Best of LinkedIn: MarTech Insights CW 48/ 49

00:00:00: This deep dive is provided by Thomas Allgeier and Frennis, based on the most relevant LinkedIn posts about MarTech in calendar weeks, forty-eight and forty-nine.

00:00:09: Frennis is a B to B market research company helping enterprises gain the market, customer, and competitive insights needed to drive growth and success.

00:00:19: Welcome back.

00:00:20: We're here to cut through the noise on BDB.

00:00:21: Martek and the last two weeks of source material have been fascinating.

00:00:25: They really have.

00:00:26: There's this huge tension right now in the conversation.

00:00:29: On one hand, you have this almost breathless race towards a fully agentic AI-powered staff.

00:00:36: The vendors are making these huge announcements.

00:00:38: Exactly.

00:00:39: Unified platforms, automated orchestration.

00:00:41: It's all about the future, you know, speed and integration.

00:00:43: But then you look at what practitioners are actually posting and it's... It's a completely different story.

00:00:48: It's almost painful.

00:00:49: It is.

00:00:50: They're on the ground, quietly fixing funnels, cleaning up just massive data leaks.

00:00:54: It's all the boring foundational work.

00:00:57: The work that actually makes the exciting stuff possible.

00:01:00: That is the paradox right there.

00:01:02: Rashida Padacharya put it really well.

00:01:03: She said, Martek is what we need to scale and stop, depending on sheer luck for conversions.

00:01:09: But if you try to scale dysfunction with these new autonomous agents,

00:01:12: you just get chaos.

00:01:14: but faster.

00:01:15: So much faster.

00:01:16: And that's a terrifying thought.

00:01:17: It is.

00:01:18: So today, our mission is to dive into these two conflicting trends.

00:01:22: We're going to give you the key clusters that are really defining what the Mar-Tech landscape is going to look like.

00:01:27: Let's start with the shiny stuff.

00:01:28: Theme one, the agentic leap.

00:01:31: OK, let's unpack this.

00:01:32: Because for most of this year, AI in marketing meant content generation.

00:01:37: We saw posts from Australopas, Cuesta, and Gabe Larson confirming that was the number one use case.

00:01:43: Right.

00:01:44: The conversation now is shifting.

00:01:46: It's moving away from just generation towards AI taking actual autonomous action.

00:01:50: That's

00:01:51: the leap from creation to orchestration.

00:01:53: And you can see the big vendors are all over this.

00:01:55: Adobe, for instance, is pushing for what Corey Haldeman and Veronica Bruce called a creative OS.

00:02:00: The creative OS.

00:02:01: Yeah, integrating multiple models to own the entire creative workflow.

00:02:05: It acts like a true operating system.

00:02:08: And we saw Insider rebranding to Insider One, which is all about that single integrated platform.

00:02:13: It's a direct shot at the fragmentation problem.

00:02:16: They're betting that a unified system is the only way to manage this complexity.

00:02:21: But there's a deeper strategic shift happening here, isn't there?

00:02:24: It's not just about better tools.

00:02:26: Not at all.

00:02:26: The future stack isn't about systems that just predict what's going to happen.

00:02:30: That's today's tech.

00:02:32: It's about systems capable of action.

00:02:35: Tobias Knitzer framed this perfectly.

00:02:37: He said, leaders are now seeing AI as a way to reduce outcome uncertainty.

00:02:41: Outcome uncertainty.

00:02:43: That sounds big.

00:02:44: What does that actually mean for a marketer day to day?

00:02:46: Well, it means you stop just forecasting what a customer might do.

00:02:50: Instead, you focus on the causal effect of your actions.

00:02:53: You're using AI to figure out the next step you can take to, you know... guarantee a result, or at least dramatically improve the odds.

00:02:59: So your KPIs shift from just tracking behavior to tracking the effects of your intentional automated actions.

00:03:05: Exactly.

00:03:06: You move from being a passive observer to an active influencer in the pipeline.

00:03:10: Okay.

00:03:10: But if you have multiple agents acting across your stack, that requires a massive amount of coordination, right?

00:03:17: Sure.

00:03:17: How do you stop them from tripping over each other?

00:03:19: And that is the huge architectural challenge.

00:03:22: Florian Delval highlighted this.

00:03:24: Orchestration is becoming the new strategic layer because these agents need to coordinate, they need to share information, manage handoffs.

00:03:32: So how do they... How do they talk to each other across all these different systems?

00:03:36: Well,

00:03:36: that's where you see frameworks like the model context protocol or MCP starting to emerge.

00:03:40: It's basically a standard language that helps these agents find tools and maintain the same context, same state,

00:03:48: you know?

00:03:49: So they're all on the same page.

00:03:50: Exactly.

00:03:51: Without something like MCP, your agents are just working in silos, which totally defeats the point of automation.

00:03:56: And if you connect this idea to paid media, this is where it gets really interesting for the bottom line.

00:04:01: And Neil K. Pandit pointed out, that AI is already reshaping planning and optimization in ad tech way ahead of twenty-twenty-six.

00:04:09: And look at attribution.

00:04:11: Seville Lustinoff's post on this was great.

00:04:13: He said it's not just a scoreboard anymore and not just a backward-looking report for the CFO.

00:04:18: It's becoming a steering wheel.

00:04:19: It's becoming agentic.

00:04:21: It combines signals with small, quick tests, and it tells you what your next move should be in real time.

00:04:26: It's not a history lesson.

00:04:28: Okay, so if attribution is now a real-time steering wheel.

00:04:31: The next question has to be, what is that steering wheel attached to?

00:04:36: Because the foundations for most companies are, let's just say, a bit shaky.

00:04:41: That's

00:04:41: the perfect transition.

00:04:42: Which brings us to theme two, data foundations, CDP architecture, and operational readiness.

00:04:48: Because that's the catch.

00:04:49: It's the whole story.

00:04:51: Paul Duravall said it perfectly.

00:04:52: If your only job is to accelerate execution, AI will just accelerate the mess if your data is flawed.

00:04:58: It's brutally exposing all the inconsistent schemas and data silos that leaders thought they'd fixed years ago.

00:05:04: Yeah,

00:05:05: the sources showed a huge data readiness gap.

00:05:07: Carly Miller detailed the problem so well.

00:05:09: She said personalization gaps aren't about a lack of tools.

00:05:12: It's that the data is, quote, locked in places only IT can reach.

00:05:17: So it's inaccessible.

00:05:18: Or it's just too slow for marketers to use when it matters.

00:05:22: And this is so critical now.

00:05:24: Niels van Mier Johnson gave this really important warning.

00:05:27: AI agents just assume their inputs are reliable.

00:05:30: So if your data isn't semantically correct.

00:05:32: Meaning different fields have the same name but mean totally different things.

00:05:36: Right.

00:05:36: The AI will just scale those errors.

00:05:39: It won't fix them.

00:05:39: You'll just get hyper-personalized bad campaigns sent out at lightning speed.

00:05:44: And as Peter Rogers noted, we need quality data at speed.

00:05:49: So what's the architectural solution here?

00:05:50: You can't just throw out your existing data warehouse.

00:05:53: No, of course not.

00:05:54: And that's why the composable CDP model is getting so much traction.

00:05:57: Logan Woodbridge explained it really well.

00:06:00: It's built on top of your existing data warehouse, your snowflake, or your BigQuery.

00:06:04: So you don't have to do a massive migration.

00:06:06: You don't.

00:06:07: It organizes the data you already have.

00:06:09: It uses what you've already built.

00:06:10: And once it's clean, you have to activate it.

00:06:13: That's where Dan Maga's point about reverse ETL comes in.

00:06:16: You're flipping the warehouse from just being cold storage.

00:06:18: Where

00:06:18: data goes to die.

00:06:20: Exactly.

00:06:20: And turning it into a revenue driver by pushing that clean data back into your marketing and sales tools.

00:06:26: For any company that's serious about agent-driven marketing, this foundation is just.

00:06:31: It's non-negotiable.

00:06:32: San Ramiki and Oscar Lopez Cuesta both agreed.

00:06:35: Centralization, semantic layers, real-time decisioning.

00:06:39: Without those, deploying autonomous agents is frankly irresponsible.

00:06:43: Which sets us up perfectly for our third theme, strategy, operating models, and human agency.

00:06:49: Because having the cleanest data in the world won't save you if your core strategy is broken.

00:06:54: And this might be the most humbling part of the whole conversation.

00:06:57: Mark Staus raised this really shocking paradox.

00:07:00: Yeah.

00:07:00: The evolution of MarTech over the last decade has actually coincided with the most significant decline in go-to-market performance in twenty years.

00:07:07: Wait,

00:07:07: how is that even possible?

00:07:09: With all these incredible tools and all this data, GTM performance went down.

00:07:12: Staus' argument is that the technology didn't fail.

00:07:15: It worked perfectly.

00:07:17: It faithfully scaled the flawed worldview that leaders gave it.

00:07:20: That worldview being the deterministic funnel.

00:07:24: the rigid linear stages.

00:07:26: Exactly.

00:07:26: A model that just doesn't reflect how a B to B buying actually works today.

00:07:31: But the dashboards looked clean, right?

00:07:33: It preserved this illusion of control while the market was becoming chaotic.

00:07:37: The tech just helped you fail faster, but with much prettier reports.

00:07:40: And that kind of flawed strategy leads directly to tool hoarding.

00:07:45: Franz Riemersma made that connection.

00:07:48: A poorly defined strategy always leads to buying more tools to compensate.

00:07:53: I mean, think about that anecdote he shared.

00:07:55: A client with a hundred and thirty-two tools in their stack.

00:07:58: One hundred and thirty-two.

00:07:59: And twelve of them were just for collaboration.

00:08:02: That just screams strategic failure.

00:08:04: It means no one is truly accountable because you haven't standardized anything and you cannot safely automate what you haven't standardized.

00:08:11: So

00:08:11: the focus has to shift from the stack to the system.

00:08:14: Jared Fackle called it a systems problem, not a tool problem.

00:08:18: It's about subtraction, not addition.

00:08:20: And moving from a MarTech step to what Benny Lukas called a marketing intelligence

00:08:24: stack.

00:08:24: Which requires a whole new level of operational maturity.

00:08:27: It does.

00:08:28: Mike Rizzo pointed out that the team of one era in marketing ops is over.

00:08:32: It's shifting to small specialized pods, which means you need crystal clear accountability.

00:08:38: RCI two point oh, lifecycle documentation for every automation.

00:08:42: The high functioning teams are already doing this.

00:08:45: But we also saw that organizational design is still a huge barrier.

00:08:49: Fabio Canalesi noted that CRM is often misplaced in the org chart and just deeply undervalued by senior leadership.

00:08:56: So it's always about governance and alignment.

00:08:59: Always.

00:08:59: And because of that, the profile of a MarTech leader is changing.

00:09:03: Pale Patel said they have to go from being technical experts to being strategic translators and diplomats.

00:09:08: balancing the deep tech knowledge with massive stakeholder alignment.

00:09:12: Yeah, and Kristen Connell said the most valuable human skill now is orchestration.

00:09:16: Your job is to arrange the symphony, not play every single instrument.

00:09:20: Which brings us back to the ultimate goal of AI.

00:09:23: Dave Golden put it so well.

00:09:24: It's not to replace your judgment.

00:09:26: It's to be a thinking partner that helps you ask better questions.

00:09:29: The AI gives you the data.

00:09:30: The marketer provides the curiosity and the critical judgment.

00:09:33: That focus on context?

00:09:35: On judgment?

00:09:37: It brings us to our key takeaway.

00:09:39: And it's a real dose of reality about the limits of all this technology.

00:09:42: It

00:09:42: really is.

00:09:43: While we obsess over collecting all this digital exhaust, the clicks, the page views, we often have zero visibility into a buyer's true internal decision state.

00:09:52: The

00:09:53: difference between knowing what someone did and understanding why they did it.

00:09:57: Ben Scanlon shared this brilliant analogy.

00:10:00: Imagine if a store like Trader Joe's was run entirely on martech logic.

00:10:04: Okay.

00:10:05: So you, the customer, you walk in needing a very specific brand of tomatoes.

00:10:09: Why?

00:10:10: Because your anxious in-laws are coming over and that brand is like a core comfort food for them.

00:10:15: That's the crucial context.

00:10:16: Where's

00:10:16: human motivation?

00:10:17: But the AI, it only has forty seven behavioral data points.

00:10:21: It knows you looked at peanut butter cups for four point seven seconds last week.

00:10:25: So it assumes you have high purchase intent for candy.

00:10:27: So the

00:10:28: agentic AI sends me a coupon for peanut butter cups.

00:10:31: Exactly.

00:10:32: It optimizes for the data it has, which is completely misleading.

00:10:36: Meanwhile, you're desperately trying to find those tomatoes to soothe your in-laws.

00:10:40: Behavioral data alone can never capture that context.

00:10:43: So Martek can make you incredibly efficient, but it can't solve for a lack of customer empathy.

00:10:49: The tools are just multipliers of the human insight you feed them.

00:10:52: Absolutely.

00:10:53: Which brings us to a final, provocative thought for you, the listener.

00:10:57: Thomas Elblum shared a wild statistic.

00:11:00: Fifty percent of consumers are already using AI search.

00:11:03: Fifty percent.

00:11:03: And that's shifting twenty to fifty percent of traffic away from the channels we've been optimizing for years.

00:11:08: The battle for the customer is moving from the SRP to the AI's answer.

00:11:12: Which means visibility in AI search isn't a marketing problem anymore.

00:11:16: It's a data and architecture discipline.

00:11:18: So the question for you is this.

00:11:21: Are your data feeds?

00:11:22: your products backs, your docs, your case studies, are they clean and structured enough to even be mentioned when a customer's agent asked that very first question?

00:11:30: That's the real challenge.

00:11:32: If you enjoyed this deep dive, new episodes drop every two weeks.

00:11:36: Also check out our other editions on account based marketing, field marketing, channel marketing, AI and B to B marketing, go to market and social selling.

00:11:43: Thank you for joining us.

00:11:44: Be sure to subscribe so you don't miss the next one.