When Good Tools Go Bad
You've probably cursed OKRs and Agile at some point. They feel like bureaucratic hoops designed to make your life miserable. I get it. But here's the thing: these tools aren't the problem. They're just tools. The real issue is how they're wielded by people who don't understand them.
Last semester, I took a Strategy & Management course. We dissected these frameworks as rational solutions to organizational problems. They made sense in theory. But in practice, they often become cudgels. The disconnect between classroom theory and office reality is stark.
The 70% Trap: OKRs as Performance Metrics
OKRs are designed to be uncomfortable. The whole point is to set ambitious goals that you can only achieve 70% of. That's by design. It pushes you to stretch. But some managers, bless their hearts, use that built-in discomfort as a performance review tool. Suddenly, everyone's running on fumes.
This is a textbook misapplication. OKRs are not a performance evaluation framework. They're a goal-setting framework. The Objective is the direction. The Key Results are milestones to measure progress. The system's purpose is to align the entire company toward a shared vision, preventing tunnel vision. It's about driving change.
KPI, on the other hand, monitors the health of existing operations. You can have all green KPIs, meaning the business is humming along. But that doesn't mean the company has a sense of direction. That's where OKRs come in. They answer, "Where do we go next?"
Management textbooks will tell you: keep OKR self-assessments separate from KPI-based reviews. Give employees a clear signal that OKRs don't affect their bonuses. Only then will they feel safe to challenge themselves.
But when OKR completion rates are tied to compensation, the safety net vanishes. People start protecting themselves. They set goals they know they'll hit 100%. The tool meant to encourage exploration and bold bets becomes a stage for performing loyalty. Ambitious folks set moonshot goals, hit 70%, and get labeled failures. The calculating ones set easy targets, hit 100%, and get praised as stars. Congratulations, you've made OKRs even less effective than KPIs.
Don't Get It Backwards Either
The reverse is also true. You can't apply the "70% is fine" mentality to KPIs. Nobody accepts a 70% uptime for servers. (Well, maybe GitHub, but that's a joke.) Baselines are baselines. That's what KPIs are for.
Mixing the two creates a credibility crisis. No one knows the actual standard. Everyone's just guessing what the boss wants.
Fake Agile: The Worst of Both Worlds
Now, Agile. The core principles are short iterations, tight collaboration, and embracing change. "Embracing change" often gets used as an excuse for chaotic product decisions. That's a recipe for burnout. The result? We get the worst parts of both Agile and Waterfall.
Waterfall is top-down. It assumes you can fully understand requirements before writing any code. You write a perfect spec, and everyone executes it faithfully. The flow is one-way: research, requirements, architecture, development. No surprises.
Agile, on the other hand, admits you can't know everything upfront. Requirements emerge as you go. Each sprint produces a small result, which you show to users. Their feedback shapes the next sprint. Sometimes they overturn your assumptions entirely. That's the real "embracing change."
But in toxic environments, managers slice a giant Waterfall plan into two-week chunks. They call them sprints. At each checkpoint, they report progress against the original plan. This fake Agile only addresses the efficiency of execution, not the uncertainty that real Agile tackles.
Real Agile is about dealing with the unknown, the dynamic demands of the market. The whole workflow is organized around that reality. Fake Agile is just a Gantt chart with a new name.
When "Change" Becomes a Weapon
"Embracing change" was meant to accommodate feedback from users, not the whims of a product manager who can't make up their mind. That's not market feedback. That's just chaos.
Even in the orthodox Scrum framework, change has a specific mechanism. Sprints last two to four weeks. During that window, the sprint goal and scope are locked. No one can barge in and change tasks mid-sprint. This protects the developers' expectations. Product managers can update the backlog anytime, but new requests only get picked up after the current sprint ends. They don't crash into ongoing work.
Refactoring Is Not Optional
Agile is built on gradual evolution. It doesn't assume you'll design a perfect architecture on day one. (You can't.) Instead, it expects you to maintain the architecture through continuous refactoring. Refactoring is as natural as breathing. When you add a new feature and the current structure can't handle it, you refactor first. You adjust the code so it can support the new capability.
Every sprint should include refactoring. In Agile, a feature is only "done" if it passes tests, has been refactored, and meets code quality standards. If you skip refactoring to cram in more features, you rack up technical debt. That debt makes the code rigid. The next change becomes exponentially more expensive. Keeping an eye on technical excellence is a core Agile principle. Refactoring is how you preserve it.
It's Not the Tools, It's the Authoritarian Culture
Why do so many people in Chinese tech circles hate OKRs and Agile? I think the hatred isn't aimed at the concepts themselves. It's aimed at authoritarian structures. OKRs and Agile were meant to bring a human touch to development. But in a system that's already authoritarian, these humanistic intentions get twisted into tools for relentless exploitation.
Before developers are "human resources," they're human beings. But toxic workplaces treat people like AI, then expect AI to behave like a human. It's a lose-lose.
So, what can you do? First, recognize the difference between good use and abuse. If your OKRs are tied to bonuses, that's a red flag. If your "sprints" are just deadlines for a Waterfall plan, that's another. If "embracing change" means your PM changes their mind daily, run.
You might not be able to change the whole system, but you can start a conversation. Share this article. Educate your manager. Or, if that fails, update your resume. Life's too short to be a cog in a broken machine.
Comments (0)
Please sign in to post a comment.
Don't have an account? Create one
No comments yet. Be the first to comment!