Looking Upstream: A Conversation with Craig Taylor on Designing the Conditions for Success

Undercurrents surfaces insights from leaders driving capacity-building
and performance improvement beyond traditional L&D
to explore where learning happens and where it’s headed.

By Mark Britz

One of the easiest mistakes people make is believing their job is defined by its outputs.

We naturally pay attention to results because they’re easy to see. A customer renews. A learner completes a course. Revenue grows. Churn increases. By the time those numbers appear on a dashboard, however, the conversations, behaviors, and decisions that produced them have already happened. We celebrate success or scramble to explain failure after the greatest opportunity to influence either has already passed.

That thought stayed with me throughout my conversation with Craig Taylor.

Craig has spent more than a decade building customer success departments after beginning his career as a military instructor before moving through learning and development (L&D) and learning technologies. On paper, those career moves appear significant, but listening to him, I found they all shared a common thread. Every role, regardless of title, revolved around understanding how people succeed and then designing the conditions that make that success more likely.

When I asked what had remained consistent throughout his career, Craig didn’t mention learning, technology, or customer success. Instead, he paused briefly before answering, “I would say the consistency is the importance of understanding what the desired outcome is.”

As our conversation unfolded, I realized that wasn’t simply an answer to my question; it was the philosophy underneath everything else he described. Whether we were talking about military training, software adoption, AI, or organizational leadership, Craig always seemed to return to the same place. Before discussing solutions, he wanted to understand outcomes. And before talking about activity, he wanted to understand purpose. Everything else, in his mind, flowed from there.

When Training Isn’t the Finish Line

Like many people who begin their careers training adults, Craig initially focused on designing effective training. Now, however, he sees that as only one part of a much larger system. He recalled working within the nuclear industry, an environment where technical competence is non-negotiable and mistakes carry consequences few industries experience. Their training programs were rigorous, engineers and technicians left the classroom prepared to perform highly specialized work, and by almost every measure available to an L&D department the organization was succeeding. Yet something continued to bother him.

“We were training people very, very well,” he told me, “but we weren’t great at supporting them in the flow of work.”

That observation, it seems, became one of those moments that quietly reshaped his career. The issue wasn’t the quality of the instruction, it was the realization that instruction alone wasn’t enough.

Once people returned to work, they were expected to rely on knowledge and memory, even in environments where conditions changed and decisions carried enormous weight. Craig could see that capability wasn’t created simply by preparing people well. It also depended on how effectively the organization supported them after the training had ended.

Listening to Craig describe those experiences, I found myself wondering whether his move into customer success had actually begun years before he ever held the title. He had already started looking beyond learning events toward the broader system surrounding performance. The course was no longer the destination, it had simply become one point along a much larger journey.

Start with the Business Problem

Craig consistently begins with a question rather than a solution. As we talked about customer success, he described what should happen long before implementation begins. During the sales process, he believes organizations should resist the temptation to talk immediately about software features or functionality. Instead, they should ask about something fundamental: What business problem is the customer trying to solve?

The answer, he explained, is almost never, “We need new software.” It might be reducing accidents, improving compliance, increasing sales, shortening onboarding, or creating greater operational efficiency.

The software is simply a means to an end. If it can’t help the customer achieve that end, Craig argued, it probably shouldn’t be sold in the first place. If it can, then customer success inherits a very different responsibility from what many people imagine: The job isn’t to teach customers how to use a product, it’s to help them achieve the outcome they originally came to accomplish.

That may sound like a subtle distinction, but I found myself thinking about how often L&D encounters exactly the opposite. Requests arrive asking for leadership training, communication skills, or another compliance course, and it’s remarkably easy to begin designing the intervention before anyone has properly explored the underlying business problem. Craig reminded me that solutions are rarely the best place to begin. Understanding the desired outcome almost always is.

The Difference Between Measuring Success & Creating It

As we shifted deeper into customer success, I expected to hear about renewals, churn, customer health scores, and the metrics that dominate most discussions of the discipline. We did talk about those things, of course, but Craig framed them in a unique way: “They’re all lagging indicators.”

He explained that retention, expansion, and churn don’t tell customer success professionals how to influence the future. They simply confirm what has already happened. By the time a customer decides to renew (or decides to leave) the relationship has already produced its result. The numbers matter, but they’re evidence, not causes.

Then he offered a crucial definition: “A leading indicator is an input into the relationship rather than an output.”

It’s a deceptively simple statement, yet it fundamentally changes where attention is placed. Rather than celebrating renewals or worrying about churn, customer success increasingly focuses on the conversations, behaviors, and signals that make those outcomes more or less likely. Did everyone agree on the customer’s desired outcome? Is the customer making meaningful progress toward it? Are there early indicators that expectations and reality are beginning to drift apart? Those questions that the future—long before the renewal conversation ever arrives.

Everything he described was upstream of the metrics themselves. He wasn’t dismissing measurement, he was simply more interested in understanding the conditions that produced the measurements in the first place.

Perhaps that’s why his thinking resonated so strongly. For years, L&D has argued that it should be measured by business outcomes rather than training activity, yet we often find ourselves having that conversation after the intervention has already been designed. Craig wasn’t just making the same argument from another profession. He was describing what it looks like when an entire discipline organizes itself around influencing outcomes before they appear.

Activity Is Easy to Measure; Value Isn’t

Craig shared a story that illustrates the difference between measuring activity and creating value: A customer barely used the software they had purchased. Their adoption statistics would have disappointed almost any software company. They touched only a fraction of the available functionality, yet they renewed every year—because those few capabilities solved the problem they had originally set out to solve.

Then there was another customer. On paper, they looked like a customer success dream. They embraced almost every feature available, explored new capabilities as they were released, and their usage data suggested a healthy, highly engaged account. Until they left.

Craig summarized the lesson simply: “Product usage does not guarantee that the customer is achieving the outcome. … Equally, low product usage does not mean that the customer is not getting value from the product.”

Learning teams have their own version of this misunderstanding. They are asked about course completion and attendance rates because these are measurable. Yet neither tells us whether someone solved the problem that mattered in the first place.

Activity leaves behind convenient numbers, while value often leaves behind better decisions, stronger relationships, improved performance, and sometimes nothing more obvious than the absence of a problem that never occurred.

The longer I listened to Craig, the more I felt that he has spent much of his career refusing to confuse those two things. He doesn’t dismiss or discount the importance of metrics; he simply refuses to mistake them for evidence that value has been created.

Seeing the Whole System

One surprise was how little Craig spoke about process. Instead, he kept returning to information. His priority wasn’t another workflow or another methodology—it was visibility.

He described organizations where customer information was spread across CRMs, ticketing systems, product dashboards, emails, calendars, and countless other applications. Individually, every system worked well, but collectively they created fragmented understanding.

A customer success manager might know about product usage but not recent support issues. Sales might understand the customer’s original objectives but not how those priorities had evolved. Valuable information existed everywhere, yet meaningful insight remained surprisingly difficult to assemble.

Craig suggested that operating this way wasn’t impossible—but it simply wasn’t commercially viable anymore. Organizations cannot afford to spend hours piecing together fragments of information before every customer conversation, particularly when hundreds or even thousands of customer relationships are involved.

His example was quite practical.

Imagine reaching out to discuss an upcoming renewal without realizing the customer has spent the previous three months escalating unresolved issues through support. You may genuinely care about the relationship, and you may approach the conversation with empathy and good intentions. Yet without seeing the whole picture, you’re almost guaranteed to say the wrong thing.

I wondered how often L&D does something similar. Learning data is in one platform while operational performance is somewhere else. Customer data, workflow analytics, employee feedback, and business metrics all remain disconnected. We then ask why demonstrating impact feels so difficult!

Technology Creates More Time for People

Like almost every conversation today, ours eventually turned toward AI. Given Craig’s background in customer success, I expected to hear about things like automation, predictive analytics, and autonomous workflows. Those topics certainly came up, but what interested me most was where he chose to place the emphasis: Not on replacing people, but on creating more opportunities for them.

Craig acknowledged the value of AI for recognizing patterns across large customer portfolios, identifying emerging risks, recommending next steps, and, of course, automating the administrative work that consumes so much of a customer success manager’s day. None of those capabilities impressed him because they reduced headcount but because they gave people something increasingly valuable. The opportunity of AI lies in “freeing up two hours a day so that they can have better conversations with another human being.”

As Craig sees it, technology isn’t replacing judgment; it is creating more opportunities for people to exercise judgment. With software handling the repetitive work, people are able to focus on understanding customers and solving their problems.

It also explains why Craig spent so much of our conversation talking about outcomes instead of technology. For him, technology has never been the destination, it’s simply another way of removing friction between people and the results they’re trying to achieve.

Culture Before Function

Toward the end of our conversation, I asked Craig how his 16 years in the military had shaped his leadership style. I half expected him to talk about all the typical things you hear about military leadership; things like discipline, accountability, or decision-making under pressure. Instead, he talked about looking after people.

He spoke about supporting the team around you, helping people grow professionally, involving them in designing the systems they will ultimately use, and making sure they know someone has their back when things become difficult.

That simple observation said more about Craig’s leadership philosophy than any discussion about discipline ever could. Throughout our conversation, he consistently described leadership as something that enables rather than controls. Whether discussing customer success managers, customers, or colleagues, his instinct was always to ask what people needed in order to succeed.

That same philosophy appeared again when we began talking about customer success itself. Distinguishing between the customer success team and customer success, Craig pointed out that one describes a department, the other describes a culture—an ethos.

Organizations don’t become customer-centric because they hire customer success managers. They become customer-centric only when marketing, sales, product, engineering, finance, support, and customer success teams all begin making decisions through the lens of customer outcomes. Customer success becomes less about a function and more about a shared way of thinking.

Similarly, having an L&D department doesn’t make an organization a learning organization. That happens only when the focus on success extends beyond the boundaries of any single team and becomes part of how the organization operates.

The Undercurrent

In my conversation with Craig, I learned surprisingly little about customer success software—and an enormous amount about how thoughtful business leaders approach complex problems.

Craig kept returning to the same idea, regardless of whether we were talking about learning, customer adoption, AI, military leadership, or organizational culture. Before deciding what to build, what to teach, or what technology to implement, he wanted to understand the desired outcome. Everything else followed from there.

It’s a deceptively simple idea, yet one we all often overlook. It’s tempting to jump straight to solutions. We build solutions before fully understanding the performance problems and we measure results before asking whether we’re measuring the things that actually matter.

Craig’s perspective suggests a different approach: Start with the outcome, and design the conditions that make that outcome more likely.

Perhaps that’s what it really means to look upstream. Craig offered a perspective I find both simpler and more compelling. Instead of becoming better at measuring outcomes, perhaps we should become better at designing the conditions that create them.

In an increasingly complex world, where information is abundant, technology is accelerating, and change has become a constant, that may be one of the most valuable leadership capabilities any of us can develop.