Loom Year Three: The Async Video Verdict

Loom's thesis at peak hype was that async video would replace large categories of synchronous meeting time. The thesis was directionally right and quantitatively too aggressive. Three years on, async video has become a useful productivity tool for specific workflows rather than the dominant medium. Loom's value proposition has narrowed accordingly. The product is not in trouble; the category is no longer the structural unlock it was sold as.

We tracked Loom usage at five companies over 18 months and tested the current product against the built-in async-video features that have proliferated in the rest of the productivity stack.

Where Loom still wins

Recording quality and viewer experience. The recording experience is the cleanest in the category. Viewer-side controls — playback speed, transcript navigation, time-coded comments — remain meaningfully ahead of the built-in alternatives in Slack, Notion, and Microsoft Teams. The product is still the best at being a Loom.

The AI summarization features have improved markedly. The transcript-based summary, action-item extraction, and the new "video as searchable knowledge" feature work well enough to be useful. This is where the Atlassian acquisition has shown up most visibly — sustained investment in AI features that the previous management was visibly slower to ship.

Async video is a useful productivity tool. It is not the dominant medium. Loom's value has narrowed accordingly.

The third durable strength is the use-case-specific wins. For design feedback (record over a Figma file with timestamped comments), for customer-support recordings (record once, distribute as a help-center asset), and for prospect-tailored sales demos, Loom remains the best tool in the category. These are real workflows that produce real productivity gains.

Where the use case has narrowed

Built-in async video has proliferated and absorbed a meaningful share of what used to be Loom's universal use case. Slack now has video clips. Notion has embedded video. Microsoft Teams has built-in async video that is good enough for most internal use. For the modal "quick recording for a teammate" workflow, the built-in tools are now sufficient and the marginal value of a Loom seat has decreased.

The "video as primary async medium" thesis — replace the meeting with a video, replace the Slack message with a video — has not aged well. In our tracking, video-message frequency peaked in late 2023 and has declined modestly since. Knowledge workers prefer text for most async communication. The loud minority who prefer video has not grown.

On pricing

Loom's pricing has crept upward over the last 18 months. The Business tier is now $15 per seat per month, which is defensible for power users and hard to justify for the casual user. We have watched two of our test customers downgrade meaningfully in the last year — moving from team-wide paid plans to a smaller paid pool with the rest of the team on the free tier. This is a coherent response to the narrowed use case.

The Atlassian question

The Atlassian acquisition closed in 2024. Eighteen months in, the product has not (yet) been visibly degraded by the acquisition. Some of the integrations into Atlassian's tools (Jira, Confluence) have deepened, which is meaningful for organizations that use those tools. The roadmap velocity has been steady; AI features have shipped at a respectable cadence.

The risk for Loom remains the structural risk of any product that has been absorbed by a larger company: the slow displacement of customer-driven priorities by acquirer-driven priorities. We have not seen this yet. We expect to be watching for it through 2026.

The verdict

Loom is still a real product with real value for specific workflows. It is no longer a default tool for the modal knowledge-work team. Buyers should evaluate whether their team's actual usage justifies the per-seat cost and consider downgrading to the free tier or a smaller paid pool for users with high-value workflows. The product earns its 3.9 on the strength of the recording experience and the AI features. It does not earn the universal-default position the marketing once implied.