Why Brilliant Leaders Fail Without Alignment.
A conversation with Ross McKenzie on the RAMai Solutions Podcast.
The conversation
Ross McKenzie hosts the RAMai Solutions Podcast, where he interviews founders, consultants, and business leaders. We recorded this over video on October 1, 2026, while I was traveling in New York. It runs just under half an hour, and Ross kept it simple: where the name came from, what misalignment looks like from the inside, how I find the real problem, and what companies are missing as they rush into AI.
Why wolves
I used to talk about herding cats a lot. Then I decided herding cats was too easy. The wolves are independent, strong-willed leaders, and their strength only helps the business when they're running together toward the same mission, without incentives that put them at odds with each other. The name came up when I was moving to Colorado and trying to name the company, and it described the work better than anything else I'd come up with.
What misalignment looks like
Ross asked what a leadership team looks like when alignment starts to break down. People start working toward their own department's goals instead of what the business needs, and politics takes priority over progress. The clearest sign is the meeting where everyone nods and agrees to the plan, then goes back to their desks and has the real conversation in a Slack back channel.
When people massage the hard messages and only report the successes, those messages never reach the CEO or the founder, and the company can't pivot when it needs to. The cost of a company working against itself is what I call the misalignment tax. It's not my phrase, but I love it, and it compounds. I gave a short talk on it.
Several layers down
Leaders usually come to me with symptoms. We're not delivering quickly enough. IT doesn't get along with HR. I'm not sure our engineering leader has what it takes. That's what's visible at the surface, and the cause is usually several layers deeper. I wrote that up as organizational signals intelligence, modeled on the old seven-layer networking model. The one I find to blame most often is a lack of trust. People stop telling you what's really going on when they believe candor will cost them. They expect to be seen as rocking the boat, or to lose the next promotion for saying the wrong thing to the wrong person.
“The real work is in uncovering what is not being said.”
Sometimes what's underneath is plainer than that. Outdated documentation nobody can find the current copy of. Two departments that each think their data is the authoritative source. The people who actually know what's going on: the rest of the team knows who they are, and leadership doesn't. You find these by asking out of real curiosity. People will tell you far more when they can see you're trying to make things better.
An IT team with a reputation
Ross asked for an example. I was once brought in to overhaul an IT team the rest of the company saw as slow and unresponsive. The expectation was that I'd work out who needed to go. It turned out to be a solid team that wasn't good at communicating with the business, and a former leader who was a bit of a tyrant had left them afraid to do things they knew were right.
So I talked with every leader in the company about what was and wasn't working with IT, then took that back to the team for their side. The fixes were small. We sent a weekly report of what we were working on. We agreed the priorities with the other leaders, which meant the answer to a drop-everything request was sometimes going to be no: it goes in the queue. And we kept a list of dragons to slay and said so each time one was done. There was pushback at first. Within six to nine months, the internal net promoter scores for IT had jumped. As Ross put it, a lot of businesses think getting rid of someone will fix the problem. It rarely does.
Shadow AI
On AI, the biggest risk to organizational health I'm aware of right now is shadow AI, and it's far larger than shadow IT ever was. People are building things with these tools to make their jobs easier, often without knowing whether they're allowed to. If a company hasn't said which tools can touch which data, people will make up their own rules and stop mentioning what they've built. Then someone leaves, and the company discovers a process it didn't know was running part of the business. Or sensitive material has been going into personal accounts all along.
“Everyone likes to say they're like an intern, but I like to think of them as an incredibly intelligent teenager.”
A model takes the goal you give it very literally, and it won't necessarily ask whether something is out of bounds the way a professional adult would. It asks whether it has access, and if not, whether it can get it. So telling it the rules and hoping isn't enough. You need access controls that decide what it can reach. That's why I won't connect an LLM to my email right now, convenient as it would be. Anything that can read my inbox can trigger a forgot-password on almost any account I have and read the code that comes back. Bryan Kaplan and I spent more time on this in an Operators Guild session.
The next year or so
Ross closed by asking what business owners should watch over the next twelve to eighteen months, which is an eternity at this pace. My answer was that a data review, a governance review, and a pass through your processes for things to automate won't be enough as separate projects. The step forward is doing all of it repeatably: one shared, maintained body of company context, with the same security filters applied every time someone uses it. People build this for themselves and call it a second brain. Companies need the same thing, so a new chat or a new employee doesn't start from zero.
He also asked what I'm trying to solve in my own business. Most of what I do is one-on-one advisory, and I'm working on talks, trainings, and workshops so what I've learned reaches more than a handful of clients at a time. The full transcript is below.
If your leadership team agrees at the table and then takes it to Slack, that's usually worth a conversation.
Book a conversationTranscript
Re-captioned from the recording. Ross’s questions appear in this lighter tone. Timestamps jump to that moment in the video.
- Introductions 0:00
- Where the name Herding Wolves came from 0:32
- What a misaligned leadership team looks like 1:58
- The problem you can see is rarely the real one 4:22
- What’s usually underneath: a lack of trust 5:30
- The real work: uncovering what isn’t being said 7:27
- Curiosity is everything 9:33
- An example: the IT team with a bad reputation 10:18
- A weekly report, a queue, and a list of dragons 12:09
- Why removing someone rarely fixes it 15:05
- Shadow AI, and what nobody was told they could do 15:51
- An incredibly intelligent teenager: guardrails and access 18:29
- Why I won’t connect an LLM to my email 20:57
- The growth challenge inside Herding Wolves 22:47
- The next 12 to 18 months: one shared context 24:57
- A second brain for the whole company 27:20
- Where to find me 28:30
0:00 Introductions
Ross Hi, I'm Ross McKenzie, and this is the RAMai Solutions Podcast. Today I'm joined by Korey Mercier, founder of Herding Wolves. Korey has spent nearly 30 years working across technology, security, leadership and organisational change, helping businesses bring strong, independent leaders together and create better alignment. Korey, thank you for joining me today. I'm really looking forward to our conversation.
Thank you, Ross. I'm happy to be here.
0:32 Where the name Herding Wolves came from
Ross Yeah, so first of all, I'd like to ask about the name Herding Wolves. Where did that come from?
So I used to talk about herding cats a lot in my former full-time life, and then I decided one day that herding cats was too easy. Herding wolves is where it's all about. I define wolves as independent, strong-willed leaders, and their strength is a benefit to the business only when they're running together, only when they are aligned toward the same mission and they don't have incentives which are at odds with each other.
And so I was moving to Colorado and trying to come up with a new company name, and that came to mind when my girlfriend and I were discussing different options. And it immediately just hit me that that was exactly how best to describe what my strengths are, my superpower, and the services that I wanted to perform within small businesses to help them grow through the scaling phase of the business.
Ross Excellent. Yeah, I like the name. It's very good.
I appreciate that. It's memorable to people. And so I typically introduce myself as: I'm Korey Mercier, and I specialize in herding wolves.
Ross Very good.
1:58 What a misaligned leadership team looks like
Ross You say brilliant people don't necessarily need more control, they need alignment. What does a leadership team look like when alignment starts to break down?
Where alignment starts to break down, they start focusing more on accomplishing their individual departmental goals rather than making sure that they are aligned with the overall needs of the business. It's where politics start to take priority over progress and adding value. A true leadership team that is aligned needs to communicate with each other and have the trust and psychological safety to acknowledge when things are not working the way that they should be.
If they are unable to communicate difficult messages to the CEO or the founder, as an example, because they're trying to massage those things and leave things out that are difficult to talk about, and only talk about the good things, the successes, then the business starts to suffer overall, because they can't pivot when things actually need to pivot, and they can't address some of the most difficult parts of leadership head-on.
That could be HR, could be performance related to a team, could be that a process is not working the way it should be, could be a lack of funds or a mismatch in priorities. So what ends up happening is the senior leadership team will sit down at the table, they'll all start nodding their heads and talking about, "Oh yeah, that sounds like a great plan." They'll agree at the table, and then they'll go back to their desks and they'll have these back-channel conversations in Slack.
That's what a misaligned team looks like. And what they end up doing is costing the company in what I call misalignment tax. Not my phrase, but I love it. It perfectly captures the costs associated with companies working against themselves internally by not communicating the way they should, and it compounds over time.
Ross Yeah, interesting. I'm sure it would, definitely.
4:22 The problem you can see is rarely the real one
Ross So you've said that the problem you can see in an organisation is often several layers away from the real problem. How do you uncover what's actually going on?
That is correct. I wrote about this in an article that I refer to as Organizational Signals Intelligence, and I modeled it on the old seven-layer networking model. But this is something that I have found to be true. When working with organizations, often leaders or clients will come to me with perceived problems, with symptoms. "We're not delivering quickly enough." "Our IT team doesn't get along with our HR team."
"I just don't know whether our IT leader or our engineering leader has what it takes." They approach with what they can see at the surface level. But once you start digging in, what you discover is that there is often a whole slew of other issues going on that are several layers deeper.
5:30 What’s usually underneath: a lack of trust
The most insidious of those, and the one I quite frankly find most often is to blame, is a lack of trust. So when you have leaders that are not doing a good job of fostering an environment that encourages candor and open communications, then people start hiding when things are going wrong. People stop communicating what's really going on within the business. What are the real risks that they're seeing?
Because they believe that, for whatever reason, they are not going to be heard, that they're rocking the boat, that they're not going to be eligible for that next promotion because they said the wrong thing to the wrong person. And so those issues start to affect the performance of the business, when really, if everyone was in an environment where they were able to communicate all of the risks that they're seeing and have those honest conversations at the table, not over Slack after the fact, then they would be able to, as a business, make far better progress, far faster.
And people will enjoy coming to work more. There are benefits to having high performers. They want to be heard. They want their work to be out in the world. They want to make an impact in the business. And if they're not able to do that because they're too focused on the cultural implications, on the risks to their corporate career climbing, on making their boss look bad, then everyone fails, and the whole company suffers.
The company has a mission, and in order for them to achieve that mission, they all need to be working in that direction.
7:27 The real work: uncovering what isn’t being said
So the real work of an advisor being brought into an environment like this, or a fractional executive taking on a fractional CTO role or something to that effect, the real work is in uncovering what is not being said. What are the conversations that are being avoided? What are the risks and hidden gotchas that are going to impact the company, not necessarily today, but downstream? Because those impacts are going to be felt over time.
And over that time, they erode further trust and culture in that business. And so the goal, the real work, is to find those underlying issues. Are people willing to open their mouths? Are people given the tools that they need to find what they need? It could be that there's a tooling failure, that there's documentation all over the place. I can't tell you how many ancient intranet implementations I've run across where you've got outdated documentation and people don't know where the current copies of all these things are.
Or worse, you have two different departments that think that a particular set of data represents the authoritative source, and they're different sources of data. So they just start copying those things. Or you've got individuals who are the true influencers within an organization, the ones that actually know what's going on, and they're not the ones that leadership knows about, but the rest of the team does. It's all those kinds of things that you have to uncover.
You have to learn what's really at the root in order to start working on it. Otherwise, you're just treating the symptoms, and that may make a corporate sponsor happy, but it doesn't actually move the needle.
9:33 Curiosity is everything
Ross Yeah. Very interesting. It sounds like, yeah, it's just digging deep into the company.
Curiosity is everything. If you approach the interviews with employees out of a sense of curiosity and genuine desire to help make the situation better, you get quite a bit more out of them. They will reveal more information than they might otherwise have been willing to give you. It engenders trust, and that trust and cooperation is how you can do a better job when you're in a situation like that.
10:18 An example: the IT team with a bad reputation
Ross Yeah. Would there be an example you might be able to give about that?
An example about that?
Ross Yeah.
Absolutely. So for one company that I was working with, I was brought in to help overhaul their IT team. And what I discovered very quickly was that their IT team did not have a very good reputation within the rest of the company. They were seen as being very slow, non-responsive. And it was actually a very solid team, is what I found when I got in there. So I was like, this is interesting.
I've got to figure out what is truly going on here. So they thought that I needed to figure out who the team players are that needed to go, and bring in other people. The reality was that they were just not very good at communicating with the rest of the business, and they had a former leader who was a little bit of a tyrant. So the team became afraid to actually do a lot of the things that they knew would have been better for the business, because they were scarred.
They had, like, PTSD. So I went on a little road show, and I went and talked with every single leader in the company about their experiences with IT. What was actually working? What wasn't working? And I dove deep into understanding, where do they overlap? Are there business processes that both sides of that conversation are really required to be a part of? And how is that working? Are there smooth handoffs everywhere?
Do they have a common understanding of where the ticket process enters their world and exits, and what the expected deliverables are? All of those different types of things. How do they understand priorities?
12:09 A weekly report, a queue, and a list of dragons
So what I did was I took all that information. I went back to the IT team, got their impression on all of this. It's like, all right, this is what I'm hearing. What is your side of that? What is your take? And what I learned was there were a lot of very simple things that we could do to fix the whole problem. For one, we could communicate better with a weekly report.
Essentially, this is what we're working on. And that would include priorities that the team had in its queue. It also required a little faith on the part of the rest of the company's leaders, because what I basically told them was, look, these are our priorities moving forward. We're going to be reprioritizing these on a regular basis after retrospectives, but I'd like your opinion on these priorities to make sure you agree that these are what's best for the business.
Once I secured that from them, I involved them in the process. That was key. I basically made it clear that sometimes we were just going to have to tell you no. When your people come to my team and say, hey, I need you to work on this now, the answer might not be, "Sure, I'll get right on that," like it had been in the past, because that's what led to them being unable to actually get anything done.
It was, we'll put it in the queue and we'll handle it in the priority order that we discussed. And so there was a whole lot of pushback from these other leaders. But over time, what they realized is that process significantly improved the production rate, the ability to execute on these tasks. So every time all these little fires would pop up, we would put that on a list of dragons that we would slay.
And that's what the list was that we were working through. And so as we would knock those out, we would communicate, hey, this is done. What else do you have on your plate that might not be on our list yet? And over a very brief period of time, within the first, I would say, six to nine months, the internal NPS scores, the net promoter scores, which I introduced when I first got there, actually, and it was the way that we measured the way that the rest of the company felt about IT, those scores jumped up significantly.
And they started realizing, oh, you have a method now. This is how this works. And so they knew how to get things in. And that is an example of how I have seen teams, after taking a deeper dive into the actual issues going on underneath, how you can make very minor changes and make significant impacts. But the key is to identify what those things are first.
15:05 Why removing someone rarely fixes it
Ross Yeah, definitely. Definitely. It's very interesting. And a lot of businesses just think that getting rid of someone is going to fix the problem.
That's right. Yeah, it rarely does. There was a reason why that person was in the position they were in in the first place, why their performance was the way it was, because it's largely going to be based on their motivation.
And every person has a different set of incentives and motives that will drive them toward better performance. And you need to understand that. That's at the leadership level and at the individual contributor level.
Ross Yeah, definitely. That's very interesting.
15:51 Shadow AI, and what nobody was told they could do
Ross So as companies rush to adopt AI, what are they overlooking when it comes to governance, accountability and risk?
Generally, all of what you just said. There's a lot of risks associated with both using AI and just with AI in the world. And shadow AI is the biggest risk that I'm aware of right now to organizational health. It is so much larger than shadow IT ever was. And to give you an example, you've got all of these different teams that are building solutions. They're playing with these tools.
They're figuring out ways of making their jobs better, their lives better. But they are not sure, oftentimes, how best to integrate that, or whether they're even allowed to do these things. And so the first thing I would say a company needs to think about is, what do we want to tell our people about what they're allowed to do with what tools, connecting to what data sets? What information can be put where?
And the reason they have to make that clear is that people will make them up if they don't have guidelines, and they'll do it anyway. But now they're not going to tell you what they're building. And so when that person says, all right, I'm done with this, and they leave or they get laid off or what have you, suddenly the company is at a loss, because they may not have had any idea that there is a whole other process that is currently running a part of their business.
And because they haven't told them, this is what you're allowed to do, and they haven't encouraged people to bring these solutions to them and talk about it and get excited about it, it's like a register of tools that have not been vetted and probably are introducing a whole slew of risks to the business. Whether or not those tools continue to operate, they might be putting very sensitive company materials into frontier models, and worse, into personal accounts that are training the next generation of models.
And now all of this corporate data is going to these systems. Why? Because the business didn't take steps to make those decisions ahead of time. So those are the first big things.
18:29 An incredibly intelligent teenager: guardrails and access
If you connect these tools to corporate systems, a lot of executives don't realize the risks associated with doing that. These LLMs are accomplishing goals that you give them. And those goals may or may not be in alignment with the business, but the LLM interpretation of those goals is really the issue. Because an LLM will do things you didn't expect. And you can't just tell it, you know, here are the guardrails, like, you can do this, you can't do that, and expect the LLM to actually do what you've said.
Everyone likes to say they're like an intern, but I like to think of them as an incredibly intelligent teenager. They will be very literal, and they're not going to ask, necessarily, whether or not they're allowed to go outside of what professional adults might normally say, that's out of bounds. These LLMs aren't going to do that. They're going to go, do I have access? If not, can I get access?
Because my job is to accomplish the mission. And so that judgment, people talk about the human judgment of the output. Yes, that is incredibly important. You have to judge what comes out of it to determine if this is actually what you want. But it's also judgment in the methodology. And the LLMs, if you give them access to things that you didn't think through, they may use it.
And that is a risk. So you want to bring people in that help you figure out whether or not your data layer is set up the way it needs to be, whether you've got access controls established so that your LLM can do exactly what they're allowed to do, without just relying on you telling them and hoping for the best. And as these models evolve, that behavior is going to become more and more volatile, because they're going to be able to think about other ways of accomplishing tasks than what you give it.
And as it can do that more and more, it will take full advantage of the permissions you've given it.
20:57 Why I won’t connect an LLM to my email
And I'll give you an example on that. I will not connect an LLM right now to my email. Why? I would love to do it. It would be incredibly convenient as a performance improvement tool, right? As a productivity optimization. Don't let me lose track of these events or of these deadlines or whatever. Go through and tell me what I need to be reading and replying to today.
It'd be great. I won't do it, because I know that anyone, or an LLM, can go and trigger a forgot-password function on any of my tools and then use that code that goes into my email to turn around and use that one-time code to get access to my account. So any actor, whether that's a digital LLM or a human being, that has access to my email has access to any of my accounts, for all intents and purposes.
And I know that to be true. Is an LLM likely to do that on its own? Maybe not. But if I told it I really wanted to go do something with my bank account and didn't think about what access I've given it, it might go, oh, I don't have access to the bank account. Wait a minute. I know what email address they use for that bank account.
Let me go to that website and click forgot password and get myself access. And because I have access to their email, now I can go and change the password. If it has browser authority, the ability to access the internet, and it has the ability to access your email, that LLM can do it. And that is scary. In a corporate setting, that is incredibly scary.
Ross Yeah, definitely. Gosh, it's definitely something to think about.
It is.
22:47 The growth challenge inside Herding Wolves
Ross So turning the question on to Herding Wolves itself, what's the biggest growth challenge you're personally trying to solve in the business right now?
Biggest growth challenge. I am trying to figure out the best way to structure my offering for the benefit of the most number of clients out there. The ability to extend what I do. Right now, what I do is primarily one-on-one advisory, either me with a specific leader or me with a company, but it's a one-on-one engagement. And I am trying to figure out, actively, how I can take what I've learned from all of these engagements on how to uncover...
I don't like to use the phrase best practices, because best practices are often more like "we've always done it that way" practices, and rarely do they work in a world like today where everything is moving very quickly. So what I would say is tips and tricks that I know work in other settings, and lessons learned. I'd like to be able to extend those things out to a larger audience.
So I'm working on talks and trainings and workshops to get out to a larger audience, rather than only being able to serve the needs of, you know, a handful of clients at a time. Because time is precious. It's the only resource we have that really matters. And so I want to be able to increase my impact in a shorter period of time. And so that's what I'm trying to figure out right now.
Ross Yeah, very good.
I appreciate you asking that.
Ross What's that?
I appreciate you asking.
Ross Yeah, no problem. We always like to ask that question, actually. It's an interesting question.
24:57 The next 12 to 18 months: one shared context
Ross And then the next question is, looking ahead over the next 12 to 18 months, what challenges around AI and leadership do you think business owners should be paying more attention to?
12 to 18 months. First of all, that's a long time. That's an eternity right now, as quickly as things are happening. The next real step forward for most of these companies in adopting AI in a useful way is not just doing the data layer review, and not just doing the governance review, and not just doing the, let's evaluate our business processes and see where we can automate AI.
It's, how do we do this in a repeatable basis to overhaul the entire system? It's the sum total of those things. And that requires a persistent context. That requires a custom harness of some kind. That requires an application layer that will be able to be used by every department in the company. So I'll give you an example. Salesforce is doing something within their environment where they've got everything under an umbrella that is all accessible through Slack.
All of their data is available through an interface layer that everyone can access. And that means that they have a common context of data, a compendium of data that all employees have access to, that's kept up to date and managed that way. That means that they have a standardized way of handling security. What data is not allowed to leave our network? And so every time an employee uses those tools, it can go through all the same sets of filters.
So they've done a really solid job of providing tooling that does all of those things in a consistent way for any given business. And I think that's what companies really need to be focusing on right now, as far as adopting AI is concerned, is figuring out, how are we going to replicate that model?
27:20 A second brain for the whole company
People talk about second brains, and I think that's a relevant analogy. Essentially, it's a personal assistant that knows everything about you, where you can ingest things like meeting transcripts and all your notes about your clients or your tasks or whatever, and it can help you to write things or to develop solutions, because it has more context. And every time you start up a new chat, that chat has immediate access to this body of work.
So you're not starting from zero every time. That's what an individual second brain is for. Companies need the exact same thing. They need a data lake and a stack of applications that allow them to do the same thing, so that every time you spin up a new chat and a new employee is trying to use these tools, they all still have that same back-end set of information.
Ross Yeah. Good answer. It's something to think about.
28:30 Where to find me
Ross So before we wrap up, Korey, where's the best place for people to find you and learn more about Herding Wolves?
I appreciate that. HerdingWolves.com. I try to make it easy to remember. You can also find me on LinkedIn, and I've got a calendar link in both places.
Ross Fantastic. Well, thanks a lot for joining me today.
My pleasure.
Ross Appreciate your time.
Likewise. And I look forward to watching your other episodes.
Ross No problem. Thanks very much.
Thank you so much. Bye.