When Fewer Tools Mean Higher Efficiency—A Minimalist Workflow Experiment

At the end of 2025, I ran an experiment: for an entire month, I cut my work tools in half.

Not because of some philosophical pursuit of productivity minimalism. I was just tired of too many tools. Every morning I'd open my computer and spend half an hour context-switching between Notion, Linear, Slack, Figma, Gmail, Calendar, 1Password, Arc, and three different AI tools—just the mental overhead of "loading each tool's context" was eating 30 minutes of cognitive energy before I'd done anything.

Each tool is good on its own. Stacked together, they're an endurance race through noise.

The Rules

I set three rules for myself:

One tool per function. Overlapping tools get cut.
New tools get 30 days. If they don't earn their place, they go.
"Just in case" is not a valid reason to keep anything.

What I cut: Notion (replaced with plain text + Git), Linear (replaced with a kanban board inside Obsidian), Figma (design goes to a dedicated design agent; I no longer touch it directly), two AI tools (consolidated to Claude).

What I kept: Obsidian, Arc, Claude, calendar, email. Five tools.

Week One: Pain

The first seven days were uncomfortable. Not because functions were missing—because I kept reaching for tools that weren't there anymore.

That reflex itself was revealing. How much of what you think you need is real functional need, and how much is just the weight of habit?

By day three, I found myself using Claude conversations as a "temporary Notion"—pasting thoughts into a chat and asking Claude to organize them. It was faster than manually formatting blocks in Notion. Not the design, just reality.

Weeks Two and Three: Adaptation

Some problems I expected never materialized.

I thought losing Figma would make design communication painful. What actually happened: the design agent's deliverables came with enough context that plain text comments were sufficient. The reason Figma occupied my workflow was that I'd been doing something I shouldn't have been doing in the first place.

Other problems were real. Plain text + Git isn't as smooth as Notion for cross-device sync. Switching to iCloud Drive with Obsidian helped considerably, but I still hit occasional sync conflicts that needed manual resolution.

After One Month: The Data

I tracked "tool switches per day" (rough count):

Before: approximately 65 switches per day
After: approximately 28 switches per day

Subjectively: deeper work blocks. Before, finishing one task often meant "touring" the tools before starting the next. Now the momentum is more direct—the next task is just sitting in Obsidian waiting.

An unexpected finding: email processing got faster. Reason: I stopped using the "inboxes" of various tools—Slack notifications, Notion reminders, Linear issues—and consolidating everything into one place made prioritization clearer.

What Didn't Change

Number of tools doesn't equal quality of attention. Cutting tools just removed one layer of noise. The underlying attention fragmentation problem is still there.

I still pick up my phone when I should be focused. I still open the laptop "one more time" when I should be resting. Tool management and attention management are different things. This experiment only solved the first one.

Conclusion

Minimal workflow isn't a silver bullet, but it does force an honest question: of the tools on your list, how many do you actually use, and how many are cognitive rent you're paying for "might need someday"?

If you feel like you keep adding tools but efficiency keeps plateauing, try forcing a cut. The worst outcome is discovering which tools actually were essential—at least then you've confirmed their value.

The biggest lesson from a month of this wasn't that any specific tool was redundant. It was realizing I'd consistently overestimated how much tools contribute to efficiency, and underestimated the cost of switching between them.

SFD Editor's Note

SFD Lab runs 14 AI agents in parallel, each with hard tool boundaries—content agents don't touch code tools, research agents don't touch operations tools. That's team-level "minimal workflow" by design: each tool appears only where it's strongest. The benefit is low coordination friction; the cost is that you have to think through the boundaries upfront, or gaps appear at the edges.