Two tools. Two failure modes. One person who tried both.
Our interviewee runs operations at a digital agency in Switzerland. Having spent years working across development projects, he works heavily with AI — particularly Claude. He's not someone who avoids complexity for its own sake. He can handle technical tools.
But when it came to building a daily workspace, two of the most popular tools in the space didn't fit. Notion felt like too much overhead. Obsidian felt like it spiraled out of control. The failure modes were different — almost opposite — but the result was the same: neither became a tool he actually used day to day.
This is the pattern worth paying attention to. Most people who switch to xTiles don't switch because their old tool was bad. They switch because it was wrong for how they actually work. This case illustrates both failure modes clearly.
Notion: when power becomes overhead
Notion's reputation is well-earned. It's one of the most flexible workspace tools available — databases, views, linked properties, formulas, templates. For teams building internal wikis or structured documentation, it can be excellent.
But that flexibility has a cost. Before you can use Notion effectively, you have to design it. Every database needs properties. Every view needs configuration. Every workspace needs a structure you've thought through before you start. If you open Notion wanting to capture something quickly, you often spend more time setting up the container than putting the thing inside it.
"Notion is just too complicated for me." Head of Operations, digital agency
For an operations professional who needs to context-switch quickly — between client projects, internal processes, and personal goals — this setup cost is the problem. Notion rewards patience and architectural thinking. It penalizes urgency.
You open Notion to capture something fast.
There's no obvious place for it, so you create a new database.
The database needs properties, a view, maybe a template.
Twenty minutes later, the original thought is lost or oversimplified.
People don't leave Notion because it's bad. They leave because the setup tax is higher than the daily return.
Obsidian: when flexibility becomes chaos
Obsidian takes the opposite approach. No databases, no properties — just Markdown files and links. For writers and researchers who think in connections, it's a powerful system. The verdict from our ops lead: it's fine. But then: it just gets too messy.
"I've tried Markdown — which is fine with Obsidian — but it just gets too messy for me." Head of Operations, digital agency
The messiness Obsidian users describe tends to compound over time. You start with clean folders and disciplined naming. Then projects multiply. Orphaned notes appear. The graph view becomes a tangle. Finding anything requires either perfect recall or a search that returns too many results.
Obsidian is built for people who find structure motivating — who enjoy maintaining a personal knowledge graph and thinking in linked notes. If you mainly want to capture, act on, and move past information — as most operations work requires — the overhead of maintaining that graph quickly outpaces its value.
- No visual layout — everything is a flat file; spatial organization is hard to express
- File management overhead — folders, tags, and naming conventions need active maintenance
- Plugin dependency — useful features like calendar or task views require third-party plugins that can break
- Scale problem — Obsidian is beautiful with 50 notes; it becomes a maze with 500
Why xTiles worked where the others didn't
When this ops lead came across xTiles, the initial reaction was that it felt like "a nice fit" as a project tool. Not a revelation — more like the absence of friction. No need to design a structure before starting. Open a project and immediately begin.
The visual tile layout is part of this. In xTiles, content has spatial weight — you can place things next to each other, resize tiles, embed links, images, or videos into the same view as your tasks. Notes sit alongside tasks; references sit alongside plans. This is closer to how project thinking actually works.
Work projects and side projects at the agency
Embedding Jira tickets directly into project views
Personal tracking — fitness goals and longer-term personal projects
Daily planning with the xTiles Planner
AI-assisted briefings and task management via Claude MCP integration
The xTiles template gallery also resonated. Rather than building a structure from scratch, you start from a working layout and adjust it. And when nothing in the gallery fits, the built-in AI can generate a starting point from a plain-text description.
The Claude MCP connection: xTiles as an AI context source
The feature that stood out most as a next step was the Claude MCP integration. Already working with Claude extensively, the question wasn't whether to use AI — it was how to connect the tools used for planning and tracking with the AI used for execution.
xTiles MCP makes that connection bidirectional. Claude can read from and write to your xTiles workspace — pulling task context, updating project notes, generating daily briefing pages from calendar events and inbox summaries. The workspace becomes a context source, not just a note-taking destination.
Connect xTiles to Claude via MCP
In your xTiles profile under Integrations, create an API token. Add it to Claude's developer tools configuration — the setup takes a few minutes and requires no code. Step-by-step guides are available at xtiles.app/en/imagine.
Generate your daily planning page automatically
Once connected, Claude can pull your xTiles Planner tasks, calendar events, and other sources to build a daily briefing page — and save it directly into your workspace without any manual input.
Use xTiles as context for Claude
With MCP active, your project notes and tasks in xTiles become available to Claude as a knowledge source. Ask it to summarize a project, draft a client update from your notes, or surface tasks that haven't moved in a week.
For someone running operations across multiple projects simultaneously, this changes the maintenance equation. The time cost of keeping a workspace current drops significantly when an AI agent can update it, summarize it, and surface what needs attention.