Moving Faster Without Breaking Everything - AI, Risk, and the Human Side of Change with Andrea Tarrell

WEBVTT
00:00:24.810 --> 00:00:25.770
Hello, everyone.
00:00:25.770 --> 00:00:31.289
Welcome to another episode of OpsCast, brought to you by MarketingOps.com and powered by all those MoPros out there.
00:00:31.289 --> 00:00:32.730
I am your host, Michael Hartman.
00:00:32.730 --> 00:00:40.250
Today I'm joined by Andrea Atero, president of the tech services line at Trilliad and CEO of Circante, a Trilliad company.
00:00:40.250 --> 00:00:48.490
Andrea has spent years helping organizations navigate marketing operations, Salesforce ecosystems, data strategy, and large-scale technology transformation.
00:00:48.490 --> 00:00:54.409
But our conversation is going to be less about the technology itself and more about the human side of change in the AI era.
00:00:54.409 --> 00:01:09.930
We are going to talk about why so many organizations are struggling to move quickly, how fear and risk tolerance are shaping AI adoption, why governance models may need to evolve, and what separates organizations that are actually activating AI from those that are stuck debating it.
00:01:09.930 --> 00:01:18.009
Should be a thoughtful conversation about speed, trust, process, and the trade-offs companies are making right now as AI shapes how go-to-market teams operate.
00:01:18.009 --> 00:01:21.530
So with that, Andrea, welcome to the show.
00:01:22.009 --> 00:01:23.609
Thanks for having me.
00:01:24.009 --> 00:01:24.969
Yeah, it's great.
00:01:24.969 --> 00:01:27.129
Well, it's been a while for us to get this started.
00:01:27.129 --> 00:01:38.409
So when we talked before, um, you know, that we've talked about that for years, the hard part of operations work was building technology.
00:01:38.409 --> 00:01:46.969
But now it seems like technology is getting easier, at least more democratized, maybe, while people are, you know, while the people and process side is getting harder.
00:01:46.969 --> 00:01:49.849
So A, do you think that's true?
00:01:49.849 --> 00:01:56.090
And then B, like, what do you think has fundamentally changed that has kind of led us to that spot today?
00:01:56.810 --> 00:01:59.689
Yeah, we we see this everywhere.
00:01:59.689 --> 00:02:12.889
Um, and obviously, kind of being in the technology consulting space, like we have a lot of respect for um like the very smart technical minds and technologists that build awesome solutions for customers.
00:02:12.889 --> 00:02:18.569
Um, but that soft side of change, like the human side, um, is more relevant than ever.
00:02:18.569 --> 00:02:26.329
And I think we're seeing that right now with AI and just like convincing people to use the tools and to its fullest potential.
00:02:26.329 --> 00:02:30.009
Um, that's where a lot of the hard, hard work is happening right now.
00:02:30.009 --> 00:02:37.530
And it's not a limit of the tech platform that's holding us back, but um just kind of where the rubber meets the road with the team.
00:02:37.530 --> 00:02:52.169
And um, something that illustrated this for me recently was like, I think the like the stereotypical tech founder used to be like the antisocial dude in a hoodie that like was a coding expert and was respected by all the developers.
00:02:52.169 --> 00:03:08.729
Um, but the Wall Street Journal had a headline last week where um Silicon Valley VCs are sending their their tech founders to etiquette school to like figure out how to talk at a cocktail party and uh social etiquette and social norms.
00:03:08.729 --> 00:03:19.530
And um, I just I I think that that like really emphasizes how important it is to pair the human element with the tech element, uh, if you really want to have change that sticks.
00:03:19.849 --> 00:03:20.729
Yeah, it's interesting.
00:03:20.729 --> 00:03:22.090
Uh it's funny that you bring that up.
00:03:22.090 --> 00:03:29.370
I didn't see that headline or that article, but um having kids who are in that kind of college age, right?
00:03:29.370 --> 00:03:38.330
I think it's interesting to see how the interactions for them are different than what I remember when I was their age.
00:03:38.330 --> 00:03:42.409
And um, you know, yeah, free internet and all that kind of stuff.
00:03:42.409 --> 00:03:47.849
And there's and there's all kinds of books like the I just finished reading a few months ago, The Anxious Generation.
00:03:47.849 --> 00:03:50.490
Well, trust be told I didn't read it, I did the audio book.
00:03:50.490 --> 00:03:55.449
But I, you know, and I I told my kids, it's like you need to read this because this is about you, right?
00:03:55.449 --> 00:04:04.169
This is but a big part of that is about how um just in general, like communication is not happening on a regular basis.
00:04:04.169 --> 00:04:09.050
There's another article I'd I'd referenced a couple of times in recent podcasts that somebody shared with me.
00:04:09.050 --> 00:04:33.769
Now I haven't dug into the research, but somebody's done some research that says it's something like over the last 10 years, and I don't remember exact 10 years, it wasn't for 2026, but say 2024 or something, there um there's been a like a 30% reduction on the average number of words spoken by people, and I think it's fairly global, if I remember right, but certainly in the in the in the North American space.
00:04:33.769 --> 00:04:34.970
So all that makes sense.
00:04:35.210 --> 00:04:36.730
Well that's so fast fascinating.
00:04:36.730 --> 00:04:55.049
And and honestly, like the I think like ops folks have been on the front lines of this for for a long time, but like you can build the most perfect tech solution in the world, and if people don't use it the way that you expect, or you can't win over hearts and minds with the change that you're looking to drive, like it's just it's never gonna succeed.
00:04:55.049 --> 00:04:57.929
And I think we're just we're seeing that now more than ever.
00:04:58.250 --> 00:04:59.610
Yeah, I I think that's true.
00:04:59.610 --> 00:05:09.129
I mean, it's a combination of this like rapidly evolving technology plus so these other so social economic kind of things playing into it.
00:05:09.129 --> 00:05:11.370
So so oh so let's go there.
00:05:11.370 --> 00:05:31.049
Like, so how much of this is is for AI adoption, then I guess in particular specifically, is like how much of it you think is a technology one versus a trust one or change management, or is it something else completely that's like causing it to be a little bit because it feels like it's very uneven how it's being adopted across different organizations?
00:05:31.529 --> 00:05:36.009
Yeah, I I think it's primarily a change in a leadership problem.
00:05:36.009 --> 00:05:42.809
Um I mean, okay the the tech tools are like they are what they are, they're not perfect.
00:05:42.809 --> 00:05:48.649
There's certainly like still gaps in LLMs and some of the the tooling that's available to us.
00:05:48.649 --> 00:05:56.970
But in general, like the potential of the tools has far outpaced our ability to effectively leverage them within our within our organizations.
00:05:56.970 --> 00:06:08.570
But I think too many leaders are are coming at AI with the angle of like, okay, this we're gonna use this to eliminate jobs, or we're gonna use this to like work so much harder.
00:06:08.570 --> 00:06:14.649
And um, it's just it's not really falling the they're not striking the right tone with their team.
00:06:14.649 --> 00:06:22.090
And individual contributors are looking at their heavy workload and the tools that they're being provided to do it.
00:06:22.090 --> 00:06:26.570
And when they're asked to innovate with AI, they're thinking, like, who is this for?
00:06:26.730 --> 00:06:30.090
Like, how many world am I gonna fit this into my day-to-day job?
00:06:30.169 --> 00:06:33.529
Well, I'm yeah, like is this so the CEO can get a raise?
00:06:33.529 --> 00:06:37.370
Is this so I can like like produce more hours?
00:06:37.370 --> 00:06:39.450
Like, like who is this for?
00:06:39.450 --> 00:06:49.049
And um, I feel like there's not enough messaging and communication and like trust building around like this can be a tool for all of us.
00:06:49.049 --> 00:06:57.210
Like it doesn't have to be a tool that lets you turn 40 hours of work into what used to take 60 hours to to output.
00:06:57.210 --> 00:07:06.330
Like this can be a tool for us to all be more effective, to live more balanced lives, um, to generate better outcomes for the company, but also for individuals.
00:07:06.330 --> 00:07:14.409
Um, and I I feel like we're we're overdue for a narrative shift around like what is AI and what who is it really for?
00:07:15.049 --> 00:07:15.289
Yeah.
00:07:15.289 --> 00:07:18.730
I mean, and it feels that point about individual contributors.
00:07:18.730 --> 00:07:21.610
I don't even think it's a due to individuals contributors.
00:07:21.610 --> 00:07:24.009
I think also middle managers are under a lot of pressure.
00:07:24.009 --> 00:07:24.570
Yeah.
00:07:24.570 --> 00:07:40.250
Uh I think and you know, it feels like we're recording this late May 2026, and it feels like in the last few weeks, it feels like almost two, three, four a week, there's announcements of companies cutting workforce.
00:07:40.250 --> 00:07:45.929
And on the surface level, they're saying a part of that is AI driven, right?
00:07:45.929 --> 00:07:58.490
So if you're coming in as a leader saying we need to be doing more with AI, and you're hearing as a person in these organizations about all these jobs being lost, like, like why would I spend the time if what you're trying to do is eliminate my role?
00:07:58.889 --> 00:08:00.409
Yeah, a hundred percent.
00:08:00.409 --> 00:08:07.850
And I I I feel like those headlines of the mass layoffs because of AI are are not doing any favors for the cause.
00:08:07.850 --> 00:08:15.289
Um because the vast majority of those of those layoffs don't even really appear to be AI or efficiency led.
00:08:15.289 --> 00:08:22.250
It's just like a convenient label to excuse why you're having to trim trim your workforce.
00:08:22.250 --> 00:08:35.850
And I think that that like AI is gonna take your job narrative, like has a lot of team members rolling their eyes because it's like, okay, the person in the C-suite that's making that proclamation doesn't even know what their day-to-day job is.
00:08:35.850 --> 00:08:36.570
Like, right.
00:08:36.570 --> 00:08:44.090
So, like, how how are you gonna automate away a job that you don't even understand or don't even like don't even understand why it exists?
00:08:44.090 --> 00:08:58.250
Um and thinking about AI adoption in an organization, like I um I think it's less about like resistance to change in general and more like how how do we give people a reason to want to change?
00:08:58.250 --> 00:09:02.569
Like, what's the pot of gold at the end of the rainbow if if folks are leaning in?
00:09:02.569 --> 00:09:05.610
And how do we create space for for that change?
00:09:05.610 --> 00:09:25.370
Because if your plates already full, you have like X number of hours to execute like the tasks that are on your plate, you're you're not gonna be able to pull up and spend an hour or two like redefining your workflow or testing different LLMs or like researching new tools or learning new things.
00:09:25.370 --> 00:09:29.129
Uh we have we have to create that space for learning and evolution.
00:09:29.129 --> 00:09:35.449
Um, and it it can't just fit in the cracks of like what else is happening day to day.
00:09:35.850 --> 00:09:37.049
I would love to get your take.
00:09:37.049 --> 00:09:43.209
So there was somebody who I admire, uh CMO on LinkedIn, I guess it was last week.
00:09:43.209 --> 00:09:56.809
It may have been over the weekend, but certainly within the last week, it posted something about how like adding that that extra layer of work on top of an already busy person is not likely to be successful.
00:09:56.809 --> 00:10:02.809
And the claim he was making was that essentially feels it because I kind of pushed back.
00:10:02.809 --> 00:10:11.049
I was like, to me, this is not all that much different than other technologies that have come in and could innovate and make a job more efficient or whatever.
00:10:11.049 --> 00:10:14.649
And his take on it is like this feels fundamentally different.
00:10:14.649 --> 00:10:25.689
And I'm I'm torn because I don't know if it feels fundamentally different because it's like where we are in the middle of it, or if it truly is different than say, I don't know, the steam engine, right?
00:10:25.689 --> 00:10:27.209
If you go far enough back.
00:10:27.209 --> 00:10:28.649
What do you think?
00:10:29.289 --> 00:10:41.449
I one of the pieces that feels different to me is um like a lot of talk about like making work more efficient assumes that like the outcome is like the output will be higher.
00:10:41.449 --> 00:10:52.169
So like if I have four clients and now I can use AI to be 30% more efficient or save X number of hours, then maybe I can service eight clients in like the same amount of time.
00:10:52.169 --> 00:10:58.490
Um, but what we're seeing happen more often is like small tasks are getting compressed.
00:10:58.490 --> 00:11:01.929
So what used to take 30 minutes now takes five.
00:11:01.929 --> 00:11:08.809
Um and what's being done with that extra space is not more output, but higher quality.
00:11:08.809 --> 00:11:13.449
And that's really hard to quantify and really squishy.
00:11:13.449 --> 00:11:21.529
And so I think like measurement frameworks and how we think about productivity is gonna need to evolve to keep up with that.
00:11:21.769 --> 00:11:22.089
Yeah.
00:11:22.250 --> 00:11:32.970
Um because I like many of the stats that I see about from AI tools are like, okay, we took this workflow from three hours to whatever time frame.
00:11:32.970 --> 00:11:39.610
And um, but CEOs are looking and CFOs are looking at that and saying, like, where did that time efficiency go?
00:11:39.610 --> 00:11:49.929
Like, it doesn't mean we are all of a sudden hiring X percent less people, or people all of a sudden have eight hours of free time if they never had before.
00:11:49.929 --> 00:11:58.250
Um, like there's a very long backlog of things that can fill people's plates once things are made more efficient in the day-to-day.
00:11:58.250 --> 00:12:04.009
And so it feels like it's harder to capture some of those efficiency gains or measure it than it has been before.
00:12:04.329 --> 00:12:04.569
Yeah.
00:12:04.569 --> 00:12:16.089
I mean, to me, one of the things I've seen, or I I think would be a a benefit of sort of getting that more of that time back.
00:12:16.089 --> 00:12:27.929
So I'll take I'll put it in the terms of like a market, like a demand gen marketer, right, who's gonna build out campaigns and say they've got this great idea and it ends up working well, and they go like, oh, this is our audience.
00:12:27.929 --> 00:12:45.610
So rather than sort of segment that audience into smaller chunks, it is because there's time constraints, you sort of cast a wider net that's a little less, a little more generic, a little less on point for any one of those subsegments, and um, and then you kind of hope for the best.
00:12:45.610 --> 00:13:05.370
Where I think like now, if you could take that and go like, actually, I can now, because I've got more I've got a time benefit from using AI, I can then narrow those into four separate things that are all part of one big thing and get better messaging and better results for each one of those, like that's a multiplier effect on output or yeah, output, but results.
00:13:05.370 --> 00:13:09.209
Like, and I think that to me is like, but you're right, like how do you measure that?
00:13:09.529 --> 00:13:09.769
Yeah.
00:13:09.769 --> 00:13:22.009
How do you say that thanks to AI, like we move the needle on on click-through rate, or we move the needle on like opportunity influence for this certain content channel?
00:13:22.009 --> 00:13:27.529
It's um, it's really hard to draw those like super crisp conclusions.
00:13:27.850 --> 00:13:36.089
Yeah, and then you get into things if it's thinkers in the marketing rate, you get into the whole like is attribution and like what should marketing be measured on anyway, right?
00:13:36.089 --> 00:13:38.250
And who gets credit for it?
00:13:38.250 --> 00:13:43.610
And I think it it's like it it's a little bit of a circular problem that doesn't have a real good answer.
00:13:43.610 --> 00:13:52.409
I hadn't thought about the the that squishiness of be able to measure the impact outside of uh the efficiency one.
00:13:52.409 --> 00:13:55.289
I like that's a really interesting point there.
00:13:55.289 --> 00:14:11.209
Um, so do you like do you think that some of this challenges is really is it because this is such a different thing and that there's this fear of replacing jobs, or do you think it's just is there inertia in some of these organizations that is preventing that?
00:14:11.209 --> 00:14:17.449
Like to your point, like not creating the space to to build on that of those skills and capabilities?
00:14:17.929 --> 00:14:26.250
Yeah, I mean, I'm I'm torn because that there definitely are some parts of this shift that are like new and novel and unprecedented.
00:14:26.250 --> 00:14:32.490
Um, but there's a lot of similarities with like every wave of tech change that has come before it.
00:14:32.490 --> 00:14:48.809
And I think maybe the the primary thing that's different is it's just faster and the fear factor is dialed way up, like the the narrative around like you're gonna be left behind, you're missing an opportunity, your job's gonna be eliminated, you're gonna become irrelevant.
00:14:48.809 --> 00:14:55.449
Like, there's so much negative talk about it that it um it's just it's it's a lot of noise.
00:14:55.449 --> 00:15:00.329
So yeah, I'm I'm torn on like is it is it same old, same old?
00:15:00.329 --> 00:15:02.809
Is it like new and novel?
00:15:02.809 --> 00:15:04.569
Um, I think it's a little bit of both.
00:15:05.129 --> 00:15:05.449
Yeah.
00:15:05.449 --> 00:15:13.049
So like one of the things that I hear people say is AI won't replace your job, but you know, people who use AI more efficiently will could replace your job.
00:15:13.049 --> 00:15:19.049
Like, what's your do you think that's like I think we talked about this before, but like how do you think about that?
00:15:19.049 --> 00:15:20.889
Does that resonate with you?
00:15:20.889 --> 00:15:28.169
Um, and then how are you seeing that play out in reality with the way leaders are were working with their teams?
00:15:28.730 --> 00:15:34.809
Yeah, I I think that that is very, a very true statement when you think about it at the organizational level.
00:15:34.809 --> 00:15:41.209
Like, and we we've tried to talk talk through this with our our team internally at Circonte and Trilliad as well.
00:15:41.209 --> 00:15:45.289
Because if you're not evolving, somebody else is evolving.
00:15:45.289 --> 00:15:50.889
And we're not looking at our team saying, like, oh, if you're not using AI, we're gonna replace you with someone who is.
00:15:50.889 --> 00:16:01.529
But the reality is like our competitors are are racing to figure out how to be more efficient, be more nimble, um, charge higher rates, deliver the same outcomes in less time.
00:16:01.529 --> 00:16:06.569
And if they crack the code before we do, like we're all we're all gonna be out of jobs.
00:16:06.569 --> 00:16:09.370
Like we're we're just slowly gonna start losing more deals.
00:16:09.370 --> 00:16:15.529
We're not like so, and we can't control the future unless we actively participate in it.
00:16:15.529 --> 00:16:26.649
So we want to disrupt ourselves so that we as an organization don't get disrupted by um people that are more nimble and able to embrace change.
00:16:27.850 --> 00:16:40.649
So yeah, so and then there's this pressure, like it feels like everyone's being told like go go do quote, do do AI or yeah, incorporate AI into your day-to-day activities uh and to do it fast.
00:16:40.649 --> 00:16:49.929
But then there's seeing, I think then there's you hear stories about um IP being exposed and things like that, right?
00:16:49.929 --> 00:17:09.850
Um, so how how are you seeing people manage that I guess tension for lack of a better term, between yeah, wanting to move fast and to like to minimize the risk you're talking about while also minimizing the risk of potential downside of AI being incorporated?
00:17:10.890 --> 00:17:27.210
Yeah, I I feel like AI is challenging us to redefine like how we think about security and governance more generally, because um organizations that are super conservative in their in their posture towards that, they're just they're not invading.
00:17:27.210 --> 00:17:37.210
Like their teams are frustrated because they're filling out 14-page questionnaires about like the security of a tool that they want to just like pick up and tinker with.
00:17:37.210 --> 00:17:47.049
And um, so IT and security are becoming bottlenecks and um barriers to innovation rather than leading the organization through this change.
00:17:47.049 --> 00:17:52.809
Um, what we see happen more often than not is like shadow IT popping up in response to that.
00:17:52.809 --> 00:18:00.170
Like if it's too hard to get approval for a new tool, people are just gonna sign up for a license with their personal email address and use it on the side.
00:18:00.170 --> 00:18:09.450
And that to me, that's the worst possible outcome because um you're not mitigating risk, you're just removing your visibility to it.
00:18:09.450 --> 00:18:12.889
And um that's not the solution.
00:18:13.210 --> 00:18:20.250
It's so funny because I was thinking I I know some companies that like they restrict access to uh like LLMs in general, right?
00:18:20.250 --> 00:18:23.450
Uh through IP stopping, but here's what they don't stop, right?
00:18:23.450 --> 00:18:26.649
100% our personal devices.
00:18:26.649 --> 00:18:29.690
Yeah, and it's it's funny you bring up the the shadow ITs.
00:18:29.690 --> 00:18:43.690
Like this reminds me, the reason I was sort of laughing is because this reminds me exactly the same thing I dealt with when I was doing, I was asked to take on a uh an initiative to try to reduce uh what I would call is like not reduce, it's sort of a similar thing, right?
00:18:43.690 --> 00:18:51.369
The the initial was like go reduce cost on support as revenue was increasing, and typically it'd been in parallel, right?
00:18:51.369 --> 00:18:53.769
The way I framed it was twofold, right?
00:18:53.769 --> 00:19:04.649
One, I want to bend the cost curve for sure, but that doesn't mean we are going to cut costs, we're just gonna change the rate of growth, which is a different thing, and we're gonna capture more of it.
00:19:04.649 --> 00:19:23.450
But in the process of doing that, working with the support teams and the leaders, like it turned out because there was no centralized sort of knowledge base of ways to provide answers quickly to customers when they came up, is they all had these, they had a document on their desktop or a spreadsheet, or they had their own little database.
00:19:23.450 --> 00:19:29.690
So they already had the solution that we were proposing, even though there was resistance at the top, yeah, that we should do something centralized.
00:19:29.690 --> 00:19:36.089
It feels like this is why I'm like, it feels this is why I struggle with like, is it really different or not?
00:19:36.089 --> 00:19:47.129
I yeah, to me, because I've gone through something like this personally, where I'm like, yeah, like I think the way you address it is not to say it's gonna replace your job, it could, right?
00:19:47.129 --> 00:19:50.329
We do need to adopt adapt to it and adopt it.
00:19:50.329 --> 00:19:57.289
But the way that I in that case, right, it was like this is gonna free you up to work on the more interesting, challenging issues, right?
00:19:57.289 --> 00:19:58.410
If that's what you care about.
00:19:58.410 --> 00:20:06.329
Now, if all you want to do is come in and check a box and you know, you know, we solve some things, then it's probably not the right thing for you.
00:20:06.329 --> 00:20:09.609
This is probably the same case for AI tools.
00:20:09.609 --> 00:20:12.649
So I mean, are you seeing stuff like that too?
00:20:12.649 --> 00:20:19.450
Like where that messaging is kind of not going through the way they expect it to as leaders are communicating.
00:20:19.930 --> 00:20:23.450
Yeah, yeah, no, I I see exactly what you're what you're seeing there.
00:20:23.450 --> 00:20:38.410
And um, the example of like reps with their own spreadsheets and Google Docs and like things saved to their desktop or saved to their like notes on their phone or whatever the case, like um I mean everyone has to find a way to get their job done.
00:20:38.410 --> 00:20:50.410
And yeah, if you make it too difficult to go through the the the main channels, people are gonna find side doors to accomplish what they need to accomplish and keep things moving in their job.
00:20:50.490 --> 00:21:02.170
So I mean that thing about this stuff is being hidden from like if you don't open up that, then this stuff is gonna happen, it's gonna be hidden from you is a really big thing I don't think a lot of teams think about.
00:21:02.730 --> 00:21:03.049
Yeah.
00:21:03.049 --> 00:21:10.490
Yeah, I think that like the best hope is to make like the the secure way of doing things the easy way of doing things.
00:21:10.490 --> 00:21:31.690
Like give your team the tools that they need to do their jobs, give those tools access to the data um needed to provide context to to power some of those LLM conversations and like pro make that the default versus like forcing people to find these side doors to to get what they get done what they need to do.
00:21:33.049 --> 00:22:08.809
So so I guess this is to some degree this implies like they're using frameworks for governance and whether it's data, technology, things like that that are rooted in uh do we call pre-AIs or a term for like pre-AI world um models that like seem like they don't work as well today because there's I mean, the one thing it feels like there's a proliferation of this, like these small one-off tools with like purpose built with AI underneath the hood that solve specific problems that people can get used very quickly.
00:22:08.809 --> 00:22:12.089
Um, so is that I mean, do you think that's part of the problem?
00:22:12.089 --> 00:22:19.609
Is that the that mindset of like how you m you manage this from a governance standpoint and the technology and risk management has got to change?
00:22:20.329 --> 00:22:29.129
Yeah, I I do think the pendulum is swinging the like the But a different way than it has for the last decade on the talk of point solutions.
00:22:29.129 --> 00:22:41.450
Because for the last several years, we've been talking about okay, how do we eliminate point solutions, bring everything onto a central platform or system of record, and um like eliminate tools that do the same thing?
00:22:41.450 --> 00:22:48.250
But with the way that AI has moved, to your point, like there's a lot of individual things that do a task super well.
00:22:48.250 --> 00:23:03.369
And so instead of like, okay, it has to be all on a single platform, it's like how do we connect sort of all this infrastructure so that we're not copying data, we're not moving things around in an insecure way, um, but we're picking like the right tool for the task.
00:23:03.369 --> 00:23:12.569
And also recognizing that with how quickly things are moving, the right tool for the job in six months from now may be totally different than what's the best like now.
00:23:12.569 --> 00:23:17.529
And how do we make it like how do we prevent like being handcuffed to vendors?
00:23:17.529 --> 00:23:20.970
How do we make it easier for our teams to experiment and move quickly?
00:23:20.970 --> 00:23:24.329
Um, so that the narrative on that has totally shifted.
00:23:25.210 --> 00:23:25.849
Yeah.
00:23:25.849 --> 00:23:40.329
So for the audience that we have as ops people, like, how should they be thinking about managing this speed and versus agility versus control and risk management, those trade-offs between those things?
00:23:40.329 --> 00:23:42.329
How should they be thinking about doing that?
00:23:43.129 --> 00:23:50.649
Yeah, I um I mean, I feel really strongly that speed and agility are like that's the top priority right now.
00:23:50.649 --> 00:23:52.009
Um I'm with you.
00:23:52.009 --> 00:23:59.849
And too many organizations operate in like a slow, very waterfall thinking decision by committee type of approach.
00:23:59.849 --> 00:24:07.609
Um, and it's just it's resulting in organizational gridlock where nothing gets approved and um folks have to work around the system.
00:24:07.609 --> 00:24:18.569
If it's like a six-week approval process to run an experiment that takes like all of six minutes, you're you're on like the wrong side of the the risk to the risk to speed trade-off.
00:24:19.129 --> 00:24:22.170
So no, I don't think that's just a a a theoretical thing.
00:24:22.170 --> 00:24:24.889
It feels like that's been a real experience that you've you see.
00:24:25.289 --> 00:24:30.329
Oh yes, we we we see that we see that a lot, unfortunately, in our in our customer orgs.
00:24:30.329 --> 00:24:31.769
Um that's crazy.
00:24:31.769 --> 00:24:41.049
And something that our our team talks about internally that um I I love as a concept is like asking yourself what is the blast radius of a mistake.
00:24:41.049 --> 00:24:43.930
Like if it's if it's oh that's an interesting way of thinking about it.
00:24:43.930 --> 00:24:49.450
If the impact can be continued, cleaned up, triaged quickly, it's probably not that big of a deal.
00:24:49.450 --> 00:24:56.410
And you you don't need to have sort of a really robusture around like mistakes that will cost you very little.
00:24:56.410 --> 00:25:00.809
Like building an email in production versus a sandbox is an example of that.
00:25:00.809 --> 00:25:06.970
Like the maximum damage of of doing it wrong in prod is usually not that high.
00:25:06.970 --> 00:25:19.289
Um, but on the other hand, if if a oops moment is like, okay, you're losing data, you're getting fines, you're the next big headline, like obviously you need to have your talks narrow on a on a from a security perspective.
00:25:19.289 --> 00:25:25.690
So, like, what are another an alternate term for that is like Amazon uses the term like one-way door versus two-way doors.
00:25:26.170 --> 00:25:26.410
Yeah.
00:25:26.410 --> 00:25:31.289
Like I think you're the one who told me about this now that I think about it, because I've been that's been in my head a lot lately.
00:25:31.690 --> 00:25:32.329
I love it.
00:25:32.329 --> 00:25:38.009
It's like if you could just walk back through, like, oh, I don't want to do that, like don't overthink it.
00:25:38.009 --> 00:25:42.889
But if it's like, all right, this is a permanent decision that's gonna have lasting consequences, right?
00:25:42.889 --> 00:25:45.450
You obviously need to think about those a little bit differently.
00:25:45.769 --> 00:25:58.329
Well, I love the I love the addition to that model or framework, the the blast radius one, because I think they're kind of addressing two different sort of components of a decision-making process.
00:25:58.329 --> 00:26:04.970
When one is if I make this decision beforeward, can I undo it and go back to where I was or go to another door?
00:26:04.970 --> 00:26:15.210
The other one is if I go through that door, am I exposing our uh me or position to a risk that is actually one that is, even if I could go undo it, right?
00:26:15.210 --> 00:26:16.889
Once it's out there, it's out there.
00:26:17.049 --> 00:26:17.690
Yeah.
00:26:17.690 --> 00:26:20.009
Yeah, a hundred percent.
00:26:20.410 --> 00:26:21.210
Uh okay.
00:26:21.210 --> 00:26:22.889
Yeah, I like that combination.
00:26:22.889 --> 00:26:25.049
That's a that's a really interesting way to do it.
00:26:25.049 --> 00:26:27.930
There's probably a good someone's gonna come up with a good acronym for that.
00:26:27.930 --> 00:26:32.009
Um, kind of uh some sort of good matrix will come out of that, you know.
00:26:32.009 --> 00:26:32.809
Yeah.
00:26:32.809 --> 00:26:36.329
Um yeah, so that's interesting.
00:26:36.329 --> 00:26:43.690
The um like I still struggle with like I don't think a lot of companies or organizations think about trade-offs like that, right?
00:26:43.690 --> 00:26:49.609
They think they they don't think of them as a kind of a spectrum, or in this case, right, two dimensions at least.
00:26:49.609 --> 00:27:00.250
So um how are you how are you trying to get your clients, or it sounds like you do a lot of this internally too, applying what you do with your clients.
00:27:00.250 --> 00:27:04.089
Like, how are you doing that with clients, like getting to start to think a bit differently?
00:27:04.089 --> 00:27:12.089
Um, because I did like the one thing, like I said, I'm struggling with like is this a fundamentally different thing than past big changes?
00:27:12.089 --> 00:27:19.369
Like the one thing that does feel like it is is different is that the amount of new stuff you're pointing about, like six months from now, things are different.
00:27:19.369 --> 00:27:20.889
Like, I've been here struggling.
00:27:20.889 --> 00:27:27.289
I invested in building out a lot of deep knowledge about me and how I work with ChatGPT, it just happened to be the one I used.
00:27:27.289 --> 00:27:29.450
I keep hearing all these great things about Claude.
00:27:29.450 --> 00:27:32.089
I'm like, should I jump and move and like all that work?
00:27:32.089 --> 00:27:36.089
And I'm like, I I can't decide, but then like they're gonna leapfrog each other.
00:27:36.089 --> 00:27:41.690
So, like, is that I mean, are you like is that really the fundamental change that's happening now?
00:27:41.690 --> 00:27:48.329
Is it's just happen changing so quickly, and it's hard to like go like, how am I locking myself into something I can't undo?
00:27:48.970 --> 00:28:11.690
Yeah, yeah, I think the the concept of like shelf life is rapidly evolving because I mean we we used to talk all the time doing CRUB and marketing automation implementations, like you want to build for the long term, like you want to pick a platform that you're gonna be on for the next three to five years and not make architectural decisions that you're gonna have to undo.
00:28:11.690 --> 00:28:14.490
Um, I just don't think that's the the reality anymore.
00:28:14.490 --> 00:28:27.450
I mean, like the the big LLMs like Chat GPT, Claude, Gemini, um, it's been fascinating to watch how like one pulls ahead, one drops behind, one pulls ahead, one drops behind.
00:28:27.450 --> 00:28:38.410
And where my head is out, that is like I don't want to pick a winner or like make a long-term commitment to any platform until I absolutely have to.
00:28:38.410 --> 00:28:52.970
So like shorter term contracts, allowing our team to choose the tools that they want to use to for different types of tasks, um, and testing the models against each other, um feel feels like where we need to be right now.
00:28:52.970 --> 00:29:03.289
Um I've gotten to a place where most sort of like routine queries, if I if I have the time, I I'll put it in the Gemini Claude Chat GBT.
00:29:03.289 --> 00:29:08.649
And it's fascinating to see where some miss and where some don't.
00:29:08.649 --> 00:29:17.049
And um it's just a great sanity check that like we're not really ready to hand over the full keys to the yeah, the AI overlords just yet.
00:29:17.049 --> 00:29:24.410
Like they're human in the loop definitely needs to still be in the picture um because no no model is getting it right 100% of the time.
00:29:24.730 --> 00:29:25.690
Yeah, it's interesting.
00:29:25.690 --> 00:29:28.089
I've I've only done what you described a few times.
00:29:28.089 --> 00:29:29.609
I should probably get back to trying that.
00:29:29.609 --> 00:29:35.129
But one thing I did do recently was we're was collaborating on a project with someone, and I was using ChatGPT.
00:29:35.129 --> 00:29:40.809
That person was using Claude and uh generated something out of Claude.
00:29:40.809 --> 00:29:43.690
I used Chat GPT to evaluate it, tweaked it.
00:29:43.690 --> 00:29:48.649
Like I think the combination ended up being something better than either one of them would have done by itself, too.
00:29:48.889 --> 00:29:49.529
Yeah.
00:29:49.529 --> 00:29:52.569
Yeah, I think there's a there's a lot of value in that.
00:29:52.569 --> 00:29:59.210
Like multiple team members using different models and using it to come up with the best possible output.
00:29:59.690 --> 00:30:22.809
Which like now gets back right back to this like, how do you like this is why that idea of giving people a little more freedom to pick and choose what they're comfortable with, as long as the I I think I think this is where this is where the problem of these people are thinking that middle managers are gonna be less and less relevant, right?
00:30:22.809 --> 00:30:35.609
I think this is where they can play a role is like evaluating the the output, you know, and guiding and and helping people, independent whether or not they know how the LLMs all work, which they probably should know, at least to some degree.
00:30:35.609 --> 00:30:42.809
But like that to me is one of these fallacies is out there, is like, oh, this work is gonna come out and it's gonna be better, more efficient, etc.
00:30:42.809 --> 00:30:44.170
But is it really?
00:30:45.210 --> 00:30:47.769
I mean, yeah, who's the judge of that?
00:30:48.089 --> 00:30:48.730
Right.
00:30:48.730 --> 00:31:09.529
I mean, what somebody I heard talking the other day, and I think he kind of alluded to something like this in earlier in our conversation today, is that if you are a low, you know, a bottom 50% kind of person of whatever it is you're producing stuff that you can now use LOM, like you're gonna your work is gonna get elevated.
00:31:09.529 --> 00:31:17.849
But if you're in the top 50, it's actually gonna bring it down because the the models essentially are geared towards getting to the middle, right?
00:31:17.849 --> 00:31:19.210
What's average or typical?
00:31:19.210 --> 00:31:26.250
And and like I I think that's when I started hearing that, I was like, oh, that fundamentally describes my experience, right?
00:31:26.250 --> 00:31:31.450
With this with when I'm dealing with something that I know really well, I don't find the output is good.
00:31:33.049 --> 00:31:33.369
Yeah.
00:31:33.609 --> 00:31:34.250
Are you seeing that?
00:31:34.250 --> 00:31:37.129
Like, does that ring with true with you too?
00:31:37.849 --> 00:31:39.289
That does ring true with me.
00:31:39.289 --> 00:31:46.329
Um, and that I do think that the the power of human discretion paired with these models is really important.
00:31:46.329 --> 00:31:58.009
Um and on that like bottom 50%, like we I see I see some folks that maybe weren't the highest volume producers in in the past era.
00:31:58.009 --> 00:32:08.329
Like now you can get Claude to spit out a like 50 page, 50 slide deck for you that like looks really smart, sounds really smart, it's really beautiful.
00:32:08.329 --> 00:32:12.009
And but when you start digging into it, it's like it it just looks smart.
00:32:12.009 --> 00:32:14.410
Like it isn't it isn't actually smart.
00:32:14.410 --> 00:32:36.089
And like that discretion of like, okay, the LLM is gonna take us through the first 70, 80 or whatever, but that human is refining the last 20 to make it really commercially viable and impactful and um just sense checking it, like that that 20% we're not we're not gonna lose the the value of.
00:32:36.649 --> 00:32:43.930
Yeah, at least I I always uh I want to tap like it doesn't feel like that's anywhere in the near term, right?
00:32:43.930 --> 00:32:44.809
Maybe one year.
00:32:44.809 --> 00:32:45.289
Yeah.
00:32:45.529 --> 00:33:03.690
And honestly, I wonder if this problem actually gets harder over time, because to your point on like the draw towards the average, like if I'm writing an email in Chat GPT and you're writing an email in Chat GPT, and we're just sending them back and forth to each other, and the models are getting trained on like more AI output.
00:33:03.690 --> 00:33:09.609
At one point, at what point does to mean start losing variability and like everything just sort of becomes like mid?
00:33:09.609 --> 00:33:14.569
Um it'll be it'll be interesting to see how we navigate that.
00:33:14.889 --> 00:33:24.170
Yeah, I mean that's kind of um I a term I I like to use is like this there's a fallacy of best practices, which feels like it's a similar thing, right?
00:33:24.170 --> 00:33:28.569
Where people are like, oh, this is the way everybody does it, so we should do it that way.
00:33:28.569 --> 00:33:31.609
Yes, but yeah.
00:33:31.930 --> 00:33:33.930
I'm with you 100% on that.
00:33:34.089 --> 00:33:34.329
Yeah.
00:33:34.490 --> 00:33:42.649
I love that term, the fallacy of best practices, because they're they're best practices for a typical use case or an average scenario.
00:33:42.649 --> 00:33:51.930
It doesn't, it doesn't mean that it's the silver bullet that's gonna work for you in your exact situation that um is probably different than the person sitting next to you.
00:33:52.250 --> 00:33:52.569
Yeah.
00:33:52.569 --> 00:34:04.250
I mean, the way I like to think about them is what are the principles that were in that quote best practice, end quote, or the thing that was successful that can be replicated and applied creatively in different ways.
00:34:04.809 --> 00:34:05.849
I I love that.
00:34:05.849 --> 00:34:09.690
Like the the principles, like why did that make that successful?
00:34:09.690 --> 00:34:13.530
And are those those same variables true in this era?
00:34:13.530 --> 00:34:16.809
If they're not, that best practice is dead.
00:34:16.809 --> 00:34:30.809
And yeah, um, I think we're we're on the front lines of a lot of that happening in marketing right now, where like the content marketing, the ABM playbook, like things that have just been like, oh, this is this is just how we do it, they're becoming less effective.
00:34:30.809 --> 00:34:33.370
And so we're gonna need to rewrite the rules.
00:34:33.370 --> 00:34:38.490
And um yeah, it's it's a brave new world for for marketing.
00:34:38.810 --> 00:34:39.450
Yeah.
00:34:39.450 --> 00:34:43.290
So maybe not the best book to refer to at this point.
00:34:43.450 --> 00:34:44.810
Um yeah.
00:34:46.170 --> 00:34:52.730
Um, yeah, it's a little bit too too close to the through to the reality, maybe.
00:34:52.730 --> 00:34:54.330
I don't know, maybe not that close.
00:34:54.330 --> 00:35:02.490
Um, no, but uh so when we talked before, you talked about the difference between I think the term governance and activation, it feels like maybe that's what you're talking about, right?
00:35:02.490 --> 00:35:06.250
Is there's like a new way of thinking about how to manage this change.
00:35:06.250 --> 00:35:07.290
Like, what does that mean to you?
00:35:07.290 --> 00:35:10.090
And like how would it apply for the people who are listening?
00:35:10.410 --> 00:35:20.330
Yeah, I have to give credit to my friend Corey, Corey Snow for this paradigm shift, but like reframing like governance as a function to the act the activation team.
00:35:20.330 --> 00:35:25.850
So instead of seeing governance as like the department of no, it's like the department of how to.
00:35:26.490 --> 00:35:26.810
Yeah.
00:35:26.970 --> 00:35:33.610
Um, governance like implies that you need to be governed, you're creating rules, you're like blocking things from happening.
00:35:33.610 --> 00:35:37.130
Um, but many of our users are pretty ungovernable.
00:35:37.130 --> 00:35:50.970
So instead of trying to make them not do the things that they want to do and doing governance to them, like how do we give them the training, the tools, the knowledge, um, the data to do what they want to do in a like safe and compliant way?
00:35:50.970 --> 00:36:08.650
Um so getting the governance team or the activation team or how whatever you however you want to frame it to ask like what are the business outcomes that you're looking to achieve and what is the safest, most compliant way to get you there, versus like, oh no, that doesn't pass our 14-page questionnaire.
00:36:08.650 --> 00:36:09.610
So we're not right.
00:36:09.690 --> 00:36:10.730
We're not right, right.
00:36:10.730 --> 00:36:27.050
Like it needs and it needs to be like I think there's a lot of governance approaches, ones, even ones that I've built before, that I have, I think, a relatively mild assessment of the costs and benefits or risks and things like that, but they still take time.
00:36:27.050 --> 00:36:38.730
And it feels like that needs to be even lighter weight than what I've thought about to be like what is enough information that can be defended to say yes, we can move forward with this, right?
00:36:38.730 --> 00:36:46.410
That feels like it needs to be a pretty quick conversation as opposed to a lengthy, like you said, 14-page form to fill out or whatever.
00:36:46.810 --> 00:36:51.930
And maybe that initial like blast radius assessment dictates like what path it goes down.
00:36:51.930 --> 00:36:59.930
Like if it's something that's not a big deal, like like put the 300 question security security questionnaire down.
00:36:59.930 --> 00:37:08.810
If it's something that is like significant and mission critical, absolutely do the due diligence, but like not over processing for process sake.
00:37:09.130 --> 00:37:09.610
Yeah.
00:37:09.610 --> 00:37:17.050
Well, and and so now I'm gonna I'm kind of laughing at myself here a little bit because like now I think the matrix that I said could be a part of this, right?
00:37:17.050 --> 00:37:23.690
The the can I go through the door and come back, or do I is it a permanent door that I'm going through or and the blast radius, that combination.
00:37:23.690 --> 00:37:30.650
Like if it's one that I can't come back from and the the blast radius could be big, like that recruit that should get more scrutiny.
00:37:30.810 --> 00:37:31.530
Yeah, right.
00:37:31.530 --> 00:37:36.810
Yeah, it should be hard to request that data, request that platform or whatever.
00:37:36.810 --> 00:37:43.050
Um, like users should have to do work to to walk through that door if if that's the real implication of it.
00:37:43.290 --> 00:37:58.490
Well, and I love anything that's gonna get people to start thinking about things in terms of the risks and trade-offs, because I think there's very few things in in life in general that are clearly, you know, right and wrong, yes and no, right?
00:37:58.490 --> 00:38:00.090
There's a lot of gray area in life.
00:38:00.090 --> 00:38:06.490
And I think that's the like this this feels like the exact scenario where that plays a role, right?
00:38:06.490 --> 00:38:11.370
Where you've like you need to have this mindset to evaluate things honestly, right?
00:38:11.370 --> 00:38:19.130
Just because you think it's a great idea doesn't necessarily mean it's not a high risk one, or you know, even if it's an easy thing to do, right?
00:38:19.130 --> 00:38:43.930
And I think that's that's a that's a a framework and a model that I would totally support if I was in a leadership role, like because if you're teaching people to think like that, and maybe that's really like what needs to happen is these organizations, we need to be thinking about how do we get our teams to be thinking about like the criteria like that in a general sense so that that when they're thinking about things, it's not like yes or no, I should do that.
00:38:43.930 --> 00:38:56.010
It's like is the if we take the step forward of doing this thing, you know, are we are we do we feel like the upside is gonna be better than the down potential downside?
00:38:56.010 --> 00:38:58.010
And can we come back from it?
00:38:58.410 --> 00:38:59.050
Yeah.
00:38:59.050 --> 00:39:00.010
Yeah.
00:39:00.010 --> 00:39:04.330
I I I feel like there's a there's a slide in the making with like a visual for that.
00:39:04.570 --> 00:39:06.570
I've got like in my head, I've got the same thing.
00:39:06.570 --> 00:39:06.970
Yeah.
00:39:07.130 --> 00:39:07.769
Yeah.
00:39:09.290 --> 00:39:27.130
Um, but I do like I think that you you're right, that like that idea of activating and becoming more of a yes with with guardrails and sort of clear expectations about um how things are gonna operate and yeah, when you're gonna make decisions.
00:39:27.130 --> 00:39:40.170
I mean, this is gets back to a little bit of like one thing a um lesson I learned from a decision-making class I took years ago at a company that did provide a lot of training was this like the sunk cost fallacy, right?
00:39:40.170 --> 00:39:48.330
I see that so often where these are like, oh, we spent a million dollars or whatever it was, right, on this thing, like and it's not quite there.
00:39:48.330 --> 00:39:50.330
We should like keep spending more money.
00:39:50.330 --> 00:39:59.850
It's pretty with the people who are looking at it like usually know like it's not gonna work, like, and so spending more is not really the right thing to do, even though you've already spent a million dollars.
00:40:00.170 --> 00:40:20.810
Yeah, yeah, I think the sunk cost fallacy is like where it's like front front and center, because that a lot of ops teams have infrastructure that they've spent years building out and refining and perfecting, and um, now there's a whole new set of market entrants that have different capabilities and different implementation timelines.
00:40:20.810 --> 00:40:36.810
And like if it's gonna take you 200 grand to optimize your existing platform to do the thing versus buying something off the shelf that um there, there's yeah, it's you you can't you can't be anchored to the past and thinking about like what are your decisions for the future.
00:40:37.130 --> 00:40:50.890
Yeah, I've always been a fan of trying to keep things as simple as possible because like you, like I am a big fan, I'm a big believer that speed and agility can trump, you know, even a poor strategy, right?
00:40:50.890 --> 00:40:57.769
Because if you have enough strategy and you can move quickly and you can adapt adapt or cut it if you like, and then move to the next thing.
00:40:57.769 --> 00:41:04.890
I think you've got an advantage over people who spend a lot of time on a great strategy that ultimately may work, but if it's too slow, they may miss the opportunity.
00:41:05.210 --> 00:41:05.850
Yeah.
00:41:05.850 --> 00:41:06.490
Yeah.
00:41:06.490 --> 00:41:10.170
I speed and agility win every day in my book.
00:41:10.170 --> 00:41:18.490
Um because and I I feel like we we see this trade-off in like on a lot of marketing teams and even in a lot of sales organizations.
00:41:18.490 --> 00:41:58.170
Like you start the year with your strategy deck of like, okay, this is these are the activations we're gonna do, these are the strategic initiatives, and then the rubber meets meets the road of just life, and like your customers are doing new things, your competitors are doing new things, and like the faster you can adapt systems, get the date, get data in the right hands of of your internal teams, get campaigns out the door, like the organization that moves the quickest is has major advantages over one that has a perfect strategy, but is slower to react to to what's coming down the the pipeline from the like the broader universe that we all all exist in, kind of thing.
00:41:58.490 --> 00:41:58.650
Right.
00:41:58.650 --> 00:41:59.370
Yeah, yeah.
00:41:59.370 --> 00:42:24.090
So if if for the people who are listening or are just like if they're an organization or trying to think, okay, how maybe they're actually working on a strategy for how do I incorporate AI into our go-to-market operations and our workflows, what like what are the things you're seeing that the best organizations are doing that seem to be, I'll use that same like have principles that can be adopted by other places, maybe not best practices.
00:42:24.490 --> 00:42:25.130
Yeah.
00:42:25.130 --> 00:42:40.810
Yeah, I I mean the the organizations that that seem to really be hitting their stride with this have like top-down support and alignment for the need to innovate and the the potential of AI in the organization.
00:42:40.810 --> 00:42:49.290
So I do think there's an aspect of like you gotta work for a team that sort of gets the vision and and sees the light kind of kind of thing.
00:42:49.290 --> 00:42:59.769
Um I'm also seeing like high-performing teams, like they they're providing their team with tools, but they're also teaching them to fish with those tools.
00:42:59.769 --> 00:43:10.410
So teaching them the principles, not just like making rules, but explaining like, okay, this is what we're trying to do, this is how we think about these things, this is how we make decisions in this era.
00:43:10.410 --> 00:43:22.810
Um, so that like as process changes are being pushed out top down, bottom up, the team is innovating and being agile and embracing change from their seat.
00:43:22.810 --> 00:43:35.210
Um, so it's it's kind of a uh so if I could boil it down to like the three things, it's like leadership alignment, um like top-down process evaluation and change.
00:43:35.210 --> 00:43:41.130
So like these are tools and processes that we can standardize across across our whole team.
00:43:41.130 --> 00:43:57.050
And then that teaching people the fish bottom-up innovation, where you're getting people the tools to be their own change agents and improving their own workflows versus waiting to be told, like, okay, this is the process that's changing because of AI.
00:43:57.450 --> 00:43:58.170
Yeah, okay.
00:43:58.170 --> 00:43:59.769
I mean, I love that anyway.
00:43:59.769 --> 00:44:02.650
Uh, can I somewhat clarify on the leadership alignment piece?
00:44:02.650 --> 00:44:16.090
Is that that the leadership agrees that there's an AI benefit and that they agree on what they expect that benefit to be, and then that's communicated, or is it more of a mandate kind of thing?
00:44:16.090 --> 00:44:18.650
Like how did like what do you what do you see?
00:44:18.650 --> 00:44:22.410
Because I I could see lots of different interpretations of what that means.
00:44:22.890 --> 00:44:23.130
Yeah.
00:44:23.690 --> 00:44:27.530
And I um I I feel like it has a lot of legs.
00:44:27.530 --> 00:44:35.610
Cause I mean, like I feel like every CEO on the planet right now is getting the message from their board, like, we gotta do more AI, we need to do more with AI.
00:44:35.610 --> 00:44:44.490
And then many of them are just turning around and repeating that message to their organization without a lot of clarity on like what does it mean to do AI and like how are you measuring that?
00:44:44.490 --> 00:44:46.810
And like what does good even look like?
00:44:46.810 --> 00:44:57.610
Um, I think also we're seeing a lot of CEOs beating the like we gotta do AI drum, and then like security and infosec don't get the memo.
00:44:57.610 --> 00:45:06.090
And so it's like we gotta do AI, and the teams are trying to embrace innovation, and then they're getting like handcuffed um to these giant security bustard airs again.
00:45:06.090 --> 00:45:06.490
Yeah.
00:45:06.490 --> 00:45:19.290
Um and if if it's like ops versus security, and ops is responsible for convincing security that they need to loosen their posture, that's usually not a winnable battle.
00:45:19.290 --> 00:45:22.890
Like you you really do need the C-suite to be aligned on.
00:45:22.890 --> 00:45:28.010
Like we get we have to be agile, we have to lean into innovation, we have to change.
00:45:28.010 --> 00:45:31.610
Um or it just it just doesn't end up happening.
00:45:31.610 --> 00:45:33.130
Um yeah.
00:45:33.130 --> 00:45:36.650
So it's it's a it's a it's a tough one.
00:45:36.650 --> 00:45:50.730
Like I I feel like uh the the C-suite in general is getting a lot of pressure around AI, and postures need to evolve for for their teams to to really successfully partner with them on that change.
00:45:51.130 --> 00:45:53.370
Yeah, it's just like something that's popped in my head.
00:45:53.370 --> 00:46:10.970
I mean, it feels like this um the challenges with evaluating the ROI or the sort of the financial benefit of adopting AI do seem like really difficult, kind of similar to what marketing has faced for years, right?
00:46:10.970 --> 00:46:14.970
I was marketing measures self as a from a financial standpoint.
00:46:14.970 --> 00:46:25.130
Um and it it almost feels like there needs to be I don't want to say no accountability to the financial performance, because that's gotta be realistic.
00:46:25.130 --> 00:46:36.730
And like one of the things like I have avoided using a lot of automated AI stuff is because I've when I tried it, like unexpectedly high volumes of transactions that each have a cost to them, right?
00:46:36.730 --> 00:46:50.650
Um and but I do think there's almost needs to be a what are the again, I'm gonna use the word principles, but what are the principles we believe we're uh in terms of like the major things we're gonna get as a benefit?
00:46:50.650 --> 00:46:59.130
And do we feel like we're making progress towards that in kind of a larger point of view?
00:47:00.250 --> 00:47:11.610
I think the analogy to to marketing's measurement problems actually is super relevant because um I mean we've been chasing like perfect marketing attribution for eternity.
00:47:11.610 --> 00:47:20.810
And yeah, some things you can very clearly point to, like, okay, this is the pipe generated or the um the pipe that was matured by this marketing initiative.
00:47:20.810 --> 00:47:24.170
And like we have better metrics than we ever had before.
00:47:24.170 --> 00:47:29.850
But there's still this whole category of stuff that's like you either see the value or you don't.
00:47:29.850 --> 00:47:40.810
And like you're not gonna get a a dashboard element that tells you, like, yep, you sent a client newsletter, and like you either believe in a client newsletter or you don't.
00:47:40.810 --> 00:47:46.250
Like it you either anecdotally see the value and you like see the metrics that it's performing.
00:47:46.250 --> 00:47:52.730
Um, but like there's no amount of data that can convince you that like talking to your customers is a good thing.
00:47:52.730 --> 00:48:03.530
And similarly, like with AI, I mean, there's some things we can measure, but there's there are also things like, all right, a search that used to take me 15 minutes now takes two minutes.
00:48:03.530 --> 00:48:07.370
Do we believe that faster search is valuable?
00:48:07.370 --> 00:48:09.290
Like, I would say yes.
00:48:09.290 --> 00:48:12.970
And yeah, but like are we gonna see that drop to the bottom line?
00:48:12.970 --> 00:48:14.170
Probably not.
00:48:14.330 --> 00:48:17.210
Like at least not in the in the near term, right?
00:48:17.210 --> 00:48:19.610
And I think that's the that's the challenge.
00:48:19.610 --> 00:48:22.170
Yeah, it's been uh this has been a fun conversation.
00:48:22.170 --> 00:48:24.330
I guess maybe I'll leave it like wrap up here.
00:48:24.330 --> 00:48:32.330
Like, is there anything we haven't covered yet that you think that um you know our audience, primarily operators, would benefit from hearing that we hadn't covered?
00:48:33.210 --> 00:48:35.130
Um, I think we covered a lot of ground.
00:48:35.130 --> 00:48:40.970
Um I mean, I I think the the bottom line is like the space is evolving really quickly.
00:48:40.970 --> 00:48:47.290
The human aspect of change is equally important, if not more important than the actual tech itself.
00:48:47.290 --> 00:48:58.890
Um and ops teams who are agile and um deliver like focus on time to value of the things that they're that they're shipping, I think those are those are the teams that are gonna pull ahead and win.
00:48:59.210 --> 00:48:59.769
I love it.
00:48:59.769 --> 00:49:01.370
This has been so much fun, Andrea.
00:49:01.370 --> 00:49:02.170
Thank you so much.
00:49:02.170 --> 00:49:09.610
If folks want to continue with you or learn more about what you're doing or what's happening at your organizations, what's the best way for them to do that?
00:49:09.850 --> 00:49:12.410
Um, I'm on LinkedIn, love to chat on LinkedIn.
00:49:12.410 --> 00:49:23.210
Um, and then we uh Trilliad.com, um, we have a great blog and um full webinar light up and um a lot of great content uh to share with listeners.
00:49:23.530 --> 00:49:24.090
Terrific.
00:49:24.090 --> 00:49:24.810
That's been great.
00:49:24.810 --> 00:49:25.930
Well, again, thank you, Andrea.
00:49:25.930 --> 00:49:26.730
It's been a lot of fun.
00:49:26.730 --> 00:49:28.090
I've really enjoyed the conversation.
00:49:28.090 --> 00:49:30.410
It's actually kind of crystallized some thoughts in my head.
00:49:30.410 --> 00:49:32.410
So awesome, awesome stuff.
00:49:32.410 --> 00:49:33.130
I appreciate it.
00:49:33.130 --> 00:49:35.690
Thanks to our audience out there for continuing to support us.
00:49:35.690 --> 00:49:40.970
If you have ideas for topics or guests or want to be a guest, you can reach out to Nanny, Mike, or me.
00:49:40.970 --> 00:49:43.130
We'd be happy to get the ball rolling until next time.
00:49:43.130 --> 00:49:44.330
Bye, everybody.