Best of LinkedIn: MarTech Insights CW 46/ 47

00:00:00: This episode is provided by Thomas Allgaier and Frennis, based on the most relevant LinkedIn posts about MarTech in calendar weeks, forty-six and forty-seven.

00:00:08: Frennis is a B to B market research company helping enterprises gain the market, customer, and competitive insights needed to drive growth and success.

00:00:17: Welcome back.

00:00:18: So for today's deep dive, our mission is pretty sharp.

00:00:22: We're doing a high speed distillation of the most crucial insights from the MarTech community on LinkedIn.

00:00:27: We're looking specifically at calendar weeks, forty six and forty seven.

00:00:31: And we're really going to drill down into operational excellence, the, you know, the accelerating pace of AI adoption and how everyone is grappling with performance measurement right now.

00:00:39: Exactly.

00:00:39: And this is in theory.

00:00:40: We're synthesizing what the industry leaders and the, you know, the on the ground practitioners are actually talking about.

00:00:45: And we saw four big themes just dominating the conversation.

00:00:48: First, there's the evolution of AI from just being a tool to, well, a foundational teammate.

00:00:53: Then there's this urgent need to fix data foundations.

00:00:56: After that, the classic roadblock organizational readiness.

00:00:59: And finally, what measurable performance even looks like now.

00:01:02: Okay, let's jump right in there.

00:01:04: Theme one, AI is becoming a core operating layer.

00:01:08: It's moving so fast from being just a cool feature to, you know, genuine infrastructure.

00:01:13: The vocabulary is even changing.

00:01:14: People are calling it a teammate.

00:01:16: Yeah.

00:01:16: And what's fascinating is seeing how the big DXP players, the digital experience platforms are reacting.

00:01:22: Fable Clarek pointed out that Adobe, for example, isn't just bolting on AI.

00:01:26: They're leading with custom models like Firefly plus this, this thing he calls an agentic orchestration layer.

00:01:33: Okay.

00:01:34: So when you say agentic orchestration layer, what does that actually mean in practice?

00:01:38: We're talking about AI agents that can do more than just one task, right?

00:01:41: They can string together multiple actions, make decisions, basically run a whole campaign on their

00:01:45: own.

00:01:46: Precisely.

00:01:47: They're becoming autonomous decision makers within the system, not just, you know, button pushers.

00:01:51: And we saw this huge concrete signal of this shift with the rumored Adobe acquisition of Simrush.

00:01:57: Right,

00:01:58: both Veronica Bruce and Jordy Ms.

00:01:59: Farney flagged that one.

00:02:00: They did.

00:02:01: And this is not a classic SEO play.

00:02:03: This is about securing your brand's visibility and voice in a world where search is AI first.

00:02:10: And the data on that is just, it's staggering.

00:02:13: I saw one insight showing traffic from generative AI sources to US retail sites jumped twelve hundred percent year over year.

00:02:22: I mean, think about that, the twelve hundred percent change.

00:02:24: That means the old SEO playbook, you know, the ten blue links on a page, it's fundamentally broken, or at least it's radically different.

00:02:31: It's

00:02:31: the writing on the wall.

00:02:33: If AI is taking over that research layer of the buyer's journey, then while you need to be able to influence that generative output, but it's not just the giants playing Acquisition Chess, AI is also getting incredibly tactical.

00:02:45: I thought the workflow examples were amazing.

00:02:47: Michelle Lieben reported that Claude the son at four point five can now build advanced any and workflows from just as single prompt.

00:02:53: And NEN is a pretty powerful automation platform.

00:02:56: It is.

00:02:57: And the fact that the LLM can handle what, ninety percent of the complexity, including the data flow, the logic, even the error handling, that lets marketers just skip the whole TV's build phase.

00:03:06: That frees up so much time.

00:03:08: And Justin Norris give another great example of AI tackling what he called the messy middle zone.

00:03:13: All those little paper cut tasks that just bleed your ops team dry.

00:03:17: What was the example?

00:03:18: They built an LLM solution to intercept support tickets.

00:03:22: It could diagnose and even fix issues like opportunities not showing up in sales reports.

00:03:27: And the kicker on that was how fast it was to build.

00:03:31: He said it only took about an hour per scenario to get it built, tested, and deployed.

00:03:36: That is a serious operational multiplier.

00:03:38: It is.

00:03:39: But none of that amazing automation works if the foundation is shaky.

00:03:44: Which brings us perfectly to our second theme, data foundations, CDP maturity, and architectural clarity.

00:03:49: The

00:03:50: classic problem.

00:03:51: You can't run advanced AI on messy fragmented data.

00:03:53: Exactly.

00:03:54: The consensus is clear.

00:03:55: Data governance and architecture are the new competitive battlegrounds.

00:03:59: And it seems like the conversation is evolving too.

00:04:01: It's not just about getting a single customer view anymore.

00:04:04: Now it's about needing a single meaning of data.

00:04:07: That's a much more profound idea.

00:04:09: Matthew Lee really drove that point home talking about the Salesforce and Informatica rumors.

00:04:14: The strategic goal there isn't just about mashing two databases together.

00:04:18: It's about building a unified metadata backbone for AI.

00:04:22: So that concepts like active customer or qualified lead mean the exact same thing everywhere.

00:04:28: Yes.

00:04:29: Think about it.

00:04:30: If your marketing automation system defines active one way and your sales CRM defines it another, What does your AI agent do when you tell it to target active customers?

00:04:41: It makes a mess.

00:04:42: The decisions become untrustworthy.

00:04:44: Right.

00:04:44: And this is why CDPs have evolved so quickly.

00:04:47: Shavalka Trashtha was stressing that for financial services firms especially, CDPs are capability enablers.

00:04:54: You can't start by looking at vendor slides.

00:04:55: You have to start with a strategic use case like personalization or churn reduction and then work backward.

00:05:00: That focus on the use case is so critical and it all rests on foundational best practices.

00:05:06: shared some really tactical advice for implementing tools like the Adobe Web SDK.

00:05:10: What was the

00:05:10: key takeaway?

00:05:11: That it has to be treated as the central source of truth.

00:05:15: That means a clean, structured data layer, centralized mapping.

00:05:19: And the most crucial part, your data model has to be defined before you start setting up any rules or tracking tags.

00:05:26: It always comes back to the data model.

00:05:28: But even with all this focus on clean, centralized data, there's still this huge activation gap.

00:05:35: Logan Woodbridge pointed this out.

00:05:36: I saw that.

00:05:37: So a company invests in a data warehouse.

00:05:40: They get these beautiful unified reports.

00:05:42: Great for the dashboards.

00:05:44: But when marketing wants to actually activate a niche segment like high value abandoned cart users, they

00:05:50: have to submit a ticket to the data team and wait two weeks.

00:05:53: Exactly.

00:05:54: The data is clean, but it's locked in the reporting layer, not available to the activation layer.

00:05:59: And if you can't use the data without that kind of friction, the whole investment just fizzles out.

00:06:04: Which of course moves us right into theme three.

00:06:06: Operational excellence and organizational readiness.

00:06:09: Because that friction isn't a tech problem, it's a people problem.

00:06:12: It's the ops problem.

00:06:13: Right.

00:06:14: Vanessa Budak stated it so clearly.

00:06:16: She said most marketing teams have an operations problem, not a strategy problem.

00:06:21: And she estimated they lose ten to twenty hours a week just making systems talk to each other.

00:06:25: A quarter to half of someone's work week.

00:06:27: Just gone.

00:06:29: It's exhausting.

00:06:30: And Mike Rizzo reinforced this.

00:06:31: He argued that big change initiatives always fail when the underlying systems and processes aren't ready.

00:06:37: So marketing ops ends up right at the center of every single transformation.

00:06:42: They're the gatekeepers.

00:06:43: They are.

00:06:44: And that pressure often leads to what Ariel Afrid observed, which is companies buying shiny new tech to avoid fixing the internal

00:06:50: rot.

00:06:51: Ah, the shiny new tool trap.

00:06:53: Exactly.

00:06:53: She found that ninety-nine percent of marketers admit they leave key features of their current stack unused.

00:06:59: Optimizing what you already have is almost always the best move.

00:07:02: But that's the hard work, right?

00:07:03: It's easier to get a budget for new software.

00:07:05: So much easier.

00:07:07: Adrian Questor had this piece of... dark humor about it.

00:07:11: He suggests in companies sometimes buy expensive martech specifically to avoid solving the underlying problems, like poor integration or a disconnected data team.

00:07:20: And then, surprise, the licenses get abandoned.

00:07:23: It's the ultimate symptom of organizational dysfunction.

00:07:27: And to get out of that, system unity has to be foundational.

00:07:31: Syracuse Tech found why PNY's advice was to stop treating things like Salesforce marketing cloud and sales cloud as separate products.

00:07:39: Integration isn't an option.

00:07:40: It's the whole point.

00:07:41: Right.

00:07:41: And that was the big takeaway from the MarTech Summit too.

00:07:44: Max and Pruska summarized it as focus on revenue teamwork, clear SLAs and process before you implement technology, process before tech.

00:07:53: And beyond system unity, there's also this need for conceptual unity.

00:07:56: Dr.

00:07:57: Celia Dones highlighted this risk of interpretive drift.

00:08:00: Okay,

00:08:00: what's that?

00:08:00: It's when different teams are looking at the exact same perfectly accurate data, but they reach totally incompatible conclusions.

00:08:06: Like marketing sees a super efficient campaign, a green light.

00:08:09: While the CX team sees a huge drop in customer trust because the personalization felt creepy, a big red light, and both teams are right based on their own metrics.

00:08:19: Exactly.

00:08:19: So calibration, building shared meaning, That's now an essential capability, especially with AI, which can just amplify those misaligned decisions at incredible speed.

00:08:29: Okay, so once you have the solid data, the unified systems and the aligned organization, then you can finally talk about theme four, performance, personalization, and finding that balance between creativity and data.

00:08:40: And

00:08:41: we saw some great evidence that targeted data backed fixes can yield massive returns.

00:08:45: Jessica Bayafi showed a story where she drove a sixty four percent month over month increase in leads for a client.

00:08:52: And it wasn't some revolutionary new tech, was it?

00:08:54: No, it was strategic.

00:08:55: It involved narrowing the targeting on Google Ads and, interestingly, shifting forty percent of the PMAX budget back into focused search campaigns.

00:09:02: A real win for human governance and precision over just broad automation.

00:09:06: But that brings up measurement and, well, the great attribution debate.

00:09:10: It always comes back to attribution.

00:09:12: And Brian DiDrea, I think, summarized it perfectly.

00:09:15: He said, look, both sides have a point.

00:09:18: The pro-attribution RevOps folks and the skeptical brand marketers.

00:09:23: both are right

00:09:24: sometimes.

00:09:25: So, his value is totally contextual.

00:09:27: Completely.

00:09:27: It depends entirely on your company's buying journey.

00:09:30: For some, it's vital.

00:09:31: For others, it's just noise.

00:09:33: Treating it as a universal law is the mistake.

00:09:36: And this complexity, this pressure for numbers, it can lead into some murky ethical territory.

00:09:42: Diacosta brought this up.

00:09:43: Yeah, she highlighted the existence of these gray area tactics, things like suspicious personalization or these black box intense scores being used to chase really marginal gains.

00:09:53: Her point was that ethics and brand protection shouldn't be seen as a burden.

00:09:57: Right, they should be a competitive advantage, especially as customers get smarter about how their data is being used.

00:10:03: And that ties directly into what we heard from the summits about the human AI partnership.

00:10:06: Was the consensus there?

00:10:08: Ed Kelly noted that the future is AI as the executor.

00:10:12: But human intuition and creativity are still essential.

00:10:15: The human job is shifting to creative design and ethical oversight.

00:10:26: totally redefined success metrics.

00:10:29: Yes, you're not just measuring the speed of output, but the quality and the ethical defensibility of that output.

00:10:35: And Phil Gaumache really nailed it.

00:10:36: He said AI can deliver assets ten times faster, but human governance is still required to protect the meaning and quality.

00:10:43: We're becoming stewards of brand integrity in an automated

00:10:46: world.

00:10:46: That was a tremendous deep dive.

00:10:48: The through line is so clear.

00:10:49: AI is accelerating everything, but only the companies with robust data foundations and serious organizational lines.

00:10:56: are going to actually capture that value.

00:10:58: Before we wrap, I want to leave you with one final thought.

00:11:00: It's a really powerful analogy that Keanu Taylor shared at the Martek World Forum.

00:11:05: He argued that the old cliche, building the plane while flying it, is actually wrong for Martek transformation.

00:11:10: Okay, I'm intrigued.

00:11:11: So what is it instead?

00:11:13: He described it as the ship of Theseus.

00:11:15: You don't rebuild your entire stack in one go.

00:11:18: You can't just stop operations.

00:11:20: Instead, you're continuously replacing one plank at a time, one system, one process, one data layer.

00:11:26: And that analogy really that the real strategic challenge isn't the initial purchase.

00:11:31: It's the continuous evolutionary organizational ability to keep replacing those planks.

00:11:36: That's what dictates long-term success.

00:11:38: A great reminder that the challenge is always evolution and alignment, not just installation.

00:11:44: If you enjoy this deep dive, new episodes drop every two weeks.

00:11:47: Also check out our other additions on account-based marketing, field marketing, channel marketing.

00:11:52: AI and B to B marketing go to market and social selling.

00:11:55: Thank you for joining us and make sure you subscribe so you don't miss our next deep dive.