Best of LinkedIn: MarTech Insights CW 14/ 15

00:00:00: This episode is provided by Thomas Allgaier and Frennis, based on the most relevant LinkedIn posts about MarTech from calendar weeks fourteen-and fifteen.

00:00:08: Frennis a B to B market research company that supports enterprise marketing teams in unlocking full potential of their customer data with help of AI.

00:00:17: You can find more info in description.

00:00:19: So if you've ever bought like seven different software tools You know the pain.

00:00:25: Oh, absolutely you spend months on implementation And then you just realize your revenue growth has actually slowed down.

00:00:31: right if you're listening Right now and you're looking at a dashboard that just doesn't quite match Your CRM reality well?

00:00:37: You know exactly what that kind of diagnostic muddy water feels like.

00:00:41: welcome to The deep dive.

00:00:42: yeah today we are pulling from a stack of recent highly debated LinkedIn posts From top Martek professionals.

00:00:49: We really want uncover the structural shifts happening in B to D marketing right now.

00:00:53: Exactly, we're moving past the theoretical fluff were looking directly at system architecture data foundations and well somewhat alarming reality of AI orchestration.

00:01:03: it is definitely alarming And The overarching theme across all these professional discussions Is I mean?

00:01:09: It's just a fundamental breakdown with the old models.

00:01:11: Yeah for over A decade You know the standard operating procedure In bdb marketing was To buy a core platform and then infinitely bolt on peripheral tools to solve every little micro problem that popped up.

00:01:26: Right,

00:01:26: just slapping band-aids on it?

00:01:28: Exactly but that era is effectively over.

00:01:31: the infrastructure's just buckling under its own weight.

00:01:33: let's actually start by dismantling that core concept though one BB marketers have clung to like a life raft which is The Martech Stack.

00:01:42: oh the famous stack.

00:01:43: right?

00:01:43: We're seeing industry heavyweights like Scott Brinker, Friends Ramresma and Hugo Boudry all pointing out that the traditional concept of The Stack is basically dead.

00:01:53: Yeah

00:01:53: they say it's being replaced by this composable canvas.

00:01:56: Exactly!

00:01:57: And I was mapping this out and realized the old stack was a game of Jenga

00:02:00: A Game Of Jenga?

00:02:01: Love That!

00:02:02: Think about it.

00:02:03: you start with solid block maybe your email platform.

00:02:06: Then you add an intent data tool, then an enrichment layer.

00:02:10: I don't know a lead scoring application.

00:02:12: Right just piling them

00:02:13: up?

00:02:14: Every piece that is added might serve as specific function but it makes the entire structure wildly unstable.

00:02:21: You pull wrong block or API connection breaks and whole thing comes crashing

00:02:25: down.

00:02:26: But this new system design they're advocating for, it's completely different.

00:02:30: It's more like a modern city grid.

00:02:32: The infrastructure—like the data layer is shared universally below ground.

00:02:37: Then the buildings which are your various activation tools and apps.

00:02:41: They just plug into that grid seamlessly above-ground.

00:02:45: but if the stack is dead why do organizations still operate like chronic tool hoarders?

00:02:51: Well the mechanical difference between those two models Is really where the friction lives.

00:02:55: When you build a Jenga tower, the friction comes from pieces trying to balance on top of each other.

00:03:01: Yeah that makes sense.

00:03:02: But in Martek The Friction isn't coming from software tools themselves It's coming form absolute absence Of underlying city grid.

00:03:11: There was this brilliant but somewhat painful reality Shared by Cameron Korniak recently on LinkedIn.

00:03:19: That illustrates it perfectly.

00:03:20: Oh I saw him!

00:03:21: He is hired to auto start up right?

00:03:23: Yes

00:03:24: His mandate was very specific.

00:03:26: Fix HubSpot.

00:03:27: Amanda, I'm sure half the people listening right now have been given at some point?

00:03:30: Precisely!

00:03:31: But when he actually got under the hood and mapped the data flow... He found just a complete architectural disaster.

00:03:37: Wait what did he

00:03:38: find?!

00:03:38: He found three separate customer relationship management platforms.

00:03:43: Three.

00:03:43: Oh, wow!

00:03:44: And an enterprise resource planning tool and ERP that was only integrated with the oldest of three CRMs Plus.

00:03:51: there is an inherited business intelligence tool.

00:03:53: That's supposed to be their source-of truth But literally no one in company knew how write a query for it.

00:03:58: That sounds like nightmare

00:03:59: Seven disconnected systems total With zero lateral data flow.

00:04:04: So Kornik realized this problem.

00:04:05: never hub spot.

00:04:06: The problem was trying to force a collection of siloed software, To cooperate with an accompanying culture that itself couldn't cooperate.

00:04:14: And rather than doing the grueling work Of fixing this underlying architecture, marketers just keep falling into the exact same trap!

00:04:21: We try and buy our way out of structural problems.

00:04:24: We really do.

00:04:25: Vivek Nair & Bill Habib highlighted this behavioral loop perfectly.

00:04:29: They noted that marketers will buy virtually anything that promises to fix their marketing, provided it doesn't require them actually fixing the market fundamentals.

00:04:38: It's the shiny object syndrome

00:04:40: Exactly!

00:04:41: We just keep buying new tools.

00:04:43: Only now instead of a new automation platform we're buying something repackaged as an AI agent.

00:04:49: But as Hobbit points out if you are plugging an AIAgent into system with seven disconnected CRMs You aren't innovating, you are just renaming the bloat.

00:04:57: Yeah and that behavior I mean it made sense ten years ago.

00:05:01: back then as Kenneth Wagner in Bread House note buying a shiny new tool actually gave you a competitive edge

00:05:07: because no one else had yet

00:05:09: exactly.

00:05:09: if you have an automated lid scoring tool your competitor didn't want to deal.

00:05:13: but today we're looking at landscape with over fifteen thousand distinct Martek tools.

00:05:21: That is just absurd number

00:05:23: It Is And every single one of your competitors can buy the exact same off-the-shelf software that you can, usually with a corporate credit card in under five minutes.

00:05:34: So because the tools are completely commoditized now... The competitive advantage no longer lives on the software vendor you choose.

00:05:41: It lives entirely into proprietary workflows Your unified data models and bespoke decision systems.

00:05:51: Organizations have to stop operating like tool hoarders.

00:05:53: We have to look at what actually holds the modern marketing ecosystem together,

00:05:57: right?

00:05:58: The grid

00:05:58: exactly the grid that brings us to the new center of gravity Which is the data foundation.

00:06:03: the Jenga pieces are gone and unified data Is becoming the absolute single source of truth.

00:06:08: specifically we're seeing warehouse native architectures completely replacing traditional siloed customer data platforms or CDPs And the market is aggressively voting with its wallet here.

00:06:21: Josh Kanagi and Alan Tran pointed out that Hightouch, which is a primary player in this space just hit one hundred million dollars an annual recurring revenue.

00:06:30: Wow!

00:06:31: Yeah That level of enterprise adoption Is massive.

00:06:33: validation.

00:06:35: Companies are finally realizing they want their execution tools to run on Their own proprietary data rather than Just handing it all over To third-party vendors

00:06:44: And the mechanics Of this shift Are critical to understand In a traditional CDP model, you're essentially copying your customer data from your central database moving it over to the marketing platform and storing it there.

00:06:56: Which

00:06:57: sounds incredibly inefficient.

00:06:58: It is!

00:06:58: It creates latency...it create security risks And means that your marketing team always operates on slight delay.

00:07:05: Warehouse-native architecture means the marketing tools sit directly on top of your central data warehouse

00:07:10: So theres no copying?

00:07:11: Zero copying.

00:07:12: The Tools simply query live data where already lives.

00:07:16: Michael Burton and Jessica Leeds gave a fantastic example of how this practically changes deployment.

00:07:21: Okay, what was their tip?

00:07:23: They argue that if you look at data warehouse platform like Databricks And the customer engagement platform like Brays You shouldn't think of them as two separate vendor relationships.

00:07:33: they have to be architected As single unified organism.

00:07:36: interesting

00:07:37: yeah Databricks is the bedrock.

00:07:39: It houses the absolute source of truth for customer identity, behavioral signals and predictive

00:07:44: models.

00:07:45: Embrace is simply the intelligent activation layer sitting directly on top of that

00:07:50: bedrock.".

00:07:51: And the mechanical distinction there if sitting directly up versus sitting downstream...that changes the entire operational math!

00:07:58: Completely.

00:07:58: because if your activation tool sits downstream you're syncing audiences and overnight batches.

00:08:04: You are basically reading yesterday's newspaper to trade stocks today,

00:08:07: right?

00:08:08: Your always behind

00:08:09: but when your tools it's natively on the warehouse.

00:08:12: a churn risk signal can fire in The data model at two point zero p.m.. And A targeted retention journey launches that two point Zero one p.M Before the customer even closes their browser tab.

00:08:22: That immediate causality is the holy grail.

00:08:25: yeah.

00:08:25: But you know Getting there requires navigating a structural mind-feel because organizations know they have data problem.

00:08:32: but their approach to fixing it is fundamentally flawed.

00:08:35: Oh,

00:08:36: this exactly what Gene DeLibro decepted so well.

00:08:39: he pointed out the corporate habit of conducting sequential audits?

00:08:42: Yes!

00:08:42: The phases

00:08:43: Right.

00:08:43: A company will hire consultants to audit people in phase one then they audit process and phase two Then data in Phase three And content in Phase four.

00:08:51: neatly categorized

00:08:52: Too neatly.

00:08:53: Exactly Because DeLibero argues that Mar-Tech infrastructure doesn't fail neatly inside those isolated categories It fails on white space between them.

00:09:01: Ah yes The handoff.

00:09:03: Yeah, a broken internal process destroys the data quality during the hand-off.

00:09:08: and then that bad data makes your automated campaigns actively dangerous to the brand.

00:09:13: so I have to challenge this model.

00:09:15: how does a CMO actually bridge that white space?

00:09:19: Because right now it feels like bridging requires a PhD in Data Engineering

00:09:22: And most marketing leaders definitely do not have them.

00:09:25: No they don't.

00:09:26: That gap is core dilemma for modern market leadership.

00:09:29: And it's precisely where CMOs are making their most expensive mistakes today.

00:09:33: Schimrit, Sabag and Emil Bjornsger addressed this head-on... What

00:09:37: did they say?

00:09:38: They pointed out that CMO is desperately trying to use artificial intelligence to bridge the white space.

00:09:43: They treat AI as a magical capability fix which will somehow decipher the mess.

00:09:49: Like waving magic wand!

00:09:51: Exactly.

00:09:51: but AI cannot fix your marketing if underlying identity resolution or measurement foundations are broken.

00:09:57: If a user's web browsing session can't be reliably linked to their final purchase event in your database, throwing an AI prediction model on top of that doesn't solve the problem.

00:10:06: Just

00:10:06: scales your existing blind spots?

00:10:08: Yes!

00:10:10: It simply scales the blindspots...

00:10:12: ...it essentially gives you power.

00:10:14: make highly confident terrible decisions at light speed.

00:10:17: Exactly.

00:10:18: The only way CMO bridges white space is by fundamentally changing relationship with IT and data departments.

00:10:26: The CMO has to pull up a chair at the data engineering table and explicitly claim that data infrastructure as a primary marketing asset.

00:10:35: You can't just toss it over the fence anymore?

00:10:37: No, It is no longer acceptable to say...

00:10:48: Okay so let's assume an organization actually achieves.

00:10:51: The warehouse is humming and the CMO has secured the data foundation.

00:10:56: A rare but beautiful scenario,

00:10:58: right?

00:10:58: The next logical question is what runs on top of it And the answer is shifting rapidly away from human marketers manually building segments and pushing execution buttons.

00:11:08: We are moving toward AI agents making autonomous decisions based on live signals.

00:11:13: we're transitioning From the era of cute generative AI experimentation like you know having a bot write a blog post into full-scale transactional orchestration

00:11:22: And the distinction between generative experimentation and transactional execution is massive.

00:11:28: Sarah McNamara recently shared some highly practical use cases for autonomous bots that show just how close we are to this reality.

00:11:35: Yeah,

00:11:35: her stuff is amazing.

00:11:36: She's actively building AI bots for revenue operations That function as a continuous co-pilot for ops.

00:11:44: These aren't just novelty summarization tools.

00:11:46: They're taking over complex analytical workflows.

00:11:49: The mechanics of her use cases are fascinating Mainly because they don't require massive enterprise software investments, They just requires smart prompting and clean data

00:11:59: access.

00:11:59: Right

00:12:00: Like she deployed a sales pipeline bot that autonomously checks open CRM deals every single morning.

00:12:05: It flags opportunities that have gone stale based on historical velocity And it automatically tags the specific sales reps with recommended actions.

00:12:12: That's

00:12:13: incredible.

00:12:13: Yeah!

00:12:14: She built another bot that acts as solutions brief generator surfacing accounts that need immediate renewal attention.

00:12:20: She even has a competitive intelligence bot that scrapes competitor websites and cross-references data against internal sales called transcripts.

00:12:29: See, what McNair is building represents a critical evolution from tasks to transactions.

00:12:35: Sathya Narayana introduces the concept he calls ConverX.

00:12:38: that perfectly frames them He points out.

00:12:41: in twenty-twenty four The industry figured how make AI talk to customers effectively.

00:12:46: We built conversational interfaces, barter chatbots all of this.

00:12:49: But looking toward twenty six AI's primary value won't be having open-ended conversations.

00:12:55: It will be closing transactions with verifiable kinetic outcomes in the

00:12:59: database."

00:13:00: Which

00:13:00: shifts the entire financial model of marketing, we basically stop optimizing for cost per click or costs per lead and move toward a Cost Per Outcome or CPO Model.

00:13:09: We're giving AI agents bounded autonomy to actually execute work flow Whether that is provisioning software trial Or sending custom pricing contract all based on real-time CRM data and inventory levels.

00:13:23: Bounded autonomy is the operative phrase there, The AI is authorized to drive a commercial outcome but only within strict operational guardrails.

00:13:32: I want to push back on the safety of that autonomy though because from a brand management perspective, letting an AI autonomously negotiate or execute deals sounds like a massive liability.

00:13:43: Oh it absolutely is if you aren't careful!

00:13:44: Gregory

00:13:45: Pete offered a brilliant critique Of Scott Brinker's composable canvas model That isolates this exact fear.

00:13:52: Pete warns The industry about what he calls strategic hallucinations.

00:13:56: Okay What does he mean by that?

00:13:57: Well, this isn't the classic AI hallucination where a language model just makes up a fake statistic.

00:14:02: A strategic hallucination is much more dangerous.

00:14:05: it happens when an AI makes a choice that is factually correct based on perfectly clean data but is strategically completely off-brand eyes.

00:14:13: yeah like it contradicts the core identity or pricing philosophy of the company.

00:14:17: so how do we architect our systems to put decisional governance upstream?

00:14:23: How do we ensure these autonomous agents don't just go rogue, offering wild discounts or sending completely off-tone emails?

00:14:30: Just because the math says it will optimize a conversion rate.

00:14:33: That is the defining architectural challenge of the next five years.

00:14:37: Data without an upstream decisional model just produces raw insights but data flowing through essentialized decision models produces governed verdicts.

00:14:48: This connects perfectly to John Miller's predictions regarding the future of traditional marketing automation platforms or MAPs.

00:14:55: Historically, the MAP like tools like Mercado and Eloqua was the absolute center in the marketing universe.

00:15:01: Right it is the engine room where every rule was manually coded.

00:15:05: But Miller argues that the role of the traditional MAP is shrinking dramatically As these new AI agents take over the actual execution of tasks at the edge of the network, The MAP evolves into a marketing context and orchestration service.

00:15:19: So what does that look like in practice?

00:15:20: It essentially becomes the centralized corporate brain.

00:15:23: it stores your operational guidance Your brand voice parameters you're hard governance rules Like email frequency limits And your compliance consent logs.

00:15:33: Oh okay.

00:15:33: so Functions like a highly advanced, strictly enforced corporate handbook.

00:15:37: Exactly!

00:15:38: The AI agent might spot an opportunity to accelerate the deal but before it is allowed send email or offer discount must query the MAP check rules and engagement

00:15:48: Precisely.

00:15:49: That's how you put governance upstream.

00:15:51: You bind the AIs autonomy with strategic context.

00:15:54: The agent executes transaction But the MEP ensures that never commits a strategic hallucination.

00:16:00: So if we zoom out and look at this complete picture, a universal city grid of data at the bottom.

00:16:05: And AI agents executing workflows that they cop based on strict upstream governance We have to ask what exactly is the role Of The human B-to-B marketer in This new model?

00:16:15: That Is the million dollar question

00:16:16: because If the machines are handling the execution In the data layers Handling the routing The human operating model has to fundamentally change.

00:16:25: We're seeing marketing operations evolve from tactical campaign execution into pure system architecture,

00:16:32: and the human element is actually where these models are currently failing the hardest if we look specifically at revenue operations or rev ops.

00:16:40: that transition has been incredibly rocky.

00:16:43: how so

00:16:43: Well, Mario Pyshev and Don Crow have both highlighted that RevOps as a holistic organizational concept has largely failed to deliver on its promises.

00:16:52: Wow!

00:16:53: That is harsh assessment for the framework.

00:16:55: almost every modern B-to-B enterprise claims be running.

00:16:58: It's harsh?

00:16:59: but the empirical evidence backs it up.

00:17:01: Peshev shared a highly illuminating example of a company with one hundred and twenty million dollars in ARR.

00:17:07: This organization invested five million into new sales tooling, intent data platforms And additional headcount.

00:17:13: That's

00:17:13: massive investment

00:17:14: Massive.

00:17:15: Yet after that capital injection their overall revenue growth actually slowed down.

00:17:19: Great

00:17:19: really?

00:17:20: Yes!

00:17:21: The root cause wasn't software.

00:17:23: It was the fact that their sales, marketing and customer success departments were operating in fiercely guarded silos.

00:17:30: Of course!

00:17:31: Marketing was generating thousands of automated leads... ...and the sales team was manually rejecting sixty percent because they didn't trust the scoring model.

00:17:40: The department flat-out refused to share raw data or align on shared metrics.

00:17:44: Perfectly proves the point, you cannot throw an advanced AI agent at a leaky organizational funnel where human departments actively withhold context from one another.

00:17:54: It's an organizational failure masquerading as a tech failure

00:17:57: Exactly and because of that The way operations professionals measure and communicate their value has to change immediately.

00:18:05: Wade Coorey and Fund Me Marin are urging MOPs professionals to entirely change their vocabulary.

00:18:11: Changing your vocabulary in what way?

00:18:13: Well, Marin points out that operations teams spend a massive amount of their unofficial time fixing database permissions or manually routing leads.

00:18:20: It's pure plumbing work.

00:18:21: But if you walk into a meeting with the CFO And talk about your lead routing rules you will be viewed strictly as a cost center.

00:18:29: Absolutely

00:18:29: So.

00:18:30: Kuri argues that ops leaders must translate that plumbing into commercial outcomes.

00:18:35: You don't talk about API uptime, you talked speed to lead and revenue velocity.

00:18:41: You prove how your system designed directly accelerated cash flow.

00:18:45: That is how operations secures the seat at the executive table

00:18:49: And driving that level of alignment requires an entirely new breed of leadership.

00:18:53: Anna Moran and Ken Madsen suggest organizations desperately need to formalize a new role the chief marketing technologist,

00:19:00: or a dedicated strategic owner.

00:19:01: Right someone who actually owns The entire end-to-end marketing supply chain.

00:19:06: because right now the traditional CMO is often too focused on brand and channel performance

00:19:11: And the CIO was to focus on security in infrastructure.

00:19:14: exactly neither of them Actually owns the connective tissue the data plane that allows these AI agents To function across departments.

00:19:22: this new leader has to carry the ultimate fiduciary duty for customer data.

00:19:26: Transitioning into that kind of system-level leadership is a brutal emotional shift for professionals who grew up in the trenches, though.

00:19:33: Mike Rizzo touched on the psychological transition for operations leaders and it perfectly explains why so many teams struggled to scale.

00:19:43: It's hard to let go

00:19:45: so hard.

00:19:46: Historically you get promoted in marketing ops because you are the absolute best at doing The technical work.

00:19:52: You build perfect rigid tightly controlled systems,

00:19:56: right?

00:19:56: But the moment you transition into leadership your core job stops being about the systems themselves and starts being about humans operating those systems.

00:20:04: You have to completely let go of control that made you successful in first place?

00:20:08: You really do!

00:20:09: Rizzle emphasizes, The leader's primary function becomes creating psychological safety.

00:20:14: Your team needs feel completely comfortable raising their hand at a sprint meeting saying Hey this new AI workflow is inherently flawed And it will break our routing.

00:20:23: Yeah because if don't build that safety they'll just watch it break.

00:20:26: They will watch system break quietly and a quiet, undocumented break in an automated architecture is the most dangerous kind of failure a company can experience.

00:20:37: I want to synthesize that reality with A Brilliant Point raised by Niels Van Meer de Manzey & Gagan Byssa.

00:20:43: They observe we are fundamentally moving from era of marketers who use tools To an era of operators build with AI.

00:20:51: So if human resistance to change is the true bottleneck here, how does a traditional B-to-B marketing department culturally shift?

00:20:59: To act more like a software engineering sprint team.

00:21:01: The cultural shift only happens when you completely redefine what.

00:21:05: If your team views itself purely as a campaign factory measured only by how many webinars they host or emails They send you will never make the leap right.

00:21:15: You have to start treating Your internal platforms, your data pipelines and your AI agents As actual primary products?

00:21:29: Exactly.

00:21:30: You have to shift the culture, to value the integrity of the architecture just as much if not more than the creative output of the campaigns And

00:21:38: that architectural integrity is going to be tested in ways we are only beginning to comprehend.

00:21:43: I want you leave with a final forward-looking thought based on everything we've unpacked

00:21:48: today.

00:21:48: Okay let's hear it.

00:21:49: We've

00:21:49: spent this entire deep dive analyzing how B-to-B marketing is evolving into a complex ecosystem of autonomous AI agents orchestrated by centralized platforms running on universal data grids.

00:22:01: But we have only looked at the sell side, what happens when your enterprise buyers procurement process also fully agentic?

00:22:10: What happens if their finance department deploys Autonomous AI to source vendors evaluate features and benchmark pricing.

00:22:17: Will the future of B-to-B marketing simply be our AI agents negotiating directly with their AI agents over API endpoints?

00:22:25: That is wild to think about.

00:22:27: And if so, how do we architect our data—our content and strategic governance for a reality where the primary consumer in our market isn't even human?

00:22:38: Think back that diagnostic muddy water we started with... The landscape isn't getting any simpler….

00:22:44: No it's not.

00:22:45: But at least now we know that to survive it, We have to stop staring At the broken Jenga pieces and start paying very close attention To the grid connecting them.

00:22:53: A truly fascinating And imminent reality.

00:22:57: to prepare for The architecture really is everything.

00:22:59: If you enjoyed this episode New episodes drop every two weeks.

00:23:02: Also check out our other editions on account based marketing Field Marketing Channel Marketing AI in B-to-B Marketing Go-To Market & Social Selling.

00:23:10: Thank You so much For joining us On This Deep Dive.

00:23:12: Make Sure To Subscribe

00:23:13: And Keep Questioning Your Systems.