Agentic AI in B2B Marketing: Fixing Funnels,
Measuring Impact, and Driving Revenue

Featured Speakers:-

Links and Resources

• Breaking Down Barriers: AI Tools and the Democratization of Data Analysis in B2B Marketing
• From Data to Dollars: Leveraging Predictive Account Scoring to Drive Revenue
• How To Become An AI-Ready Organization

Transcript

0:11:
Hello and welcome to another episode of the revenue focused marketer where we discuss anything and everything related to marketing as well as data.

0:20:
I’m your usual host, Hershey, and today I’m thrilled to be joined by two industry powerhouses for this conversation.

0:27:
From automation to authenticity, marketing strategy in the age of AI is really evolving.

0:34:
And to join me first up we have Chris.

0:37:
Chris, thank you so much for joining us today.

0:40:
Oh, it’s great to be here, Hershey.

0:41:
Thanks for inviting me.

0:42:
So Chris is the CMO at Sinneger, and then we have the person who’s joining him is Arpit.

0:50:
Arpit is our chief everything product officer, I would say, and he’s had a lot of experience in analytics.

0:58:
, kind of going back to Chris.

1:00:
Chris has had decades of experience in kind of launching innovative AI powered products, and he’s also built world class go to market teams, authored books.

1:11:
He also has podcasts of his own.

1:14:
So we’re kind of really thrilled to kind of have you join us here today, Chris.

1:18:
Well, thanks.

1:19:
It’s great to be here and nice to meet you, Arpit.

1:22:
Same here.

1:23:
Awesome.

1:24:
So today we’ll kind of talk about how the modern marketing engine is evolving and how AI is changing not just how we automate but also how we connect.

1:34:
And to kind of start this off, I do wanna start off with an icebreaker if that’s OK with you guys.

1:40:
Fantastic.

1:41:
Awesome, so, , maybe Chris, you can start as soft if AI could magically take one part of your job off your plate forever, what would you happily hand off to AI?

1:55:
That’s a great question.

1:56:
The part of my job that I would hand off to AI really comes back to, you know, my, my definition of marketing.

2:06:
So we should, we need to start there.

2:07:
, because historically, you know, during my tenure at Microsoft, at Dell, other large organizations, marketing sometimes falls back into being , the advertising or the publicity department for a feature crew, right?

2:24:
Like, tell us what you’re shipping.

2:26:
We’re gonna write a blog post about it once a month.

2:28:
We’re gonna update a PowerPoint.

2:30:
We’ll make sure there’s some social, we’ll go to 2 or 3 events a year.

2:33:
That’s tablestas.

2:34:
Marketing in my definition is the engine that Discovers capabilities that the market needs and turns them into businesses.

2:43:
And so all the way in the what is the product marketing in my definition is accountable for long-term product strategy.

2:50:
, not for building it.

2:53:
that typically is the R&D or the engineering divisions, but looking at that whole pipeline, then you have sort of classic inbound and classic outbound strategies in the marketing side.

3:03:
All of that is not where I would tend to focus.

3:08:
The thing that I would take off of my plate is, once you have all those capabilities, how do you and all the materials that you’re using for inbound and outbound, how do you turn those into sales enablement, channel enablement?

3:20:
That is the part, , where I think if I could hit an easy button, , AI could be even more impactful on my day to day job by looking at those repackaging aspects.

3:34:
You already have the content, so that’s a place where AI can do, can work wonders if you already have starting materials.

3:42:
Definitely love that, I think, what about you?

3:46:
Well, I’ll just keep it simple.

3:49:
Any of the meetings where I can’t add value, I’ll let my AI of that take on so that I have more time for family.

3:58:
Definitely.

4:00:
AI can definitely be something that can be kind of used to kind of balance that work-life balance that we’re kind of always looking forward towards.

4:08:
So now that we’ve kind of cleared our AI wish list, let’s dive into real work.

4:14:
, so Chris, I’ll come to you first.

4:17:
, if the funnel is broken and digital channels are kind of declining, what does a high performing B2B growth engine look like to you in 2025?

4:29:
That’s, that’s a fair question.

4:30:
I think in 2025, the place where pipelines get stuck is not at the very top, but somewhere in the 20% range.

4:44:
And it’s gonna vary based on your organizational responsibilities.

4:47:
Like, at what point is this, you know, with the synergy between marketing and the sales engine, do you transition that?

4:55:
But , Getting names, getting initial interest to be able to start a pipeline.

5:03:
, Isn’t really where things break down, in my experience.

5:08:
If you’re using metrics in marketing, if you were looking at that transition between marketing qualified lead and sales accepted lead, looking at, that is the place in the funnel where I think marketing can have the most impact on optimizing pipeline and funnel.

5:29:
Definitely.

5:31:
Are pet I’d love to also get your perspective on this.

5:35:
Yeah, definitely.

5:36:
I think it’s a very interesting and relevant questions.

5:39:
, I would say that the kind of change we are seeing in the last 12 to 24 months.

5:45:
It is definitely unprecedented.

5:48:
You pick a channel and there is definitely some challenges that every marketer is facing.

5:54:
You pick email.

5:56:
You send any number of emails, , most of them will land into junk or spam folders because the AI algorithms that are powering those email engines are very smart, and they detect all those cold emails, whatever you write.

6:11:
So there’s a challenge there.

6:12:
You pick search, you might have noticed that AI overview that is coming up on Google.

6:18:
, you, you talk to anybody, any client, most of us are seeing 50, 60% decrease in traffic.

6:27:
You pick social.

6:28:
It’s, it’s so noisy in B2B context, maybe LinkedIn’s most important channel, but it’s full of, I would say AI content and AI comments, right?

6:38:
So, and the moment you realize that, that it’s AI content, , generally, you won’t read it.

6:45:
, maybe it’s good, but if you’re getting a signal that it’s AI generated content, , definitely that’s, that’s a challenge again.

6:54:
So in my view, the whole new playbook that we all are probably doing work in progress kind of a thing, , we need to reimagine the whole go to market motion and, , really identify, , what, what can be more authentic that connects, , your brand with your audience.

7:17:
And then, , really also exploring some emerging channels.

7:24:
Take an example of Reddit, for example, , that could be a potential channels.

7:29:
We’re also seeing, , substack and even some dark social kind of channels where you have a smaller community in Slack, where you have very selected people, but then they are the influencers, micro influencers, or decision makers.

7:44:
Can you add value and participate and show what you’re solving for?

7:49:
, and those little things will have a lot of impact.

7:53:
I’m not saying that, , AI is, , not helping, it’s definitely helping, but you probably still have to own your own point of view, your, your brand story, and let AI be, , that, that force that brings the scale.

8:10:
Yeah, it’s, , I really agree with what you just said, Arpit.

8:13:
I think one of the things that’s going on is you’re looking at the evolution away from sort of traditional search to kind of emerging AI brokered answer generation.

8:23:
And it doesn’t matter whether you’re looking at Google and Bing, or you’re looking at Chat GPT Gemini Anthropic co-pilot, those answers are getting distilled and given to people who are never gonna come and find your content.

8:36:
So I think what you’re, you know, you’re exactly right.

8:39:
What I see is that, that sort of traditional approach to inbound, which is, let’s put our content out there so it will generate interest, is becoming less effective than being a little more skewed to outbound.

8:53:
Let’s get the messages to people and yes, email is becoming a harder and harder channel, but finding those emerging channels, so that you can get your brand identity into that.

9:05:
Otherwise, you know, , Chat GPT is going to carve your material out, great, but you might only be literally a footnote in what someone discovers there.

9:17:
Ah.

9:18:
Yeah.

9:18:
Yeah, definitely.

9:20:
And like, I think even, even the whole technology stack and marketing department uses so many applications.

9:27:
Maybe the, the, the tech side of the budget is, could be in the range of 1 or $2 million.

9:32:
But given all these emerging technologies we are building a website or landing page or a tool or an app, it is getting more or less commoditized.

9:44:
We’ll also have to rationalize some of that cost that marketing is owning right now, and that should be part of your go to market plan that why can’t we just invest that $2 million and Adding value to our prospects and customers versus just depending on tech stack and a lot of it is not even being utilized.

10:06:
So it’s funny go ahead Hershey.

10:11:
No, I was just gonna say you guys rightly mentioned like, you know, search is really evolving now.

10:16:
So the traditional sort of SEO is kind of really changing now.

10:21:
And do you think factors such as EEAT are still things that, you know, people should look out for, or specifically when they’re trying to grow their business, what are some factors now with like, you know, answer engine engines kind of being the main or more popular source for getting answers.

10:40:
, how does one kind of, you know, move to getting better results if people are not going to go through those 10 blue links anymore?

10:52:
Yeah, well.

10:54:
You want to go first, Chris?

10:56:
Yeah, I, you know, we have found both for our own book of business as well as for our clients.

11:01:
, you know, the fact that it’s evolving doesn’t mean that you can overlook traditional SEO because that still matters.

11:07:
It’s sort of foundational, but it’s, you know, it’s shifted some of the meta, and I mean that in the website sense, focus, , you know, towards keywords, , towards focus words rather, instead of some of the other elements, because in the end, You know, you do want to get picked up by the major AI aggregators because at a minimum, like if, if you’ve got all the right of that rest that SEO, you don’t want your words to be recomposed, you don’t want to be down in a footnote citation.

11:39:
You want your company brand to be expressed there.

11:41:
So the more you’re paying attention to classic SEO tools for doing that, the more likely you are to show up in those answer engines, even though, as we said, That’s not gonna be as much impact as regular content marketing for inbound looked like 5 or 10 years ago.

11:59:
Mhm.

12:00:
Yeah, definitely agree on that.

12:02:
The foundational things I, like you said, do matter.

12:06:
It’s just like one more dimension that you have to apply.

12:09:
, is it good for the new emerging large language models and platforms?

12:16:
, they do consider your experience, authority, trustworthiness, all those factors, and it’s not just limited to your own web property or your own website.

12:26:
It also really looks into your presence in let’s say Reddit or YouTube or.

12:33:
So they are actually training the whole model from all the different data sources.

12:37:
And if you’re repairing those and you’re participating and you have that consistent point of view and a brand or a category that you map towards.

12:47:
, very likelihood that you will show up on the, those, those searches and questions that matter to you and it’s very important to have that brand recognition, , and, and definitely there’s a process that you need to follow, , that is far more, I would say intense and in depth than, , the, the vanillaio that we’ve been all doing in the past.

13:11:
Yeah, there’s an interesting chain in the pre-pipeline there.

13:15:
I would say, one of the things that we studied is last week, , we put a blog post out about a specific piece of emerging technology, co-pilot notebooks.

13:25:
, and when we do that, we put it across all the traditional channels, .

13:31:
Google Bing, Insta, Facebook, LinkedIn, .

13:34:
You know, I play, I still play along with threads and Twitter and Blue Sky, even though I think those are sort of dead, but that’s just my personal bias there.

13:44:
What’s fascinating is we put that out there and we’ve got authoritative other things to reinforce it.

13:49:
We saw that post get hit about 550 times by AI engines in the 1st 24 hours alone.

13:58:
, so that’s a, you know, that is the next step you need to do.

14:03:
You know, I don’t know that I have perfectly have the code for making sure that That is showing up not just with authority, but with brand authority, , but , we’re getting there.

14:16:
Definitely.

14:17:
And how do you kind of balance between, like, you know, investing in AI driven experimentation with delivering kind of consistent measurable business outcomes because AI is so evolving, like, you know, we can be sleeping and something new has come out the next day.

14:35:
So it’s definitely something that does consume your time, but how do you kind of still maintain the balance between the two?

14:46:
Well, one of the ways that I try to do it, and I have to be personally careful about it, is I’m, I’m, I’m a technologist by nature, you know, I, I have multiple patents.

14:55:
I, you know, , helped shape products at the companies I’ve worked at.

15:01:
, and so, I love to tinker, kind of the way that my grandparents would go out in the garage.

15:06:
You know, it is great to be able to, OK, there’s a new LLM.

15:10:
Let’s talk through, like, how do I build an agent around this?

15:13:
How do I , use this in a content marketing strategy.

15:17:
, it can be really fascinating, but I have to reserve that for maybe only 1 day a week because there are still baseline elements, like we’re going to a conference in 2 weeks.

15:27:
, if you’re listening to this in 2 weeks, 2 weeks from now, I may not be that responsive because I may be delivering a session.

15:34:
, we can’t overlook kind of that classic side of making sure that We have a team ready to staff our booth, that the exec meetings for myself and the founders are being lined up, that we’re doing all of the right messaging around that.

15:49:
It’s got to be balanced because you’re right, AI can be so all-consuming, and for me, so fascinating.

15:56:
You know, I could start right now and look up and it would be 11 o’clock at night.

15:59:
It’d still be bright here in the Pacific Northwest that at this time of year, but you get my point.

16:05:
Art, what have you found?

16:06:
Like, how do you balance You know, the, the newness, the novelty of AI with the practicality of continuing to drive the business.

16:17:
Yeah, it’s definitely very interesting because on one side, AI has come with the promise of driving efficiency, doing more with less, but at the same time, we all, we all are feeling, feeling very exhausted and we all want to be up to speed.

16:36:
I think, , , we all are now like a kid in the candy store.

16:42:
With the new shiny object or is attracting us.

16:47:
So at the very base level, I think curiosity, , we should definitely appreciate, , and, , really try to mess up with some of the new technologies and learn from doing things versus just reading things.

17:03:
And as a leader, I think it’s more important that you have that kind of team in place.

17:10:
Which is not really very stringent about the way things have done, , or have been worked in the past, but trying out new things, be curious, be, be that stupid person who always want to start from zero.

17:25:
If that’s happening, I think that is the, the need of the R where we’ll have to go like.

17:31:
some time and really understand what’s working, what’s not.

17:35:
And, , and based on that, really consulting our clients, , show them, , what’s possible, , , by doing things, right?

17:46:
What’s interesting there, you know, my advice to, , to marketing leaders who haven’t done this yet is, you know, have lunch with your CIO.

17:54:
Make sure you unders you know, they can be, the technical side of your organization can also be a great source of guidance because it can be daunting to look at, you know, how month to month, which particular model is getting ahead.

18:08:
, you know, is chat GPT the best way to generate content?

18:13:
When should I jump to runway?

18:14:
Is mid journey still relevant?

18:16:
What about some of the other, all of these other content generation pieces which have become essential for marketing, , you can Put too much time into that level of experimentation.

18:29:
And in, similarly, you know, I was struck by this art when you were talking about it before, in the same way as that, , like, there are emerging standards in email.

18:40:
That IT may look at and not understand the impact it has with marketing.

18:45:
So that’s an important dialogue to have because in certain ways, if you’re making sure your outbound email is branded using the Beanie standard, BIMI, you’re going to do better at getting through some of those engines and people who don’t.

18:56:
That’s not something that IT might even think about.

19:00:
So that’s a good place for the dialogue.

19:02:
That’s a side piece of it.

19:04:
As a marketing leader, talk to IT, get their perspective on , first and foremost, what they’re seeing and what they’re hearing, so you don’t, you know, repeat experiments that someone else has already taken.

19:17:
And to be fair, also understand what are the, you know, ethical and compliance standards for what the organization is using, .

19:25:
We in MyOg, we have a fair bit of our client base is in regulated industry.

19:33:
And so we have some documented standards about these are the IT tools that we use the IT.

19:38:
These are the AI tools that we use and disclose.

19:41:
These are the ones that we treat as experimental.

19:42:
These are the ones that are off limits.

19:44:
And so, you don’t want to inadvertently, you know, as a con contributor to revenue, post pipeline.

19:52:
Have revenue disappear because you’re using tools that you told your clients you wouldn’t use.

19:58:
Definitely.

19:59:
I think that’s such an important piece, like, you know, as AI tools are increasing, , kind of relying on the first party data, how do you kind of manage data governance, especially when feeding models sensitive data or like, you know, behavioral data.

20:16:
, I know you shared Chris some examples of what you are kind of doing with some of your clients, but, , there are so many companies out there that aren’t even thinking about the data compliance side or like the governance side.

20:31:
So how important is that aspect, , specifically, like, you know, , for companies, like I know EU has like really , strict standards still, but there are other countries that are more open to experimentation, but there are also certain risks that people should be aware about.

20:50:
So what are your thoughts on that?

20:54:
Well, first and foremost, understand kind of where the trust boundary exists in your organization, the difference between your content, everyone else is out there on the internet.

21:05:
I was talking with a colleague last year about Related AI project.

21:10:
He’s was considering building a contract management engine.

21:15:
The problem is, if you go to a chat GPT or any of these consumer grace grade tools, they’re free, they’re fun, but they understand generic information, they don’t understand your information.

21:30:
Like AI, did you see the Volvo for Life commercial that was going around last summer?

21:35:
, no.

21:37:
So it, it was an interesting exercise in what AI can do.

21:40:
So, , and we can get a link to this in the show notes, but you can, the audience, you can just look up Volvo for Life and you see an ad, looks like something you might see during a sports broadcast.

21:54:
, Someone gets into a Volvo in a dark gray city, and they, as they drive through the city, everywhere the Volvo goes, the city bursts to life and trees start sprouting and flowers.

22:05:
It’s all about how Volvo is taking an environmentally responsible approach.

22:09:
That ad was made entirely with AI.

22:12:
Oh.

22:13:
Now, and they did it within 24 hours.

22:15:
There’s some technical things that are worth noting that they didn’t just give it one prompt and the video showed up.

22:22:
They did a lot of prompting to build each individual shot, and they assembled it, but The way that they did that was based on the fact that it was done by Volvo’s ad agency, and they had access to 30 years of brand image and brand identity and language, and so on and so forth.

22:38:
So getting that information that’s unique to you or to your client, or however you, you’re an agency, it’s important to focus on that.

22:46:
So, let’s not get to the technical level.

22:48:
, you can go to the extreme and have someone build a custom model for you that is trained on Your data in a very precise way, and you might get there.

22:59:
The other extreme is going to an open chat GPT where you’re commingling your data that probably doesn’t make sense.

23:08:
The guidance I give to lots of organizations is look for the AI that runs inside of your organizational trust boundary.

23:16:
And whether that’s Google Gemini or Microsoft co-pilot, Those engines are going to, first and foremost, work on what you give it and what you have.

23:28:
As long as getting back to that IT partnership.

23:31:
If IT has well regulated governance standards for security, for, for retention, , those sorts of things play in well to a Gemini or to a co-pilot because they respect the underlying security of those tools.

23:47:
, that’s my Fairly verbose guidance on that.

23:51:
I haven’t given you a lot of space, Arpit.

23:53:
You should, next question that Hershey comes with, you should probably take first, but would love to hear kind of your perspective on it as well.

24:01:
Yeah, like, in my view, , I think you touched upon models that there are open source options as well, , that can be basically hosted, , within your, your own, let’s say on-premise or even your cloud instance, and you sort of manage it and, , you can always add a layer of fine tuning it and, , all that is, it’s owned by you.

24:27:
So I think right now models have become a lot more like a commodity.

24:32:
They will keep on improving and there are open source side of it as well.

24:37:
It is just a matter of how you’re utilizing it and if, if security and those kind of things are concerned, you better own that end to end.

24:46:
, by picking any of the open source, , choices that are available in the market.

24:52:
Right.

24:52:
I do kind of have a follow up to this.

24:55:
If an AI tool does make a mistake, Who’s accountable?

24:59:
Is it the vendor?

25:00:
Is it the marketer, the engineer, or like, you know, how should marketing teams have that sort of accountability structured within their teams when they’re using AI?

25:14:
I would definitely say that it should be part of the process that .

25:20:
Humans should be in the loop, right?

25:23:
You can’t just give an open ATM card and then let AI drive everything because that to me is a recipe of disaster.

25:33:
And more importantly, from a marketing perspective, again, you’re not owning your story, you’re not owning what you want to focus.

25:40:
That means that derails everything.

25:43:
You, you’re just part of the noise.

25:45:
So in that sense, , there should be process and governance, , and, , things like that in place that really, , is very clear to the entire team how they are utilizing it and, , and then the final go to or, , the, the approval process should be, should be given to a person who owns let’s say the brand or a particular department, whatever it is.

26:12:
And kind of with the shift of like, you know, Everybody kind of moving towards agentic AI and agents kind of, you know, now we’re doing tasks without like generative AI was great.

26:25:
You can use it, you can, you still have that decision making.

26:28:
You can edit things and then it kind of, you know, moves forward.

26:31:
But with a lot of the focus now being on agentic AI, how do you think this kind of plays out and you know, maybe Chris, you can start off with your thoughts on this.

26:42:
Yeah, I don’t think the accountability has changed when it comes to agentic AI.

26:46:
So if you look at, you know, what’s been in vogue the first half of this year with what OpenAI did with the 03 engine, which is now showing up as researcher and co-pilot, and there’s parallels that are happening with Anthropic and with Gemini and all the rest.

27:01:
, it’s able to do a lot more and it can bring in more sophisticated capabilities, but as a result, it can also magnify.

27:11:
That mistake.

27:12:
And here’s a very recent example.

27:13:
So, I gave some coaching to my partner, our CEO on how to use AI to construct a LinkedIn follower campaign.

27:26:
Generate a series of posts, you should go through and edit them and here’s how to, you know, here’s how to coach it, here’s what to get.

27:32:
You tell it what the subject is and numbers, and she went off and she built this and came back was, fortunately, she asked for help on Getting it ingested into our social engine.

27:44:
And I took a look at the post that it generated and I’m like, you know, It’s got the same graphic every day for 30 days.

27:50:
And some of these first comments on LinkedIn, like, they point to imaginary articles, like, you know, they said, you know, it, saying, for more information, download our change management tool kit from this link on our website.

28:05:
I know perfectly well that it’s not something that is on my website.

28:09:
So in the end, , you know, I, you could say it’s the technology’s mistake.

28:14:
but, you know, becau just because now, it’s easier to use AI at scale to do 30 or 40 or 50 things.

28:22:
It also means it can amplify those mistakes 30 or 40, 50 times.

28:25:
Yeah.

28:27:
As a marketer, whether it’s, so whether I was working with a college intern or an AI intern, as a marketing leader, you know, one of my bits of guidance to people is, you’re the, you know, you’re the editor in chief.

28:38:
You know, you’re supposed to be going in and providing the final level of quality control.

28:41:
So I do think that accountability is centered on the CMO role.

28:47:
Definitely.

28:49:
I’d love to kind of shift the gears now to kind of talk about some metrics when it comes to marketing.

28:55:
I know there’s been a huge shift lately from, you know, marketing being a cost center to sort of more of an investment, and then once you kind of talk about investment, ROI is the next thing that comes into play.

29:07:
So what are certain metrics like, you know, or maybe one metric in your GTM funnel that actually improved because of AI and how did you prove it, , if you’ve had any recent examples of that?

29:22:
Hm, .

29:25:
One of the metrics I look at, , And it’s a little bit different for my book of business because we’re in management consulting and we’re not, we tend to not run at the hundreds of thousands to millions volume, , but , We look at Cost per sale is accepted lead by source.

29:51:
And which is, I, I’m a quan geek by background, so I made a couple of adjustments, some of those classic calculations.

29:58:
And we can within that kind of do the equivalent of A testing that across.

30:08:
Some things that are AI enhanced and some things that are not to see what the impact has been.

30:13:
, I think, you know, part of the problem is some of the telemetry because of that.

30:19:
Cause there’s now a level of of disintermediation that AI is bringing in the consumption side.

30:24:
It can be a little bit harder to attribute that all the way through.

30:30:
Like, was, , Let’s talk about a sales journey for an individual customer because I think it gives you some insight as to what to measure.

30:40:
So, , there’s a client that we’ve signed recently who’s in Funnel for 9 months.

30:47:
And, you know, at various points in time, they received communications that have been developed in part with the AI.

30:54:
, they jumped into our ABM strategy, which is a place I think AI has tremendous impact these days.

31:01:
, And so within all of that, like, but I mean, in the end, what closed them was that willingness to kind of Fly down to meet with them and to get the deal done.

31:18:
.

31:21:
If AI had never happened.

31:23:
You know, debatably, that last meeting doesn’t happen either.

31:28:
So that’s why like getting at least to the sales threshold is where I think you need to measure AI, but it’s all nuanced.

31:34:
Is it, you know, , did he buy because of something he, you know, in psychographics that he read three months earlier and forgot that he’d heard it from us, but it was still buried in there.

31:46:
, You know, I don’t know, but I do think that being able to AI makes it easier to sustain a volume of high quality content in in an ABM strategy that would be hard to do with people.

32:02:
Definitely maybe you can answer this.

32:05:
Are there metrics we still track kind of out of habit, but no longer really hold any value to the business, especially when you know, we’re going more forward with like AI powered GTM models.

32:19:
So are there metrics we should stop kind of looking at now?

32:23:
Yeah, let me come back to, , it, but before I answer this, , to your previous question, , like one of the, , category of campaigns that, , that is actually being powered by AI, , so Google has come up with a performance Max campaign.

32:43:
, so, as you know, AI is only as good as the data you feed it into.

32:50:
And for running those kind of campaigns, , you need like a good conversion data.

32:57:
So what we realized was that the client was not even collecting those right conversion metrics.

33:04:
And that’s why the campaign was not never got that data to understand what’s working and what’s not.

33:11:
And you, if you don’t feed that, , that data, then the campaigns would not really, , get tuned to where should they expand like doing that look alike modeling and reaching out to the right audience.

33:25:
So, , the, the, the underlying, , prerequisite to leverage AI and running campaigns that require data is that first of all, collect data, , in the right manner and then feed that, , to the algorithms so that they understand what’s working.

33:44:
And then only the AI magic will kick in and that’s what we, we actually have seen.

33:49:
Once you have those things in place where, , The algorithm is getting the feedback loop of the conversions that matter to you.

33:59:
It could be the sales qualified lead or maybe a portity was created that threshold and not necessarily a typical formfield which could be a junk lead.

34:09:
The moment you start doing it, the campaign starts really kicking in and really looking for similar audience that can work better than our, , like high quality leads.

34:20:
So for me, , lead quality as a metrics.

34:24:
, and your ability to understand what is the lead quality, , and more from a, , predictive, , analytics perspective where AI or machine learning is telling you whether it’s a grade A versus grade C is, is still very important and very relevant, and it’s relatively a lot more easy because, , there are like platforms that support the same.

34:47:
, to your next question, like what probably would not matter much.

34:54:
So for example, like in the last 1015 years, one thing we were always excited was, are we ranking number one on this keyword?

35:03:
And we will have those weekly and monthly reports around ranking whether we are ranking for those keywords or not.

35:09:
So we’ll have to take a step back because maybe those informational keywords generating content around those is not going to be meaningful anymore.

35:20:
, rather you focus more on your main category and commercial keywords.

35:25:
Don’t worry about the ranks, but as long as you are getting a brand recognition in some of these emerging models.

35:32:
, I think you, you as your team or your, , generative, , , , generative optimization, , team is doing a good job.

35:42:
So we’ll have to really look, , from each department perspective, what has, , been the, the, , the KPIs that matter earlier, , and then just rethink about, , the new KPIs.

35:58:
I’ll just give you one example about Search, but I think similar, similar dimension we have to apply to each of the channels and how they work.

36:10:
At the end of the day, what really matters is, are you able to understand the kind of activities that are leading to the close one.

36:22:
And what are those activities, , and are they influenced by search or is it social and then really double down on the channels that are leading to activities from users that in turn become, , the close one and you, you have a closed loop reporting, , then I think your marketing will be in good shape.

36:46:
Yeah, it’s just hearing you say that.

36:50:
, you’ve given me an idea I wanna play around with tomorrow when I have some downtime.

36:55:
And so I’m gonna start with storytelling cause we’re marketers, we like to just tell stories.

36:59:
, so, have either of you ever coached baseball?

37:03:
Mhm.

37:04:
Not me.

37:05:
So I coached, I coached 17 seasons of youth baseball.

37:09:
, and so, baseball as a fascinating sport, but as a data science, because there’s so many pitches and so many events, it generates a lot of data.

37:20:
And so there’s, for decades, there have been a lot of quantitative geeks who go through and analyze things, and they invent statistics.

37:27:
And so, when I was coaching, I tended to be, despite the fact that it was the youth baseball level, very quant driven.

37:33:
And so, I devised a stat called SIP.

37:36:
SWIF is the sigma of walks and hits per innings pitched.

37:40:
What it’s telling me, what it told me in that case, is What are the odds that a given pitcher is suddenly gonna go crazy and give up 10 hits in an inning and have one of those, if you’ve ever seen Little League or Pony League, you see these innings where suddenly everything goes wild.

37:58:
And I discovered that I had one pitcher on my rotation, who didn’t look like a great pitcher, but he also never went.

38:07:
wild.

38:08:
And so it uncovered some of what those gems are.

38:11:
I think looking at the points you were making our Pit about lead quality, , and sales acceptance, I think cross-capa, looking at the sigma of those is gonna tell you the campaigns, let’s assume all the contents squibbling.

38:26:
The The products and messages that you had 3 years ago are gonna connect differently in this era.

38:33:
And what might have been the winners 3 years ago are not gonna be the winners you have now.

38:38:
And so, , you know, if I am seeing, you know, that some campaigns are Consistently delivering results as opposed to, you know, that high variability means that that campaign got lucky one day and unlucky one day and may not be as reliable as the ones that are consistent.

38:58:
So, , you know, I would say, , Sigmae and Sigma MQL are gonna be the ones I’m gonna look at the most over the next 6 months.

39:11:
Yeah.

39:13:
Definitely.

39:14:
, now that we’re kind of coming towards the end of our discussion here today, I do want to transition to what I like to call a lightning round.

39:23:
So this is gonna be super fast, super quick, something that kind of stems on top of your head when I kind of share the question, and it doesn’t have to be long, just a quick response, one or two words or maybe a sentence of what you kind of think of what I’m asking about.

39:41:
So.

39:42:
This is a little tough one, but I think it’ll be good to kind of, you know, get your honest thoughts on certain things.

39:49:
So let me start up by, , you know, one overhyped AI trend in marketing.

39:56:
What would that be?

39:57:
Maybe Chris, you can start off.

40:03:
Agents.

40:04:
Agents.

40:06:
Definitely of it.

40:09:
More, , generating more content.

40:12:
Mhm.

40:13:
And one AI or analytics tool that you won’t work without.

40:22:
Researcher.

40:24:
Research.

40:26:
For my PPTs, , I like napkin.

40:29:
AI.

40:30:
It really helps create those graphics which I used to depend on the design team, no, not anymore.

40:37:
Mhm.

40:38:
, contents and let’s talk about what should marketers stop doing with content in 2025.

40:48:
Indiscriminately pushing links to everyone in your contact database.

40:54:
That’s a good one.

40:56:
I.

40:57:
I think volume doesn’t matter, quality matters.

41:01:
Mhm.

41:02:
And what one human skill that AI will never replace or at least not in the next 5 years, let’s say.

41:12:
And the curiosity.

41:15:
Empathy, oh, curiosity is great.

41:17:
Yeah, those are both great answers, , I guess my last one in this section would be.

41:24:
In one word, AI and marketing should feel, fill in the blank.

41:30:
What should it feel like AI in marketing if you could just use one word?

41:37:
Co-pilot?

41:39:
Yeah.

41:41:
Yeah.

41:42:
, It should feel Suggested.

41:49:
Mhm.

41:50:
Definitely noted.

41:53:
Yeah, that was an amazing session.

41:55:
I, that was really some punchy and crisp responses.

41:59:
So thank you for that.

42:01:
, before we kind of wrap up here, I do want to ask you guys for, you know, any last piece of advice that you would have for marketers out there that are still struggling with AI, still kind of, you know, , and figuring things out, anything in terms of marketing that they should keep an eye out for in terms of what’s next to come, or, you know, something they should keep consistently doing.

42:25:
So any advice for marketers out there.

42:28:
, I would say that the, the fundamentals are still true.

42:35:
, for example, the old age definition of marketing where it’s all about reaching out to the right people with the right message at the right time.

42:45:
It still holds true, but now you have this amazing AI technology that can be a force multiplier and really help you get more right people.

42:57:
, I will look for an AI that can give you a cohort of users that are really good fit for your ICP, a model that can tell you that can convert maybe 80% of chances.

43:08:
I mean, those information was not there.

43:11:
And then once you have the right people, it’s about the right message.

43:16:
So again, Applying that data and AI to really craft a message that not only tells your brand story, but also really personalized based on whatever you know about that company, about that person that you’re reaching out to.

43:32:
And then the reaching out in a right moment again, an AI agent that is really listening all those signals and alerting you that look, a change has happened in this organization, let’s, let’s do a quick communication.

43:47:
So I think all those things still apply.

43:51:
AI is a leverage and we should definitely embrace it, but don’t forget the, the, the foundation, the fundamentals, your own brand story, your own positioning, your messaging.

44:04:
All those concepts are still true and will remain true.

44:08:
Definitely, Chris, your thoughts?

44:12:
I don’t think agents are gonna replace marketers, but marketers with agents are gonna replace marketers without them.

44:18:
And I think there needs to be, , from a marketing perspective, you’re not in IT, but you need to understand what’s emerging and what’s quality.

44:26:
You know, there’s an agent that’s been exist around in marketing for 10 years.

44:30:
We didn’t call it an agent.

44:32:
But everyone knows it and everyone is frustrated by it.

44:34:
You dial into an organization and so a robot gets on the line and says, before I connect you with someone, why don’t you tell me what you’re looking for to see if I can help you.

44:45:
And what most people do, is figure figure out, what do I need to say to get to talk to someone?

44:51:
Because those agents usually get it wrong.

44:53:
We see the same thing on websites.

44:55:
I personally think that putting agents directly in communication with my prospects and clients is a poor idea.

45:03:
, but agents are not quite there yet, but in 2 years, they are going to be able to take on Offloading much more complex tasks for a marketer than is possible today.

45:17:
So study with that, study that, be cautious, but that would be, I would say, how marketers should think about AI technology moving over the next two years.

45:27:
Definitely.

45:28:
Thank you so much to both of you.

45:30:
This has been such an incredible conversation.

45:33:
I know I have personally learned a lot, so I’m sure our audience will too.

45:38:
, just, you know, as we close off, we would love to, you know, kind of, , for our audience if they kind of want to connect with you, Chris, how can they find you?

45:49:
What’s the best way to kind of reach you?

45:52:
Probably the simplest, , my company Synier is harder to spell S Y N O G U R.

45:56:
So I’ll say, yeah, you, you can find me on my personal website, chris McNulty.net.

46:01:
That will connect you to my company, to my LinkedIn, to whatever you need to be, to my upcoming schedule, places where I’m gonna be speaking and touring.

46:10:
Awesome.

46:11:
Thank you so much.

46:11:
And you are, what’s the best kind of way to reach you?

46:15:
Just connect me on LinkedIn or drop me a note on info at the dipro.com.

46:20:
Mhm.

46:20:
Awesome.

46:22:
, again, this conversation has been so exciting.

46:24:
We’ve covered so many different topics and I’ve kind of grilled you guys with some questions, some quick answers.

46:30:
So thank you so much for being patient with me today and for our audience out there, thank you for tuning in.

46:37:
, we’d love to get your reviews on this session, and we’ve got tons of more, you know, thoughtful leadership experiences and conversations coming up soon.

46:46:
So definitely stay tuned.

46:48:
And thank you once again.

46:50:
Thank you.

46:51:
Take care.

46:51:
Thank you.

See more webinar