# Multiplayer-by-default failure pattern

Pattern surfaced in [[sail-muddy-lessons]]: products built around "this should be multiplayer by default" keep failing for the same structural reason, regardless of execution quality.

## The graveyard

- **Tandem** — virtual office, raised $7.5M from a16z, exponential growth during COVID. Founder Rajiv Ayyangar's [retrospective](https://blog.rajivayyangar.com/what-happened-to-tandem-virtual-office/) names it directly: "most people don't feel a need to talk to their co-workers that often."
- **Multi** — acquired by OpenAI in 2024, sunsetted as a product
- **Screenhero** — acquired by Slack in 2015, eventually sunsetted
- **Google Wave** — 2009, tried to be email + IM + document editor at once. Operational transformation tech survived as Apache Wave and influenced Google Docs; the product didn't.
- **Microsoft Loop** — pitched as multiplayer collaborative workspace, became Microsoft's Notion clone
- **Safari Shared Tab Groups** — WWDC 2022 announcement, quietly deprioritized
- **Sail / Muddy** — multiplayer browser, infinite canvas with live websites, didn't reach broad launch

## Why it keeps happening

Rajiv Ayyangar's framing in the Tandem retrospective is the clearest version: "I wish we all wanted to talk with our team frequently and collaborate closely. But, most people don't... They value the autonomy and convenience of remote work, even if it can lead to disconnection and loneliness."

The thesis is built on how the founders want work to feel — collaborative, present, encountering teammates the way you encounter colleagues in Figma. The reality is that work is mostly individual, mostly async, and the rare moments where people actually want to be in the same surface in real time are well-served by Zoom, Slack huddles, or Figma's existing presence layer.

Multiplayer works as a feature inside a product whose core value is single-player (Figma's design tool, Notion's docs, Slack's messages). It does not work as the product itself.

## Related

- [[positioning-vs-vision-gap]] — the same products also struggled to reduce their grand framing to a single sentence users could repeat
- [[dogfooding-trap]] — multiplayer products often look great in dogfooding because the team is, by definition, a tight group of collaborators
- [[working-on-products-people-hate]] — engineering quality doesn't rescue a product the market has decided it doesn't need