20 min read
MARCUS
Monday, August 3, 2026 – 5:00 AM – Marcus’s Home Office
August: $10.
Marcus checked two numbers every weekday morning at 5 AM, before Sarah woke up, before the boys started getting ready for school. First the pre-market stock price. $10, down from $12 at IPO close, down from $16 at opening. Then the ARR dashboard. $234 million, down from $240 million at IPO.
Both numbers moving in the same direction. Both bleeding out, week after week.
The customer losses had started the day after the Meridian announcement. Hartley Shipping, the Fortune 500 company whose CTO had warned him about feature delays, exercised their option not to renew. Two days later, a mid-sized 3PL that had been with Axiom for three years announced they were “evaluating alternatives.”
By mid-August, six more customers had given notice. The migration offers from MeridianOne were too good to ignore: free data transfer, free training, 90-day implementation guarantee. Everything Axiom had once offered to steal customers from the legacy players was now being used against them.
Marcus didn’t know which number to watch more closely: the stock price eroding his net worth, or the ARR erosion that was causing it. They fed each other in a death spiral: customers left because Axiom looked weak, and Axiom looked weak because customers were leaving.
At $10, his 1.4% stake was worth $8.3 million. Still significant money. But every week it dropped, and every week more customers announced they were “exploring options.”
He couldn’t sell. The lockup held him frozen, watching both numbers bleed away one tick at a time.
Monday, August 3, 2026 – 3:00 PM – Catherine Bell’s Office
Catherine had spent the weekend writing and rewriting the letter. Not because the words were hard. Because admitting what it meant was hard.
She’d been the best at what she did. Three companies built for exit. Three successful outcomes. A career built on one skill: taking messy organizations and making them clean enough for someone else to buy. She understood financials, governance, narrative control. She understood how to get a company from Point A to the bell.
What she didn’t understand — what she’d never needed to understand — was what happened when the ground shifted underneath you. When the playbook stopped working. When “execute the plan” stopped being the right answer because the plan itself was wrong.
She’d known since July 5th. Maybe earlier. Maybe since the day she’d issued the AI ban and watched Derek Huang’s face go flat with something she’d mistaken for compliance but now recognized as resignation. Not anger. Not rebellion. Just a young man realizing his CEO didn’t understand what she was banning.
Catherine had spent her career being the smartest person in governance rooms. The AI ban hadn’t been stupid — it had been correct for the narrow problem she was solving. Protect the IPO metrics. Minimize variance. Control the narrative. Every one of those objectives was sound. Every one of them was also irrelevant if the entire competitive landscape was shifting beneath you.
She’d been right about the tactics and wrong about the war.
Marcus knocked.
“Come in.”
He sat down across from her. She’d rehearsed this conversation too.
“I wanted you to hear it from me first.” Her voice was flat, professional — because that was who she was, and she was not going to pretend otherwise at the end. “I’m stepping down.”
Marcus stared at her. “You’re what?”
“I took this company public. That was the job Blackwood hired me for. What happens next is a turnaround, and turnarounds aren’t my expertise.” She slid a folder across the desk. “My resignation letter. Effective September 1st.”
“You’re leaving in the middle of a crisis?”
“I’m leaving because there is a crisis. The board needs different leadership now. Someone who understands technology transformation. Someone who can rebuild what needs rebuilding.” She paused. “That’s not what I do, Marcus. I build companies for exits. I don’t rebuild them after exits go wrong.”
What she didn’t say: I’m leaving because I’m the wrong person, and the longer I stay, the more damage I do. That kind of honesty was for memoirs, not resignation conversations.
Marcus wanted to be angry. He wanted to call her a coward, a mercenary, a fair-weather captain abandoning ship. But when he looked at Catherine — who had stared at the ceiling of her apartment at 3 AM the night before, running the counterfactual where she’d said “yes, let them experiment, but with guardrails” instead of “anyone caught will be terminated” — he couldn’t muster the outrage.
She’d told him. Every time Derek raised AI concerns, she’d said “is this going to lose us customers?” Every time Prakash warned about competition, she’d said “focus on execution.” She’d never claimed to understand technology. She’d only claimed to understand numbers.
And the numbers had finally caught up with both of them.
“Who’s taking over?”
“The board is bringing in Richard Townsend as interim CEO. PE background. He’s led three turnarounds. He’ll have new ideas.” She stood and offered her hand. “I wish you well, Marcus. I really do.”
He shook it. The gesture felt hollow, the end of a partnership that had never really been a partnership at all.
Catherine watched him leave. Then she sat back down and looked at the empty office she’d occupied for eighteen months. The framed photo of her daughter’s graduation. The clean desk. The whiteboard she’d never written on because the numbers lived in spreadsheets, not on walls.
She thought about Derek Huang. He’d been right. She wondered where he’d ended up.
Tuesday, August 11, 2026 – 2:00 PM – Axiom Boardroom
The August board meeting was a trial.
Victoria Hartwell led the questioning with the precision of a prosecutor. Catherine Bell was notably absent; her last day had been the previous Friday. The empty chair at the head of the table said everything about how quickly things had changed.
“Marcus, what’s our AI response?”
“We’re evaluating options. Looking at acquisitions. Talking to AI vendors. Building an internal roadmap.”
“Those are activities, not a strategy. Customers are asking concrete questions. They want to know if we can match Meridian’s capabilities.” Victoria paused. “Can we?”
Marcus hesitated. The truth was ugly, but lying to the board would only delay the inevitable.
“Not in the next twelve months. Their architecture is different. They rebuilt from scratch. We’d have to do the same thing, and we don’t have the foundation they started with.”
Richard Townsend, attending his first meeting as the incoming Chairman, leaned forward. “Let me make sure I understand the scale of what we’re facing. Meridian built their new platform with fourteen engineers. We have four hundred. They did it in fourteen months. We’ve been saying ‘after the IPO’ for over a year.” He looked around the table. “Is that an accurate summary?”
Marcus nodded. The room was silent.
“Fourteen people,” Townsend repeated. “With AI tools and clear direction. They beat four hundred people with traditional processes and delayed priorities.” He paused. “And the reports say their team included Harry Thornton, the COBOL developer everyone joked about retiring. He didn’t retire. He became their domain expert. Forty years of institutional knowledge, fed into AI tools that could build faster than any human team.”
The irony was brutal. Marcus had laughed about Meridian’s COBOL developers. Now one of them had helped destroy Axiom’s future.
“So what are we telling customers?” Victoria asked.
“That we’re committed to AI. That we have a roadmap. That we’re investing in the future.”
“Is that enough?”
“It has to be.”
But Victoria’s expression said what everyone was thinking: It’s not enough. Not anymore.
Tuesday, September 1, 2026 – 5:00 AM – Marcus’s Home Office
September: $8.
The talent exodus accelerated.
First it was a trickle: an engineering manager here, a senior architect there. People Marcus had worked with for years, people who’d believed in Axiom, walking out the door with resignation letters and farewell posts.
Then it became a flood.
One week in September, three key engineers left on the same day. All three went to Meridian.
Marcus called HR. “Can we counter-offer?”
“We’ve tried. They’re not leaving for money. They’re leaving because they don’t believe in our direction.”
Our direction. What direction? The IPO that had already happened? The AI strategy that was still being “evaluated”? The roadmap that existed only in PowerPoint slides?
Derek had been right. Prakash had been right. The best people saw the future before it arrived, and the future wasn’t Axiom.
Thursday, October 1, 2026 – 5:00 AM – Marcus’s Home Office
October: $7.
Marcus finally called Edward.
He’d been putting it off for months, dreading the conversation, dreading the admission it would require. But at 2 AM, unable to sleep, staring at the ceiling while Sarah breathed softly beside him, he picked up his phone.
Ed. I need to talk. About what you’re doing at Riverton. About how you’re approaching AI.
The reply came faster than he expected: I’ve been waiting for this call.
Saturday, October 10, 2026 – 12:00 PM – Beachside Restaurant, Fort Lauderdale
They met on a Saturday. Edward flying down from Miami, Marcus picking him up at Fort Lauderdale airport. They drove to a quiet restaurant near the beach, one they’d been to years ago, before IPOs and AI and everything that had gone wrong.
“Tell me how to catch up,” Marcus said, before the drinks even arrived. “What do I need to do?”
Edward sighed. The expression on his face was sadness, not triumph: sadness for what his friend had lost, sadness for how long it had taken him to ask.
“Marcus, I want to help. I really do. But there’s no playbook I can give you that will fix this.”
“There has to be something. You figured it out.”
“I started asking questions eighteen months ago. You’re starting now. You can’t replicate eighteen months of learning with a playbook.”
“So what do I do?”
“You start with the questions. The real questions. What’s your goal now? Not the IPO, that’s done. What are you actually trying to accomplish?”
Marcus opened his mouth to answer, then stopped. The question was simple. The answer wasn’t.
“I don’t know,” he admitted. “I don’t know anymore.”
“Good.” Edward leaned forward. “Sit with that. Really sit with it. Before you jump to solutions.”
Saturday, October 10, 2026 – 1:30 PM – Beachside Restaurant, Fort Lauderdale
“That’s not helpful. I need actions. I need a roadmap.”
“And that’s why you’re in trouble.” Edward set down his coffee. “You want answers before you’ve asked the questions. You want a roadmap before you know where you’re going.”
“So I just… wait?”
“You ask. What is your goal? What are your real constraints, not the comfortable limits, but the actual constraints? What is your current situation, not the story you’ve been telling, but the reality?”
Marcus stared at his friend. The man who’d been his roommate, his best man, his brother for twenty-five years. The man who’d seen what Marcus couldn’t see and tried to warn him.
“I don’t know how to do that.” Marcus looked at the floor. “I’ve never asked those questions. I always thought I knew the answers.”
“You didn’t know the answers. You assumed them. There’s a difference.”
“What’s the difference?”
“Knowing comes from asking. Assuming comes from not asking.” Edward paused. “You assumed your goal was the IPO. But the IPO was a milestone, not a goal. You assumed your constraints were board expectations and clean quarters. But those were choices, not constraints. You assumed your SDLC was best-in-class. But you never actually mapped it, never asked where the inefficiencies were.”
The words landed like blows. Not because they were harsh, but because they were true.
“I assumed I already had the answers,” Marcus said.
“That’s exactly the problem. And these questions only work if you ask them honestly.”
Saturday, October 10, 2026 – 2:30 PM – Parking Lot
The conversation ended awkwardly.
Walking back to his car, Marcus thought about what Edward had said. Goals, constraints, current reality. Start by asking.
It sounded simple. But when he tried to apply it, tried to ask himself what his goal was, what his real constraints were, what Axiom’s actual situation looked like, his mind went blank.
He’d spent three years protecting a trajectory. He had no idea how to build one.
Sunday, November 1, 2026 – 5:00 AM – Marcus’s Home Office
November: $6.
At $6, Marcus’s 1.4% stake was worth $5 million. Down from $13.3 million at the opening bell. Down from the $10 million at close on day one. Every month, another chunk of his future evaporating.
Five million was still money. But after taxes it would be barely $3 million. Enough to pay off the $50,000 equity loan. Maybe fund one kid’s college. But the retirement rebuild? The cushion that would let them stop worrying? Gone. Four years of sacrifice, the equity loan, the no-salary years, the raided retirement accounts, and he’d end up no better than if he’d just stayed at Soren as a senior engineer.
The desperation moves began.
Wednesday, November 4, 2026 – 11:47 PM – Marcus’s Home Office
Marcus downloaded Claude at 11:47 PM on a Wednesday.
He’d been thinking about it for weeks, ever since the October conversation with Edward. “I’m going to learn to build again,” Edward had said. “Haven’t written production code in eight years. Maybe that’s the problem.”
If Edward could do it, Marcus could do it. He’d been a decent programmer once, back in business school, back when he’d built the first Axiom prototypes himself. The skills had to still be there somewhere.
He opened the laptop. Created an account. Stared at the blank chat window.
What do you want to build?
He typed: “I want to understand how AI coding assistants work. Show me something simple.”
The response came instantly. Pages of explanation. Code examples. Suggestions for starter projects.
Marcus tried to follow along. He copied the first code block into a text editor. Python, apparently. He hadn’t written Python since a weekend course five years ago.
The code didn’t run. Error message. Something about indentation.
He asked Claude to fix it. More code. Different error.
Two hours later, Marcus had a working script that printed “Hello World” and not much else. His eyes were burning. The screen had started to blur.
Edward did this in three hours and built something useful. I’ve been at this for two hours and I can barely make text appear.
He tried again. Something more ambitious. “Help me build a simple dashboard that shows our deployment metrics.”
Claude responded with enthusiasm. Architecture suggestions. Database schemas. Frontend frameworks. The response was three screens long.
Marcus read it twice. Understood maybe a third of it. The terms that had once been familiar, APIs, endpoints, async functions, felt like a foreign language he’d once spoken and had now forgotten.
He tried to implement the first step. Got stuck on environment setup. Spent forty-five minutes trying to install something called Node.js. Finally got it working. Then the next step failed.
At 3:17 AM, Marcus closed the laptop.
The realization was brutal and complete: he couldn’t do what Edward did. Not in a weekend. Not in a month. Edward had built things for twenty years. The eight-year gap was nothing compared to Marcus’s fifteen-year gap. Edward’s rust was surface-level. Marcus’s rust went all the way through.
I can’t catch up. I can’t even understand what catching up would look like.
He poured himself a drink. Then another. Sat in the dark living room until dawn, thinking about the gap between who he’d been and who he’d become.
Thursday, November 5, 2026 – 9:00 AM – Axiom Headquarters, Conference Room A
The next morning, Marcus hired McKinley & Associates.
If he couldn’t build it himself, he’d hire people who could. McKinley was one of the Big 4’s spin-off advisory practices. Prestigious name, $800/hour rates, a team of MBAs who came highly recommended.
“We specialize in AI transformation,” the partner said during the intake call. “Fortune 500 clients. Proven methodology. We can have a team on-site next week.”
Six weeks later, they delivered their final report. The presentation was scheduled for 2 PM in Conference Room A.
Marcus arrived early. The McKinley team had set up a projector, printed bound copies of something thick enough to be a novel, arranged water bottles with geometric precision. Professional. Polished. Expensive.
“Thank you for your partnership over the past six weeks,” the lead consultant began. “We’re excited to share our findings and recommendations.”
The first slide appeared: “Axiom AI Transformation Roadmap: Strategic Assessment and Implementation Plan.”
Marcus leaned forward. This was it. The answer he’d been looking for.
“Our assessment covered twelve dimensions of AI readiness,” the consultant continued. “Data infrastructure. Talent capabilities. Governance frameworks. Change management capacity.”
Slide after slide. Matrices with color-coded quadrants. Benchmarks against “industry leaders.” Capability gap analyses.
Marcus waited for the part that would tell him what to actually do.
“Based on our findings, we recommend a three-phase approach.”
Here it comes.
“Phase One: Assessment and Planning. Months one through six. We’ll conduct deep-dive workshops across all business units, establish baseline metrics, and develop detailed implementation roadmaps.”
“Wait.” Marcus held up a hand. “You’ve been assessing for six weeks. Now you want six more months?”
“The initial engagement was high-level discovery. Phase One is detailed planning. You can’t build a house without blueprints.”
That’s what the last consultants said. Almost word for word.
“What about Phase Two?”
“Pilot Selection and Governance Design. Months seven through twelve. We’ll identify two to three pilot use cases, establish an AI Center of Excellence, and build the governance structures needed for responsible scaling.”
“So twelve months before we build anything?”
“Transformation requires proper foundation. Companies that rush this phase typically see 60-70% failure rates.”
Marcus put down the report. “And Phase Three?”
“Controlled Implementation. Months thirteen through eighteen. Phased rollout of AI capabilities, with continuous monitoring and adjustment.”
“Eighteen months total.”
“Assuming no major blockers. Some clients take twenty-four to thirty-six months for full transformation.”
Marcus looked at the bound report in front of him. Flipped to a random page. “Cross-functional stakeholder engagement is critical to ensuring buy-in across business units.”
Another page. “A robust governance structure will mitigate implementation risks while enabling measured experimentation.”
Another. “Axiom should establish an AI Center of Excellence to drive organizational alignment around transformation objectives.”
He thought about Edward. About value stream mapping. About the question Edward kept asking: What’s stopping you from shipping?
“Let me ask you something,” Marcus said. “Our deployment pipeline. Do you know how long it takes for a feature to go from code complete to production?”
The consultants exchanged glances. “That would be covered in the Phase One deep-dives.”
“You’ve been here six weeks. Interviewed our engineers. Reviewed our architecture. And you don’t know our cycle time?”
“We focused on organizational readiness, not operational metrics. Those are different workstreams.”
Different workstreams. They spent six weeks studying the company and they never looked at where work actually gets stuck.
Marcus flipped to the budget section. $2.4 million for the full engagement. Plus technology costs. Plus hiring costs. Plus the opportunity cost of eighteen months while Meridian pulled further ahead.
He closed the report.
“Thank you for your work. We’ll need some time to evaluate internally.”
The consultants nodded professionally, started packing up their projector and their bound reports and their geometric water bottles.
Marcus sat alone in the conference room after they left, staring at the hundred pages of expensive nothing.
$2.4 million. Eighteen months. And they never once asked what’s stopping us from shipping.
Thursday, November 5, 2026 – 4:00 PM – Marcus’s Office
He tried hiring a VP of AI Transformation next, someone who could report to him and drive the change he couldn’t seem to make happen. HR posted the role. The responses were dispiriting.
The candidates with impressive titles had less hands-on experience than Marcus did. They’d managed AI initiatives, sat on steering committees, written strategy documents. But when Marcus asked what they’d actually shipped, the answers got vague. “We were in the planning phase when I transitioned.” “The implementation was handed off to another team.” One candidate had “led AI transformation” at three companies in four years, which meant he’d led nothing to completion anywhere.
The candidates who’d actually built things didn’t want to join a sinking ship. One was blunt about it: “I’ve seen your stock chart. I’d be inheriting a crisis, not an opportunity.”
The last interview haunted him. A woman from a Fortune 100 company, impressive credentials, the kind of person who could actually make things happen. She’d asked sharp questions about budget, authority, reporting structure. And then she’d asked one that stopped Marcus cold:
“Who would I be reporting to? You, or the board directly?”
Marcus had said “me” without hesitation. But afterward, walking back to his office, he realized he wasn’t sure anymore. Was he hiring someone to help him? Or was the board waiting for him to identify his own replacement?
The woman had seen it too. She’d smiled politely and said she’d be in touch. She never called back.
Marcus reached out to independent consultants, fractional CTOs who’d guided other companies through AI transitions. The good ones were booked eight months out. “Call me in Q3 2027,” one said. “I might have availability then.”
By Q3 2027, there wouldn’t be an Axiom to transform.
The rumors started that week. A PE firm was circling, one of Victoria’s contacts apparently, asking questions about a take-private deal. The board would restructure, bring in new leadership, try to rebuild what Marcus had failed to protect. It wasn’t official yet. But Marcus could read the writing on the wall.
Marcus announced an “AI roadmap” to customers, a glossy presentation full of buzzwords and timeline projections. The customers nodded politely. They didn’t believe it. Marcus could see it in their eyes: Too late. Too vague. Too desperate.
He offered retention bonuses to key engineers. Most of them had already accepted offers elsewhere. The ones who stayed didn’t seem any more confident than he was.
Tuesday, November 24, 2026 – 2:00 AM – Marcus’s Home Office
Late November, Marcus sat in his home office with his laptop open.
He’d found the articles Rachel Torres had sent Edward nearly two years ago. The same case studies Edward had mentioned at the family dinner. The same approaches Marcus had dismissed as “interesting but not urgent.”
He read them now with different eyes. Not the eyes of a confident executive protecting his trajectory, but the eyes of a man who’d lost everything and was trying to figure out why.
Start with the outcome. Map the process. Identify the constraints. Fix the bottlenecks.
Process improvement, not tool adoption.
What’s stopping you from shipping?
The phrases repeated like a drumbeat. How had he missed this? How had he spent eighteen months saying “after the IPO” while the answer was right in front of him?
Because he’d never mapped his own process. Because he’d assumed it was working. Because acknowledging uncertainty had felt like weakness.
What outcome am I trying to achieve?
Marcus typed the question into a blank document. Then he stared at the cursor blinking, waiting for an answer that wouldn’t come.
Where does work actually wait in our pipeline?
Another question. Another blank line.
What constraints are real and which ones did we invent?
He knew the numbers: stock at $6, equity at $5 million on paper, lockup ending in 40 days. He knew the business metrics: customer churn accelerating, talent departing, competitors pulling ahead.
But those were symptoms, not the actual process. The process was something deeper, something he couldn’t see yet, couldn’t name.
Sarah found him at 3 AM, hunched over the laptop, typing nothing.
“Come to bed.”
“I’m trying to figure something out.”
“What?”
“What Edward showed me. Value stream mapping. I’m trying to understand where we went wrong.”
Sarah sat down beside him and looked at the screen. A blank value stream map. Boxes with no labels.
“You can’t force it.” She put her hand on his. “This isn’t a problem you solve in one night.”
“I’ve been avoiding it for two years. I don’t have more nights to waste.”
“You’re not wasting them. You’re learning to look. That’s different.”
“Is it?”
“It’s where everyone starts.” She touched his shoulder. “The Marcus I married, the one who rebuilt himself after his knee blew out, who started a company from nothing, who climbed to the top of this one, he didn’t start with answers. He started by looking honestly at what was in front of him.”
Marcus looked at her. In the glow of the laptop screen, she looked tired, worried, still there.
“I don’t remember what not knowing felt like,” he admitted. “I’ve been so confident for so long, I forgot how to actually look.”
“Then learn again. That’s what humans do.”
She stood, leaving him with the laptop and the blank screen and the process he was finally ready to examine.
Sarah stood and turned off the light. Marcus stayed at his desk, the cursor still blinking on that blank value stream map.
Where does work wait?
What’s actually stopping us from shipping?
What does our process really look like?
He sat there until morning, not writing, not solving, just sitting with questions he was finally learning to ask.
It was a start.