, the idea was it could interact 59 00:02:13,090 --> 00:02:15,450 much more personally, adapt to the user, 60 00:02:15,450 --> 00:02:18,030 handle tons of users at once, obviously, 61 00:02:18,030 --> 00:02:19,930 and just feel more tailored. 62 00:02:19,930 --> 00:02:22,450 than a one size fits all static flow. 63 00:02:22,450 --> 00:02:25,410 - Okay, I can see the potential there and the result. 64 00:02:25,410 --> 00:02:26,390 - That's the kicker. 65 00:02:26,390 --> 00:02:28,570 It reportedly doubled their activation rate. 66 00:02:28,570 --> 00:02:30,950 - Doubled, wow, okay, that's not incremental. 67 00:02:30,950 --> 00:02:33,050 - Right, that's a proper step change, 68 00:02:33,050 --> 00:02:35,050 rethinking the whole thing around AI. 69 00:02:35,050 --> 00:02:36,610 - That really speaks volumes. 70 00:02:36,610 --> 00:02:39,490 Okay, so switching gears a bit, marketing's also changing, 71 00:02:39,490 --> 00:02:41,050 this Vibe marketing thing you mentioned. 72 00:02:41,050 --> 00:02:42,330 - Yeah, the sources bring this up. 73 00:02:42,330 --> 00:02:46,010 It sounds quite different from the usual demographic 74 00:02:46,010 --> 00:02:48,370 targeting age, location, that sort of thing. 75 00:02:48,370 --> 00:02:52,210 - It definitely moves beyond those dryer data points. 76 00:02:52,210 --> 00:02:55,290 Vibe marketing, it's about connecting emotionally, 77 00:02:55,290 --> 00:02:58,410 shared feelings, values, identity. 78 00:02:58,410 --> 00:03:00,430 - So less about who you are on paper 79 00:03:00,430 --> 00:03:03,730 and more about the community you feel part of, 80 00:03:03,730 --> 00:03:04,570 how you feel. 81 00:03:04,570 --> 00:03:06,170 - Exactly, it's about resonating. 82 00:03:06,180 --> 00:03:13,380 that level. Interesting. But how do you like create or scale a vibe? Does AI come back into play here? 83 00:03:13,380 --> 00:03:18,820 It seems so. The sources suggest it's a mix. Human intuition is key understanding and pinpointing 84 00:03:18,820 --> 00:03:24,500 that desired emotional connection, the vibe. Okay. And then AI comes in for the execution, 85 00:03:24,500 --> 00:03:29,780 rapidly creating and deploying experiences that actually embody that feeling and doing it at 86 00:03:29,780 --> 00:03:35,940 scale. So humans define the feeling, AI makes it happen fast and wide. Pretty much. And apparently 87 00:03:35,940 --> 00:03:39,780 challenger brands are really leaning into this, trying to build that deeper, more authentic 88 00:03:39,780 --> 00:03:45,140 engagement because, well, it feels less like you're just a target in a demographic bucket. 89 00:03:45,140 --> 00:03:52,420 That's a sharp insight. Authenticity through emotion scaled by AI. Yeah. Cool. Beyond those 90 00:03:52,420 --> 00:03:52,580 big. 91 00:03:52,720 --> 00:03:55,440 shifts there were some other useful nuggets in the sources too, right? 92 00:03:55,440 --> 00:03:59,380 Something about A/B testing. Yeah that caught my eye too. The point that A/B test 93 00:03:59,380 --> 00:04:04,320 success rates are often, well, pretty low. Lower than people think. Seems so. Yeah. 94 00:04:04,320 --> 00:04:07,840 And the source even says tests can sort of lie if you're not careful how you 95 00:04:07,840 --> 00:04:11,360 interpret them. The real focus shouldn't just be the win rate anyway. What should 96 00:04:11,360 --> 00:04:16,000 it be then? Learning. It's about the learning you get from the whole process. 97 00:04:16,000 --> 00:04:21,840 Understanding the ROI of testing, yes, but crucially using insights from past 98 00:04:21,840 --> 00:04:27,280 tests to design better future ones. Ah, so testing as a learning engine not just a 99 00:04:27,280 --> 00:04:31,040 feature-picking machine. Exactly, it's a mechanism for learning. Makes sense. And 100 00:04:31,040 --> 00:04:35,200 one last quick point, they mentioned framing ideas. Oh right, even the best 101 00:04:35,200 --> 00:04:40,600 ideas can flock if they aren't presented well. Yeah, what's the key there? 102 00:04:40,600 --> 00:04:42,710 The source emphasizes focusing on how your ideas 103 00:04:42,710 --> 00:04:44,910 solves your stakeholders' problems, 104 00:04:44,910 --> 00:04:47,290 not just why your idea is great in isolation 105 00:04:47,290 --> 00:04:48,990 or why other options are bad. 106 00:04:48,990 --> 00:04:50,810 - So frame it around their needs. 107 00:04:50,810 --> 00:04:52,850 - Right, frame alternatives positively, 108 00:04:52,850 --> 00:04:55,930 show the trade-offs clearly, build confidence, 109 00:04:55,930 --> 00:04:57,610 don't just pick fights. 110 00:04:57,610 --> 00:04:59,230 - Some really practical stuff there too. 111 00:04:59,230 --> 00:05:00,390 Okay, quite a tour. 112 00:05:00,390 --> 00:05:02,230 So remember that quiz question from the start 113 00:05:02,230 --> 00:05:04,970 about Dorkesh Patel's background before media and AI? 114 00:05:04,970 --> 00:05:07,410 - Mm-hmm, his path is pretty interesting, 115 00:05:07,410 --> 00:05:09,830 especially thinking about tech's broader impact, 116 00:05:09,830 --> 00:05:11,130 which we've been talking about. 117 00:05:11,130 --> 00:05:13,670 - Definitely, and hey, if you found this deep dive valuable, 118 00:05:13,670 --> 00:05:15,750 the easiest way to support us is just to rate, 119 00:05:15,750 --> 00:05:17,670 like, and subscribe wherever you're listening. 120 00:05:17,670 --> 00:05:19,870 It costs nothing, but it helps us a lot. 121 00:05:19,870 --> 00:05:20,790 - Yeah, it really does. 122 00:05:20,790 --> 00:05:22,970 And you can also keep the conversation going 123 00:05:22,970 --> 00:05:25,590 with us and others in our Discord community. 124 00:05:25,810 --> 00:05:32,450 do. Okay, the moment of truth. The answer to our quiz question, Dorkesh Patel's primary field of 125 00:05:32,450 --> 00:05:38,530 study before his current focus was computer science. Ah, computer science. Yeah, makes sense. 126 00:05:38,530 --> 00:05:44,130 Seeing that technical foundation connect to these bigger discussions about product marketing, AI, 127 00:05:44,130 --> 00:05:48,530 it really shows how linked these fields are becoming. It really does. And you know, 128 00:05:48,530 --> 00:05:53,490 thinking about all this, the potential for truly reimagining things with AI, connecting through 129 00:05:53,490 --> 00:05:58,290 Vibe, maybe the biggest challenge isn't just learning the new tool. What is it then? Maybe 130 00:05:58,290 --> 00:06:03,330 it's just finding the boldness, the courage to actually reinvent a whole experience, whether it's 131 00:06:03,330 --> 00:06:07,250 product or marketing, from the ground up. So maybe something for you to think about. 132 00:06:07,250 --> 00:06:12,130 What's one workflow, just one, that you could try to totally reimagine?