Behind the Build: The Years of Practice Now Living Inside Our Agents
Tracy Thayne6/11/2026·7 min read
When people talk about how a software company gets built, they usually talk about the code. The architecture, the model choices, the infrastructure. I want to talk about something that came long before any of that, because it is the part that actually makes Expona work: the two decades of practice that the code is encoding.
Expona's agents are good at marketing not because of a clever prompt. They are good because they reason through methodologies that took me years to learn, refine, and trust. Those methodologies were not invented on a whiteboard during a planning sprint. They were earned the slow way, in real campaigns, with real budgets, under real pressure, working alongside people who knew far more than I did. This is the story of where that intelligence came from, and of one person in particular who shaped how I think about all of it.
Methodology Is Not Theory You Read, It Is Practice You Survive
Early in my career I had the same misunderstanding a lot of marketers have. I thought a methodology was a framework you adopted, a diagram you printed out and pinned to the wall. Do the steps, get the result.
What I learned over many years is that a real methodology is something closer to a reflex. You do not understand a buyer journey because you can name its stages. You understand it because you have watched a deal stall at the exact stage you neglected, and you have felt what it costs. You do not understand brand voice because you have a style guide. You understand it because you have shipped a campaign that technically followed every rule and still sounded like no one in particular, and you have had to figure out why.
That kind of knowledge does not come from a book. It comes from repetition, from being wrong in public, and from having someone willing to put you in the situations where you could test what you thought you knew. Which brings me to Marc.
Working With Marc Sherman
I worked with Marc Sherman at Schneider Electric, where he ran a North American marketing team. People who worked with Marc called him Sherm, and anyone who knows him is probably smiling reading that.
What I loved about working with Marc, and what I still love about him, was his limitless enthusiasm and positivity. Marketing at that scale is not a straight line. You hit constraints, shifting priorities, and odds that sometimes look genuinely insurmountable. Marc had a way of keeping the whole team motivated and excited about the work even when the path forward was anything but obvious. That was not naivety. It was leadership. He understood that a team that believes in the work will outlast a team that is merely managing it.
Subscribe
Get notified by email when we publish a new post. No spam, unsubscribe anytime.
What that gave me, practically, was room. Because Marc kept the energy and the belief steady, I could stop worrying about the noise and focus on the two things that actually mattered: the process and the purpose. I could think clearly about why we were doing the work and how to do it well, instead of burning my attention on whether it would survive the next obstacle.
Marc did something else that mattered even more in the long run. He was genuinely supportive of putting me into situations where I could test the methodologies we were developing. He gave me the opportunity to take an idea that was still half-formed and try it against a real problem, with real stakes. That is rare. A lot of managers protect the safe path. Marc understood that you cannot validate a methodology in the abstract. You have to run it, watch it break, and rebuild it from what you learn. The frameworks I trust today are trustworthy precisely because I got to pressure-test them in those years, and I got to do that because someone believed it was worth the risk.
What Those Years Actually Taught Me
The methodologies at the foundation of Expona are not new inventions. They are the distillation of what held up across all of that practice.
The first is the six-stage buyer journey. Buyers do not move from unaware to sold in one leap. They pass through distinct stages, and at each one a different person with a different motivation needs a different thing from you. I learned to map that journey not as an academic exercise but because I watched what happened when we skipped a stage. The work taught me to respect the sequence.
The second is campaign-to-journey alignment, which sounds obvious and is almost never done well. It is the discipline of making sure that what you publish actually corresponds to what a specific buyer needs at a specific stage, rather than producing content because the calendar says it is Tuesday. Most marketing organizations have a content engine and a buyer understanding, and the two barely speak to each other. Closing that gap was the work I kept coming back to, engagement after engagement.
The third is the one I have written about most, because it sits underneath the other two: context is the whole game. A framework only produces good output when it is fed real, current, structured knowledge about your buyers, your brand, and your market. Speed without context produces noise, not signal. That lesson is the through-line of everything I learned in the field, and it is the reason Expona exists.
Turning Lived Experience Into Intelligence the Agents Use
Here is the part that still feels a little strange to say out loud. The methodologies I spent years learning are no longer just in my head. They are encoded into the way Expona's agents reason.
When an Expona agent works on a campaign, it does not start from a blank prompt and improvise. It reasons through the buyer journey. It checks what stage a given buyer is in and what that buyer actually needs there. It draws on the structured context about your company instead of generic training data. In other words, it does the things I learned to do over twenty years, except it does them consistently, at scale, and without forgetting the framework the moment the work gets busy.
That is the real transformation behind the build. We did not teach the agents to sound like marketers. We taught them to think like a marketer who has been through the methodology enough times to trust it. The lived experience, the hard-won reflexes, the lessons that used to live only in the heads of seasoned practitioners, are now an operational layer that anyone using Expona can draw on. You can read more about how that works in our platform capabilities and in the way we think about agentic workflows.
This is also why I am skeptical of AI marketing that treats a powerful model as the whole answer. The model is the engine. The methodology, the buyer understanding, the years of knowing what actually moves a deal, that is the intelligence that decides where the engine goes. Without it, you just generate faster. I wrote about that distinction in AI agents are already here, and it is the bet the whole platform is built on.
Gratitude, and the Through-Line
If you trace Expona back far enough, past the architecture and the models, you find a set of methodologies. Trace the methodologies back far enough and you find the people who gave me room to learn them. Marc Sherman is high on that list. The enthusiasm he brought, the belief that kept a team moving against the odds, and the trust to let me test ideas in the real world, all of that is part of what Expona is made of, even though none of it shows up in the code.
I think that is true of most things worth building. The technology is visible. The years of practice and the people behind it are not, but they are doing most of the work. I am grateful for both, and I am glad I finally got to say so. Thanks, Sherm.
If you want to see what those methodologies look like running inside a platform, you can explore Expona or set up a workspace and try it yourself. And if these founder reflections are useful to you, subscribe to the blog (below) for the next one.
Tracy Thayne* is the founder of Expona and a B2B marketing strategist focused on operational intelligence. Subscribe to the blog (below) for weekly insights on marketing, AI, and the founder journey.*
