Best of LinkedIn: MarTech Insights CW 04/ 05

00:00:00: This episode is provided by Thomas Allgaier and Frennis, based on the most relevant LinkedIn posts about MarTech from Gallender Weeks four and five.

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.

00:00:16: Welcome back to another deep dive.

00:00:19: Today, we're really getting under the hood of, well... the engine of modern business.

00:00:24: Marketing technology, we've been combing through LinkedIn from late January and early February to find the signals that actually matter.

00:00:33: Looking at the sources we have, the whole vibe feels different.

00:00:36: For so long, Martek felt like this, I don't know, this frantic shopping spree.

00:00:41: It was all about buying the next shiny tool and just hoping for magic.

00:00:45: Yeah,

00:00:45: swiping the corporate card.

00:00:46: Exactly.

00:00:47: But now it feels like the bill has finally come due.

00:00:49: That's

00:00:49: a perfect way to put it.

00:00:50: We're seeing this big shift away from what you could call accumulation, where more tools meant you were winning to an era of system level clarity.

00:00:59: System

00:00:59: level clarity.

00:01:00: I like that.

00:01:00: People

00:01:01: are waking up to the fact that Just adding another logo to your MarTech slide doesn't solve a business problem.

00:01:07: In fact, a lot of the time it just creates a new one.

00:01:10: Right.

00:01:10: It feels like this clash of civilizations.

00:01:12: You have these messy legacy stacks on one side and on the other you have these new, you know, agentic AI models knocking on the door and they're demanding a level of data hygiene that most companies just don't have.

00:01:25: They really don't.

00:01:27: And that brings us perfectly to our first big theme, which we're calling The Great Cleanup.

00:01:32: The Great Cleanup.

00:01:33: I love it.

00:01:34: The party's over.

00:01:35: Time to sweep the floor.

00:01:36: Well, the more is better, philosophy is officially dead.

00:01:39: Companies are just drowning in software, and they're finally starting to audit for value, not just for coverage.

00:01:45: And that distinction is huge.

00:01:47: Let's make this real, though.

00:01:48: Noomanian shared a story that, honestly, it made my jaw drop.

00:01:52: He was auditing a stack that had ballooned to nineteen different tools.

00:01:57: Nineteen.

00:01:58: Just think about that for a second.

00:01:59: It's a nightmare.

00:02:00: And we're not talking about little chrome extensions here,

00:02:02: are we?

00:02:03: No, no, these are the heavy hitters.

00:02:04: Salesforce, Marketo, Six Sense, Enterprise Grade

00:02:08: stuff.

00:02:09: But the problem

00:02:09: wasn't the quality of the tools, it was the redundancy.

00:02:13: When he looked closer, he found they had three separate tools doing lead scoring.

00:02:17: Wait, three tools scoring the same leads?

00:02:19: Three

00:02:19: of them.

00:02:20: So let me guess, none of them agreed on who was a good prospect.

00:02:23: You've got it.

00:02:24: It's precisely the problem.

00:02:26: Imagine being a sales rep.

00:02:27: One dashboard says hot prospect, another says cold.

00:02:31: Marketing is looking at a third.

00:02:33: You get total misalignment.

00:02:35: It's

00:02:35: just paralysis by analysis.

00:02:36: It is.

00:02:37: And on top of that, they had two conflicting lead routing tools, literally fighting over where to send the data.

00:02:44: So the cleanup wasn't just optional.

00:02:46: He's now in the process of taking that stack from nineteen tools.

00:02:49: Down to nine.

00:02:50: So

00:02:50: cutting it in half.

00:02:51: Pretty much.

00:02:52: And yeah, the CFO is probably happy about the licensing fees, but the real win is operational.

00:02:57: When your tools are fighting, your teams start fighting.

00:02:59: That makes so much sense.

00:03:01: It reminds me of a post from Barry Rodriguez.

00:03:03: He uses this great phrase for what startups accumulate.

00:03:06: They called it digital dust.

00:03:08: Digital dust.

00:03:08: That's good.

00:03:09: Isn't it?

00:03:10: It's all those tools you bought three years ago for one campaign used for a week and then forgot about it.

00:03:16: But it's still there.

00:03:17: Siphony data costing money.

00:03:19: Rodriguez basically says the mandate for twenty.

00:03:21: twenty six is to clear that clutter because more rarely means faster.

00:03:26: It just means more friction.

00:03:28: But it's hard to let go, isn't it?

00:03:30: It's that classic sunk cost fallacy.

00:03:33: You know, we spent six months implementing this thing.

00:03:35: We can't just turn it off now.

00:03:36: That's

00:03:37: the trap.

00:03:37: But Friends Ramersma offered a really sharp framework for this.

00:03:41: He talks about optimizing for cash versus optimizing for coverage.

00:03:45: Okay, cash versus coverage.

00:03:46: Break that down for us.

00:03:47: So coverage is fear-based.

00:03:49: It's buying a tool for every single hypothetical edge case.

00:03:53: What if a customer does this one thing on a Tuesday?

00:03:56: Ramers might usually call it stalking.

00:03:58: Stalking.

00:03:59: That's a strong word.

00:04:00: It

00:04:00: is, but it fits.

00:04:01: You know, you're trying to track everything just in case.

00:04:03: He says you should stop building for the what if and start building for cash.

00:04:07: Which means revenue, margin, and risk.

00:04:09: Exactly.

00:04:10: If a tool isn't driving one of those three things, it's just noise.

00:04:13: I love that clarity.

00:04:15: But how do you stop the bloat from just creeping back in?

00:04:18: Because we've all seen it happen.

00:04:20: You need rules.

00:04:21: Constraints.

00:04:22: Yogida Wadwa had some great practical advice here.

00:04:26: She suggested a system of record rule.

00:04:28: How's that work?

00:04:29: Simple.

00:04:30: Any new tool you want to deploy must integrate bidirectionally with your CRM or your main system of record.

00:04:36: So

00:04:37: if the data can't flow back and forth automatically, it's a no-go.

00:04:40: Exactly.

00:04:40: No more data islands.

00:04:42: That's smart.

00:04:42: And she also had a one-in, one-out rule.

00:04:44: She did.

00:04:45: A classic for a reason.

00:04:46: Yeah.

00:04:47: You want a new tool, you have to retire an old one.

00:04:49: It forces you to actually prioritize.

00:04:51: There's

00:04:51: another tension here, though.

00:04:53: Billy Bohan Chanique described it as the laboratory versus the factory.

00:04:57: Mmm.

00:04:58: This is such a crucial mental model for right now.

00:05:00: The factory is your revenue engine, your core CRM, your billing.

00:05:05: It needs to be stable.

00:05:07: It needs to be boring.

00:05:08: Just

00:05:08: have to work.

00:05:09: Every single day.

00:05:10: But the laboratory is where you test things, where you break things, where you experiment with new AI agents.

00:05:16: The problem is, most companies are trying to run both in the same place.

00:05:20: They're running chemical experiments on the assembly line.

00:05:22: Exactly.

00:05:23: And then they wonder why it breaks.

00:05:24: You need to separate them, keep the factory clean, and only move things from the lab to the factory when they're proven.

00:05:30: That transition from lab to factory leads us perfectly into our next scheme.

00:05:35: We can't talk about twenty twenty six without talking about AI.

00:05:38: But the conversation feels different now.

00:05:40: It's moved to, what are they calling it, agentic AI?

00:05:44: It's the buzzword of the moment, but this one has substance.

00:05:47: We're moving from AI as a chatbot you talk to to AI as an agent that does things.

00:05:52: It operates within your systems.

00:05:54: So instead of me copying and pasting data into a chat window, the AI is actually inside the software, pulling the levers itself.

00:06:02: That's it.

00:06:02: Spencer Tayhill talked about this with the new plugins for Claude Cowork.

00:06:06: The AI isn't working around your systems anymore.

00:06:08: It's working through them.

00:06:09: It has hands, basically.

00:06:11: Okay, but if everyone has access to these agents, where's the competitive advantage?

00:06:15: If the AI is a commodity, how do you win?

00:06:18: That is the million-dollar question.

00:06:20: And Tael argues the moat, your advantage, isn't the AI model itself anymore.

00:06:25: The moat is context.

00:06:27: Context, as in our internal data.

00:06:30: Your proprietary data, CRM history, call recordings, your sales methodology.

00:06:34: If you feed that context to the agent, it becomes a specialized employee that knows your business inside and out.

00:06:40: Without it, it's just a generic, helpful stranger.

00:06:43: That

00:06:43: is huge implications for the software vendors.

00:06:45: I saw Subra Krishnan used to dramatic term for it.

00:06:48: This is apocalypse.

00:06:49: It's a bit dramatic, but the logic is sound.

00:06:52: Think about it.

00:06:53: Traditional marketing clouds, your HubSpots, your Adobe's, they sell you a user interface for humans

00:06:58: to click.

00:06:59: Write the drag and drop builders, the dashboards.

00:07:01: But Christian's point is, if an AI agent can just read the data from the database and execute the action, what happens to that big, expensive middle layer?

00:07:09: It

00:07:10: becomes less valuable, maybe even optional.

00:07:12: Or at least, a lot less valuable.

00:07:14: If no human needs to click the buttons, why are you paying for the fancy UI?

00:07:19: The value shifts entirely to the data infrastructure.

00:07:22: The plumbing underneath.

00:07:23: Wow.

00:07:24: But, okay, let's be real.

00:07:25: We've been hearing about AI taking over for a while.

00:07:28: Is it actually happening?

00:07:29: It is, but slowly.

00:07:32: And the bottleneck isn't the tech, it's us.

00:07:35: Liza Adams shared a telling statistic.

00:07:38: Only twenty-one percent of organizations have actually redesigned their workflows for AI.

00:07:42: Only twenty-one percent.

00:07:44: Yep.

00:07:44: So most companies are just bolting AI onto old broken processes.

00:07:49: It's like having an AI teammate but they're sitting in a separate chat window totally disconnected from the actual work.

00:07:54: Like buying a high-speed train and trying to run it on old rusty tracks.

00:07:57: That's a perfect analogy.

00:07:59: The value only unlocks when you change how you work.

00:08:01: Justin Norr shared a personal example of this.

00:08:03: I found fascinating.

00:08:04: He uses an AI chief of staff.

00:08:06: I love this example.

00:08:07: It's not about booking meetings or filing expenses.

00:08:09: He uses it for strategic orientation.

00:08:12: So

00:08:12: what does that look like in practice?

00:08:14: He just pastes in his messy notes, his goals, his worries.

00:08:18: all of it.

00:08:19: And then he uses the AI as a sounding board.

00:08:22: He asks it to prioritize his focus, point out conflicting goals, give feedback on team dynamics.

00:08:27: So it's a thinking partner.

00:08:29: Exactly.

00:08:29: He's redesigned his own planning process around the tools capability.

00:08:33: That's a real workflow redesign.

00:08:35: I really like that.

00:08:36: But let's go back to that context idea.

00:08:38: If AI needs our data, where does that data actually live?

00:08:42: This brings us to the CDP, the customer data platform.

00:08:45: It feels like that war is heating up again.

00:08:47: It is, and the market is bifurcating splitting in two.

00:08:50: Real long acre broke this down really well.

00:08:52: On one side, you have the platformized CDPs.

00:08:55: So the big all-in-one suites, Salesforce, Adobe.

00:08:59: Correct.

00:08:59: They want to hold your data inside their proprietary cloud.

00:09:02: On the other side, you have these rising agentic or composable CDPs.

00:09:06: They're warehouse native.

00:09:08: Warehouse

00:09:08: native.

00:09:09: Yeah.

00:09:09: Meaning the data stays in my own database, like Snowflake.

00:09:12: Exactly.

00:09:13: You own the data.

00:09:14: The CDP is just a thin layer on top that helps you activate it.

00:09:17: Longacre thinks this is going to become the customer operating system for all those AI agents we were just talking about.

00:09:22: But

00:09:22: the big players aren't just giving up.

00:09:24: I saw Lucas Lunow called Salesforce's data three sixty Schrodinger's platform.

00:09:30: Leftism.

00:09:31: Yeah.

00:09:32: Yeah.

00:09:33: As in the cat.

00:09:34: It's both a CDP and not a CDP at the same time.

00:09:37: Salesforce is trying to blur the lines, be everything to everyone.

00:09:41: Which

00:09:41: usually leads back to bloat, doesn't it?

00:09:43: It

00:09:43: often does.

00:09:44: But is anyone actually doing this warehouse native thing at a huge scale?

00:09:48: Oh yeah.

00:09:48: Juan Mateo de Monasterio shared a case study from Mercado Libre.

00:09:52: Think Amazon of Latin America.

00:09:55: The scale is immense.

00:09:56: How immense are we talking?

00:09:57: Four

00:09:57: hundred and fifteen million users.

00:09:59: Wow.

00:09:59: And they process a hundred and eighty thousand different audience segments every single day.

00:10:03: You

00:10:04: cannot do that with a spreadsheet.

00:10:05: Well, they used to try, but they moved to a composable CDP.

00:10:09: The data stays in their warehouse, and a lightweight layer on top recalculates those audiences automatically.

00:10:15: No more CSV uploads.

00:10:16: Oh, the dream.

00:10:17: And for anyone listening trying to figure this out, Supermanian Sidurman gave a great rule of thumb.

00:10:22: He said, Store everything in your data warehouse.

00:10:26: That's your attic.

00:10:27: OK.

00:10:27: But only put actionable, governed data in the CDP.

00:10:31: Keep

00:10:31: the junk in the attic.

00:10:32: Put the fine china in the dining room.

00:10:34: That is the perfect analogy.

00:10:35: Don't clog up your activation engine with data you're not going to use.

00:10:39: So we've cleaned up the stack.

00:10:41: We've deployed the AI.

00:10:43: We've sorted the data.

00:10:45: But none of this runs itself.

00:10:47: Which brings us to our last theme, marketing operations.

00:10:52: It really feels like Mahops is having its main character moment.

00:10:55: It is.

00:10:55: Mahops is moving from the back room to the board room.

00:10:58: The conversation is shifting from technical support to strategic leadership.

00:11:02: Enrico Ferrari had some advice for new CMOs that sounded, well, backwards at first.

00:11:07: He said, fix the team before you fix the stack.

00:11:10: It sounds counterintuitive, right?

00:11:12: A new leader wants to come in and buy a new tool to show they're making an impact.

00:11:15: Yeah, look, I bought this fancy AI thing.

00:11:17: We're innovating.

00:11:18: Exactly.

00:11:19: But Ferrari's point is, if the team can't operate the tech, the best software in the world is useless.

00:11:25: It's like giving a Formula One car to someone who only knows how to ride a bike.

00:11:29: Claire Robinson even mapped out the career path for mall pops now.

00:11:33: It goes from builder to owner to architect and finally to leader.

00:11:37: And that leader part is about bridging what John Miller calls the knowing doing gap.

00:11:43: The knowing doing gap.

00:11:44: Miller points out that we've known for fifteen years that B to B buying isn't a straight line.

00:11:49: We've known it's messy.

00:11:50: But our tools forced us to use these rigid linear funnels.

00:11:54: Our software was dumber than we were.

00:11:56: Basically yes.

00:11:57: And he argues we're only now getting the AI native platforms that can finally operationalize the complexity we've known about for a decade.

00:12:04: The OMOP's leader's job is to close that gap.

00:12:07: That's all very strategic, but I want to end on something that brings it back down to earth.

00:12:12: Stephanie M. Wendell shared a story that every single HubSpot user will feel in their bones.

00:12:18: HubSpot renamed lists to segments.

00:12:20: Oh, I saw the fallout from that online.

00:12:22: It's so small, right?

00:12:23: A list is a segment, but she called the change emotionally violent.

00:12:28: Laughs.

00:12:29: Emotionally violent is a bit much, but I get it.

00:12:31: Think of the muscle memory.

00:12:33: You spent five years clicking on lists.

00:12:35: All your internal training docs say lists.

00:12:38: Right.

00:12:38: Now every junior marketer is going to say, I don't see lists.

00:12:41: It's a funny but real reminder that these tiny vendor changes can totally disrupt the factory.

00:12:47: It's a great point.

00:12:48: At the end of the day, humans still have to push the buttons.

00:12:51: And humans hate it when you move the cheese.

00:12:53: So if we pull this all together, what's the story?

00:12:56: It's a massive cleanup, a shift to AI that actually does things, a war for the data's foundation, and the rise of the ops leader who has to manage it all.

00:13:05: I think it means the industry is maturing.

00:13:07: The party of endless software buying is over.

00:13:11: We're sobering up.

00:13:12: We're stopping the frantic buying and starting the thoughtful building.

00:13:15: It's harder work, but it's where the real value is.

00:13:18: If you enjoyed this episode, new episodes drop every two weeks.

00:13:22: Also, check out our other editions on account-based marketing, field marketing, channel marketing, AI in B to B marketing, go to market and social selling.

00:13:31: Thanks for listening.

00:13:32: There's a lot to digest there, but I think the path forward is clearer than it's been in a long time.

00:13:36: Thanks everyone and don't forget to subscribe.