It’s In The Backlog

“Alright, what's in the backlog this week?” Richard said, the prompt to switch to work talk while performing the practiced tilt of his overfull glass, a lunchtime ritual he’d perfected while only in Caroline’s presence. “Still entertaining thoughts of that llama farm?”

“Oh, you know. Living the dream.” Caroline said, her fork committing small acts of violence against the remains of her lunch. “Llama farm.” Her expression drifted in and out of that thousand-yard stare peculiar to product managers of a certain age before relocking on him. “Picture this slice of hell: I’ve got an engineering leader—a mechanical engineer, by training, no less—who’s absolutely convinced that the best way to build software is with 30-page requirements docs. He wants everything defined upfront. Every. Single… All of it.” She said, with an exasperated shake of her head. “As if we're planning the D-Day invasion instead of building a web app.”

“Let me guess—he's got flowcharts about how to write flowcharts?”

“Christ, probably. Plus, he's got most of the dev team convinced this is normal. They're all playing along because it means they can hide behind documentation instead of actually shipping anything.” She released a laugh that seemed perfected in hours of therapy. “And CEO? He thinks a “functioning product team” means everyone showing up to standup in matching t-shirts.”

“I love it, the myth of perfect planning. Why iterate when you can just build it right the first time? Like a cathedral.” A sly grin curled his cheek in referencing Eric S Raymond’s low-key 1997 screed of the two conflicting approaches to building software that every dev of a certain age keeps on her desktop like a federalist with a copy of the constitution in their breast pocket.

Caroline sighed with the particular weariness of someone who'd spent too much time watching sprint points evaporate. “That's exactly his mindset. Build it like Notre-Dame. No room for learning, no space for adaptation. Just chisel the requirements in stone and call it done.”

“And I assume rational discourse is out of the question?”

“About as effectual as teaching calculus to a house plant.” She wiped carefully with a folded napkin at a pool of condensation forming around her glass of ice water.

“Then don't fight it,” Richard said, settling into the familiar rhythm of middle-management therapist. “Validate his terror. Tell him you share his fear of waste. His horror of inefficiency. Then redirect. Make it about … data. About systematically discovering and validating the space that can support change without inviting the chaos.”

“Riiight. Because he seems so open to new ideas.”

“Get him in the room. Make him part of the process. Let him feel the control while you're actually introducing additional flexibility.”

“And when he still insists on his thirty-page masterworks?”

“Then you give him exactly what he wants—”

“Oh, for Christ's sake—”

“—and you document everything. It’s data. Every hour spent, every requirement that changed, every page that became obsolete the moment it hit Confluence. Reality makes a better argument than any of us ever could.”

“So let the data make the argument for me?”

“Maybe use the data as your personal hand puppet.” He grinned.

“God.”

“Meanwhile, you build your coalition. Find the engineers who understand but are too beaten down to speak up. Turn them into allies. Make it less about you versus the mechanical overlord and more about the team discovering a better way.”

“And if by some miracle his approach actually works?"

“You mean if the requirements never change? You divine every user need before writing a single line of code?” Richard leaned forward slightly. “You know Mercedes F1 runs tens of thousands of channels of data on virtually everything that can be instrumented. The pneumatic gun for changing the wheels has eight sensors alone. They need data because they understand they can't plan for every variable—they have to discover them. Even with all their mechanical engineering might, they still need real data from real races to iterate toward winning.”

Caroline's chewing slowed to a stop.

“If your mechanical engineer wants some comfort, remind him that even Mercedes knows perfect planning is a myth.” Richard settled back.

“Okay, lovely anecdote—very persuasive. But what if his approach—?”

"Then you're fired," Richard said.

She swallowed with visible effort.

“Because you stopped paying attention.” He paused, leaning forward tracing a square on the table top with his index finger. “A waterfall approach will fail and you need to document it in real-time. You need to pay attention. This is, in a way, its own proof of concept. Treat it that way. That’s the job. A product manager who stops paying attention is as useful as a … houseplant. Worse, actually—at least the plant doesn't actively create the illusion of competence while her ship drifts toward the rocks."

“Savage.” She said, wincing. “I’ll add ‘giving it a shot’ to the backlog.”

“Remember,” Richard said, finally allowing his glass to settle, “there's always that llama farm.”

“Don't tempt me. They’re cute.”

“They spit in the face of authority. You'll survive. Even mechanical engineers can learn to iterate. It just takes time.”

“Time,” Caroline echoed flatly, mopping again at the moisture gathering at the base of her glass. “I’ll add that to the backlog too.”

The word ‘backlog’ hung there like a prayer neither of them believed in anymore, but continued to recite out of habit, a rosary for the optimistic.