Best of LinkedIn: AI in B2B Marketing CW 37/ 38

00:00:00: This episode is provided by Thomas Allgaier and Frennis based on the most relevant LinkedIn posts about AI and B to B marketing in calendar weeks, thirty seven and thirty eight.

00:00:10: Frennis is a B to B market research company helping enterprise marketing teams sharpen their strategies and outreach with customer segmentation, ideal customer profiles and deep dives, customer needs analysis and buying center insights.

00:00:24: Welcome back to the deep dive.

00:00:26: Our mission today.

00:00:27: Well.

00:00:29: It's really to cut through the noise around AI and get straight to the revenue impact.

00:00:32: Absolutely.

00:00:33: Because in the last few weeks, the conversation on LinkedIn about AI and B to B marketing, it feels like it's completely shifted.

00:00:38: Oh, totally.

00:00:39: It's not about, oh, wow, look what AI could do someday.

00:00:41: It's really zeroed in on execution, actual pipeline impact, and go-to-market GTM strategy wins.

00:00:49: People are doing stuff now.

00:00:50: That's the critical pivot, isn't it?

00:00:52: Yeah.

00:00:52: They've sort of moved past the shiny object phase.

00:00:54: Right.

00:00:55: And the key mission now for anyone to be to be is really distilling these insights.

00:00:59: We need to understand where operators are actually seeing wins, like measurable pipeline impact, practical, quick wins you can implement.

00:01:06: Stefan Reepen was really on the money when he said, AI isn't just entertainment, it's about revenue.

00:01:11: Yeah.

00:01:11: He pointed out that seventy-eight percent of sales teams are already shortening their sales cycles.

00:01:17: And how?

00:01:21: It's about using tech to get organized, handle data better, and just move faster.

00:01:25: Okay, so let's unpack this shift.

00:01:26: Where should we start?

00:01:28: I think the biggest theme emerging in the sources was this idea of agentic marketing.

00:01:33: Ah, yes, AI agents.

00:01:34: Right,

00:01:34: we're talking about AI agents that don't just, you know, draft a nice email for you.

00:01:39: They're more like autonomous systems that actually act, learn, and personalize across the whole funnel.

00:01:45: And this agentic marketing idea, it's kind of revolutionizing B to B strategy, mainly because it addresses this new type of buyer, right?

00:01:52: The high velocity, really self-educated buyer.

00:01:54: Okay.

00:01:55: Sarah McConnell described it well.

00:01:56: She said the traditional funnel just can't keep up anymore.

00:01:59: You've got buyers doing maybe six hours of research before they even talk to a rep.

00:02:02: Wow.

00:02:03: Six hours.

00:02:04: Yeah.

00:02:05: So agentic AI steps in as this like... always on personalized solution to engage them earlier and more effectively.

00:02:11: And the potential scope here seems massive.

00:02:14: I saw Robert Zimmerman outline something like fourteen different ways agentic AI transforms the marketing funnel.

00:02:20: Yeah,

00:02:21: it covers a lot.

00:02:21: Everything from like instant engagement, twenty four seven availability.

00:02:26: all the way to personalize, nurturing, and even predictive analytics for pipeline forecasting.

00:02:32: It's a whole end-to-end vision.

00:02:34: But it does make you wonder, how are these agents really different from just using a standard chat GPT prompt?

00:02:41: That's the core thing people often miss.

00:02:43: I think Summit-N put it really well.

00:02:45: He said, you know, ninety-nine percent of people are talking about AI agents, but the winners, they're the one percent who are actually building them.

00:02:51: Building them as systems.

00:02:52: Exactly.

00:02:52: As structured systems.

00:02:53: They're not just single prompts thrown over the wall.

00:02:55: The real path forward requires thinking and systems.

00:02:58: You're combining large language models, specific tools, memory and structured workflow.

00:03:03: Tell me more about that.

00:03:04: That memory component is really key.

00:03:05: It means the agent remembers past interactions.

00:03:08: It can correct its own mistakes over time and it learns.

00:03:11: It gets better.

00:03:12: Ah, okay.

00:03:13: That makes sense.

00:03:13: So it's less like a one-off command and more like a tiny a rules-based GTM operating system almost.

00:03:20: Pretty much, yeah.

00:03:21: And when people implement this right, the results are, they'll be huge.

00:03:25: Some had shared that at his company, Devcom X, they saw three to four times more meetings booked per thousand cents just by switching to these production AI SDR agents.

00:03:36: Wow.

00:03:36: It's not just a little tweak.

00:03:37: That's a fundamental shift in efficiency.

00:03:39: Absolutely.

00:03:40: The agent really acts as this ultimate orchestration layer.

00:03:44: And Katya Tarfoskaya actually provided a pretty useful five step framework for getting started with the Gentik TTM.

00:03:50: Okay.

00:03:51: It involves things like auditing your hot accounts first, then building dynamic audio.

00:03:55: She calls them MQAs, marketing qualified accounts.

00:03:57: MQAs, got

00:03:58: it.

00:03:58: And then deploying outbound AI agents specifically against those really well-defined lists.

00:04:02: And she's seeing this lead to something like a five to ten X increase in outbound scale.

00:04:06: It's precision at volume.

00:04:08: That kind of velocity, that scale.

00:04:10: It takes us right into our second major theme, SDR automation.

00:04:14: And the really contentious space of cold outbound.

00:04:19: Ah, yes, the is it dead?

00:04:21: debate exactly if these agents can orchestrate outreach at ten x scale does that mean cold outbound is finally done?

00:04:28: for

00:04:28: well according to the sources we looked at the answer seems to be no but and it's a big but it requires surgical precision.

00:04:36: okay that whole narrative that outbound is dead.

00:04:38: It got immediately challenged by people who are clearly winning with it right now, like Shapir B shared this amazing story, booked a fifty-six thousand dollar ARR deal from a single polled message.

00:04:49: Fifty-six

00:04:50: thousand from one polled message.

00:04:53: How?

00:04:54: The trick was using AI to surface these really deep third level signals.

00:04:59: Things like recent market expansions or specific leadership changes.

00:05:02: Stuff a human would probably never catch manually or it would take ages.

00:05:05: Right.

00:05:05: I remember spending hours digging for those kinds of signals on LinkedIn profiles.

00:05:09: Exactly.

00:05:10: And they could use that insight to craft a genuinely problem.

00:05:13: first email in under thirty seconds.

00:05:15: Thirty seconds.

00:05:17: Okay.

00:05:17: That fundamentally changes the SDR job description, doesn't it?

00:05:20: But.

00:05:21: that success story, it seems to contrast really sharply with some of the failures we're hearing about with mass automation.

00:05:28: Hamid Elias Karmani, for instance, argued pretty strongly that the claim AISDRs will replace human SDRs is maybe the biggest lie in sales right now.

00:05:38: Yeah.

00:05:38: And the data seems to back up that need for human oversight.

00:05:41: Hamid found that purely AI driven outreach, it gets a pretty dismal, zero point five percent reply rate.

00:05:47: Ouch.

00:05:48: Yeah.

00:05:48: Half a percent.

00:05:49: But.

00:05:50: And here's the key.

00:05:50: When you add human oversight back in, so you have a person managing the strategy, doing quality control, handling the relationship building, that reply rate jumps six times, up to three percent.

00:06:01: Six times higher.

00:06:02: Okay.

00:06:02: So the insight there seems pretty clear.

00:06:04: AI can generate the volume, sure, but it's the human who generates the signal, finds the relevance, and builds the necessary trust.

00:06:10: Which

00:06:10: I guess connects to Abba Samji's concept.

00:06:12: I found this fascinating.

00:06:14: He called it putting AI in chains.

00:06:15: Putting AI in chains.

00:06:16: Okay.

00:06:17: Yeah.

00:06:17: His point was that the fantasy of mass AI personalization just fails because it leads to hallucinations and generic, uh,

00:06:25: fluff.

00:06:26: Right.

00:06:26: We've all seen those emails.

00:06:27: Exactly.

00:06:28: So his solution, surgical targeting, moving away from broad lists and getting super specific, like VPs who move from an enterprise company to a startup in the last ninety days.

00:06:38: Really niche.

00:06:39: Okay.

00:06:39: Super targeted list first and then.

00:06:41: And then once you have that list, you constrain the AI brutally.

00:06:45: Some of these approaches basically only use AI.

00:06:48: for the first line of the email.

00:06:49: Just

00:06:50: the first line.

00:06:51: Yeah.

00:06:51: And with these ironclad constraints, maximum twelve words reference one specific fact from your research and absolutely no adjectives or fluff.

00:06:59: Just the fact.

00:07:00: Wow, that's strict.

00:07:01: Twelve words, Max.

00:07:02: By applying that kind of, you know, brutal constraint paradox, you actually end up driving quality and relevance, not just volume.

00:07:08: Interesting.

00:07:09: But this still raises a big question about the sustainability of high volume, doesn't it?

00:07:13: Even if the conversion is a bit better.

00:07:15: Marcus Stahlberg warned about this.

00:07:17: He said, yeah, AISDRs deliver volume, speed, persistence.

00:07:22: But conversion is the ultimate differentiator.

00:07:25: Without high conversion, just cranking up the volume leads you right back to low conversion bulk email territory, especially as people get better at recognizing AI-generated spam patterns.

00:07:36: And we saw a perfect, if small, example of that trust erosion.

00:07:40: in a test Kevin Paul ran.

00:07:42: Oh yeah, tell me about that one.

00:07:44: Well, it was a small sample, only a hundred and fifty emails total, but they tested the exact same email copy.

00:07:50: In one batch, they disclosed up front, hey, this email was written by AI.

00:07:54: OK, and the result?

00:07:55: Zero replies.

00:07:57: Zero.

00:07:57: Zero.

00:07:58: Compared to seven replies for the identical email without that disclosure.

00:08:02: So right now, at least, it seems like transparency about AI use might actually damage the perceived authenticity of the message.

00:08:09: That's

00:08:10: quite telling, isn't it?

00:08:10: People want connection, not just efficiency.

00:08:12: Definitely.

00:08:13: OK, let's shift gears a bit and move inside the organization.

00:08:16: Theme three.

00:08:17: accelerating workflow and dealing with data.

00:08:19: Because AI isn't just about outbound, it's becoming a massive internal time saver too, removing a lot of that administrative dead weight.

00:08:25: Right, the internal efficiency games.

00:08:27: Yeah.

00:08:28: Chase Diamond highlighted Apollo's AI assistant.

00:08:31: Apparently, he can do things like complete lead list building, find verified emails, and set up personalized sequences.

00:08:39: All in about twelve minutes.

00:08:40: No bids.

00:08:41: A task that he said previously took his team two entire days.

00:08:44: Two days down to twelve minutes.

00:08:45: That's, well, that's potentially three point five extra selling days a week back in their pocket.

00:08:50: Exactly.

00:08:51: And it echoes what Andres Perez was saying.

00:08:53: He emphasized, it's not always about closing more deals, but closing the deals you have

00:08:57: faster.

00:08:58: Right, removing friction.

00:08:59: Yeah, by streamlining things like lead prioritization, helping draft outreach, analyzing sales calls afterwards.

00:09:06: The human still does the critical selling.

00:09:08: but AI clears away all the administrative muck that slows things down.

00:09:13: And Nico Guillard pointed out something interesting too that sticking to manual processes might actually be starting to decrease credibility now.

00:09:20: Well, think about it.

00:09:21: Your prospects are interacting with companies where, you know, action items are automatically extracted from a meeting, follow-ups are drafted instantly, the CRM gets updated right away.

00:09:31: If you're still scrambling to manually take notes and follow-up days later, It just doesn't look as professional or efficient.

00:09:38: Yeah, that's a good point.

00:09:39: It sets an expectation.

00:09:40: And probably the most critical friction point AI seems to be solving internally is lead qualification.

00:09:47: Oh, definitely.

00:09:48: That's a huge bottleneck for so many teams.

00:09:50: Right.

00:09:50: Carolyn Healy provided, I think it was nine proven tactics.

00:09:54: She showed that things like predictive scoring can give you a forty-three percent improvement in qualification accuracy.

00:09:59: The

00:09:59: forty-three percent is significant.

00:10:01: And dynamic profile enrichment.

00:10:03: That yielded seventy-eight percent more accurate ideal customer profiles.

00:10:08: And Stav Aldog mentioned tools like Knock AI that help map your own first-party data into actual intense signals.

00:10:15: So sales isn't wasting time.

00:10:16: Exactly.

00:10:17: So sales teams can focus only on the buyers who are actually showing signs they're ready to engage.

00:10:22: Okay, but all the speed, all this acceleration, it only multiplies what's already there, right?

00:10:27: Which brings us to the big scary elephant in the room.

00:10:30: Data integrity.

00:10:31: Yep.

00:10:32: None of that acceleration matters if your foundational data is rotten.

00:10:37: That's the classic garbage in, garbage out problem amplified.

00:10:41: Jonathan MK stated that a staggering eighty three percent of GTM teams are struggling with AI implementation.

00:10:47: And why?

00:10:48: Let me guess.

00:10:49: Bad data.

00:10:50: Poor gate equality and a lack of governance.

00:10:53: Andrew Mewborn agreed, basically warning that if your sales process is already broken, slapping AI on top just automates the noise.

00:11:00: It makes the problem catastrophically worse,

00:11:02: faster.

00:11:03: And just imagine that kind of problem scaling up at a hundred X speed.

00:11:06: Charlie Saunders raised this really genuine risk of a data catastrophe.

00:11:09: Data

00:11:10: catastrophe sounds bad.

00:11:11: Yeah.

00:11:12: He warned that if you have autonomous agents updating opportunities in your CRM incorrectly, but doing it at hyper speed.

00:11:18: You need entirely new ways to validate and monitor that data, or you risk suddenly finding all your deals marked closed lost by mistake.

00:11:26: Yeah, that would be bad.

00:11:28: The big takeaway there feels pretty simple, though.

00:11:30: Automating a bad process with AI doesn't save time.

00:11:33: It just helps you fail bigger and faster.

00:11:35: Vail at scale, not ideal.

00:11:37: So if automating bad data leads to catastrophe, the only real solution is to fundamentally change how we build our GTM strategy in the first place.

00:11:45: Which brings us nicely to our final theme.

00:11:47: looking at AI-driven strategic alignment and visibility, how AI is changing the bigger picture.

00:11:55: Right,

00:11:55: moving up from tactics to strategy.

00:11:57: Exactly.

00:11:58: Sweeney Annamaraju challenged founders to ditch what he called the old GTM, you know, the fixed, annually refreshed ICP, the rigid annual pricing cycles.

00:12:07: Well,

00:12:07: way things have been done for ages.

00:12:09: Yeah.

00:12:09: and embrace an AI GTM instead, which means adopting speed as a core strategic principle.

00:12:15: Using AI signals to refresh your ICP maybe quarterly, tuning your product packaging quarterly too, just to keep pace with how fast the market shifts now.

00:12:23: That

00:12:23: strategic shift though, it highlights a massive readiness gap that Luigi pointed out.

00:12:28: He noted something like, ninety-six percent of leaders are stressed out by all the AI chaos.

00:12:32: I believe it.

00:12:32: But only twenty-eight percent are actually seeing real improvements yet.

00:12:37: And often the reason why is they're just bolting AI onto already broken workflows instead of embedding it strategically from the ground up.

00:12:44: Yeah, Coonstam reinforced that.

00:12:46: He basically said ninety-five percent of AI pilots flop not because the AI failed, but because the GTM leaders failed the AI.

00:12:53: Ouch.

00:12:54: So the successful companies are the ones integrating it thoughtfully into their core strategy.

00:12:59: Seems like it.

00:13:00: And while all this internal strategy is shifting, the external landscape, how buyers actually find us is changing too.

00:13:05: You mean search?

00:13:07: Yeah, but maybe not search as we knew it.

00:13:09: Marcus shared and argued pretty compellingly that AI visibility or AEO AI engine optimization is the next mega trend.

00:13:17: AEO.

00:13:18: OK, new acronym.

00:13:19: His thinking is that AI recommenders, like the ones powering chatbots and new search interfaces, will reward brands based on whether they answer human questions transparently and directly, just like a helpful human would.

00:13:31: So if you hide important info, like pricing,

00:13:33: AI might just reject you.

00:13:35: We won't recommend you because you're not being helpful and transparent.

00:13:38: Interesting.

00:13:38: That's a big shift from just keywords and backlinks.

00:13:42: And are people already acting on this AEO idea?

00:13:45: Apparently so.

00:13:46: Mojvoje confirmed that something like sixty-eight percent of searches already include an AI-generated response.

00:13:52: Sixty-eight percent?

00:13:53: Wow.

00:13:54: That's fast adoption.

00:13:55: Right.

00:13:56: And she mentioned specific B to B brands, lane chain momentum.

00:13:59: Clay, who she says are already crushing AI search.

00:14:04: And the big benefit seems to be speed.

00:14:06: AEO can apparently spike much faster than traditional SEO.

00:14:09: Faster

00:14:10: than SEO?

00:14:11: How?

00:14:12: It seems to instantly reward things like high authority backlinks and maybe enhance social media indexing.

00:14:18: So theoretically, you have a fighting chance to compete with the established giants almost immediately if you play the AEO game

00:14:23: right.

00:14:24: Okay, so that's a whole new frontier opening

00:14:25: up.

00:14:26: Definitely.

00:14:26: So if we try to synthesize the key tensions from this whole deep dive.

00:14:30: It feels like there are two strong forces pulling the BGTM world right now.

00:14:34: What are

00:14:34: they?

00:14:34: Well, first, AI agents are clearly transforming the funnel.

00:14:37: There's no denying the speed and scale they bring.

00:14:39: That's one force.

00:14:40: Speed and scale from AI.

00:14:42: But second, the most successful GTM teams, the ones really winning, seem to realize that human trust and quality data.

00:14:50: Those remain the ultimate differentiators.

00:14:52: Right.

00:14:52: Neil Weitzman and Jen Freeman both kind of underscored that duality, didn't they?

00:14:56: AI is fantastic for the mechanical stuff, the list building, the research, automating maybe tier three outreach.

00:15:04: The

00:15:04: grunt work.

00:15:05: Yeah, the grunt work.

00:15:06: But it just cannot replace empathy or intuition or the ability to build real trust when a complex deal is actually on the line.

00:15:14: As they put it, AI makes you efficient, but trust is what makes you money.

00:15:18: I like that.

00:15:19: Efficient versus money making.

00:15:21: And if we look at the pattern Liam Lawson outlined for the winners, the winners solved actual specific business problems and integrated AI into workflows people were already using.

00:15:30: The losers, they built really impressive tech demos, but require huge, difficult behavior changes from the team to adopt them.

00:15:38: So start small, solve a real pain point within the existing flow.

00:15:42: Don't try to boil the ocean.

00:15:43: Exactly.

00:15:44: Which I think raises a really important question for you, the listener listening to this right now.

00:15:48: Okay, let's hear it.

00:15:49: For your team, what is the single, small, maybe even unsexy problem?

00:15:55: a piece of administrative friction, a data bottleneck that, if you automated it with AI today, could realistically compound into measurable revenue gained down the road.

00:16:05: That small, unsexy problem that unlocks bigger games?

00:16:09: That's probably where the real leverage lies, isn't it?

00:16:11: Not in the flashiest demo, but in fixing that one annoying thing that slows everyone down.

00:16:15: Exactly.

00:16:16: Find that leverage point.

00:16:17: If you enjoyed this deep dive, new episodes drop every two weeks.

00:16:20: Also check out our other editions on account-based marketing, field marketing, channel marketing, martech, go-to-market, and social selling.

00:16:28: Lots to explore there.

00:16:29: Thank you for joining us for this analysis into the latest in B to B AI marketing.

00:16:33: Be sure to subscribe so you don't miss the next one.