Observable
Observable is a web-first platform for building reactive data visualizations and collaborative notebooks, founded in 2016 by Mike Bostock (creator of D3.js) and Melody Meckfessel (former Google VP of Engineering).
Profile
Observable makes interactive, code-based notebooks for data analysis, visualization, and collaboration—think Jupyter meets a real-time whiteboard with built-in AI.
Observable is a web-first platform for building reactive data visualizations and collaborative notebooks, founded in 2016 by Mike Bostock (creator of D3.js) and Melody Meckfessel (former Google VP of Engineering). The company raised $10.5M in Series A funding in late 2020 from Sequoia Capital and Acrew Capital, followed by a $35.6M Series B in January 2022 led by Menlo Ventures. As of 2025, Observable operates with approximately 36 employees from its San Francisco headquarters at 525 Market Street.
The platform powers over one million notebooks and serves customers including Stitch Fix, The New York Times, The Washington Post, NBC News, MIT, and HuggingFace. Observable's core strength lies in its literate programming environment—combining Markdown, JavaScript, SQL, and embedded visualizations in reactive cells that re-execute automatically when dependencies change. The company expanded its product portfolio significantly in 2025, launching Observable Canvases in April (computational whiteboards for data exploration that reached general availability in August) and releasing Notebooks 2.0 and Observable Desktop (a macOS application for local-file editing) in summer.
Both products integrate transparent, verifiable AI assistance for writing and debugging code, profiling data, and creating visualizations. Observable positions itself at the developer-centric end of the data visualization market, distinct from enterprise BI platforms like Tableau, Looker, and Power BI. The company offers a freemium model with Notebook Pro at $22/month per editor and Teams plans at $90/month.
Looking ahead, Observable's 2026 roadmap centers on bringing Notebooks 2.0 and Canvases AI features to the web, consolidating desktop and browser editors into a unified experience. The strategy balances two products: Canvases (SaaS-oriented with real-time collaboration and tight database integration) and Notebooks (open-source-oriented with developer-friendly file access and Git integration).
Who buys this
- Data scientists and analysts at mid-market tech companies (e.g., Stitch Fix) exploring APIs and prototyping analyses
- Newsrooms and media organizations (NYT, Washington Post, NBC) building interactive graphics and data stories
- Enterprise data teams building custom dashboards and ad-hoc reporting on top of data warehouses
- Research and academic institutions documenting computational workflows and sharing reproducible analyses
- Developers embedding interactive visualizations into production applications or internal tools
Publicly disclosed clients
- Stitch Fix
- The New York Times
- The Washington Post
- NBC News
- MIT
- HuggingFace
Strengths and what to watch
Strengths
- Reactive notebook architecture eliminates manual re-execution and dependency management—cells automatically update when inputs change, lowering cognitive load for exploratory analysis.
- Built on D3.js and Observable Plot with first-class visualization libraries; appeals to developers and data engineers who prefer code control over drag-and-drop BI tools.
- Large ecosystem of public notebooks and community-created templates; over 1 million notebooks provide templates, tutorials, and real-world examples available for forking and customization.
Watch for
- Non-standard JavaScript syntax historically alienated Python/R-first analysts and tooling integrations; while Notebooks 2.0 shifts to vanilla JavaScript, migration complexity and potential fragmentation across old/new formats remains to be seen.
- Removed social features (user following, trending notebooks feed) and relocation of discovery tools may reduce community engagement and organic adoption—critical for a network-effect platform competing on mindshare.
- Small headcount (36 employees) scaling dual products (Canvases and Notebooks 2.0) with significant AI integration announced for 2026; execution risk on promised web-based integration and feature parity across platforms.
Recent moves
- 1y ago Observable Canvases reaches general availability with AI integration, new chart types, and version history
- 1y ago Observable Notebooks 2.0 and Observable Desktop preview released, supporting vanilla JavaScript and local file editing
- 1y ago Observable launches AI for Canvases, enabling transparent, verifiable AI-assisted data exploration and chart creation
Key Information
- Industry
- Visualization
- Founded
- 2016
- Headquarters
- San Francisco
Frequently Asked Questions
What is Observable?
Observable is a web-first platform for building interactive, code-based notebooks and visualizations. Founded in 2016 by D3.js creator Mike Bostock, it combines Markdown, JavaScript, and SQL in a reactive environment where cells automatically update when dependencies change, enabling rapid data exploration and collaboration.
What is Observable Canvases?
Observable Canvases are computational whiteboards for interactive data exploration launched in 2025. They enable spatial organization of queries and charts with real-time collaboration, AI-assisted chart creation, and version history. Canvases focus on SaaS-oriented workflows with tight database integration compared to traditional notebooks.
How much does Observable cost?
Observable offers a free tier for individual notebooks. Notebook Pro costs $22 monthly per editor for advanced features. Teams plans run $90 per month and enable collaborative projects. The freemium structure supports both individual data exploration and team-based reporting workflows.
What's the difference between Observable and Jupyter?
Observable uses a reactive architecture where cells automatically update when dependencies change, unlike Jupyter's manual re-execution model. Observable emphasizes visualization first through D3.js integration and web-based collaboration, while Jupyter targets local Python workflows. Both serve different user priorities: exploration (Observable) versus computational reproducibility (Jupyter).
Does Observable have AI features?
Yes. Observable integrated transparent, verifiable AI assistance across its products starting in 2025. AI features help write and debug code, profile data, and generate visualizations in both Canvases and Notebooks 2.0. The 2026 roadmap includes bringing AI capabilities to the web editor for unified experience.
Which companies use Observable?
Observable powers over one million notebooks and serves enterprise customers including The New York Times, Washington Post, NBC News, Stitch Fix, MIT, and HuggingFace. Media organizations use it for interactive graphics and data stories. Tech companies leverage it for rapid prototyping and collaborative data exploration workflows.
How Observable compares
Direct head-to-head against 3 competitors. Picked by 7wData.
Observable
- Positioning
- Observable makes interactive, code-based notebooks for data analysis, visualization, and collaboration—think Jupyter meets a real-time whiteboard with built-in AI.
- Customer segments
- Data scientists and analysts at mid-market tech companies (e.g., Stitch Fix) exploring APIs and prototyping analyses
- Strengths
- Reactive notebook architecture eliminates manual re-execution and dependency management—cells automatically update when inputs change, lowering cognitive load for exploratory analysis.
- Watch for
- Non-standard JavaScript syntax historically alienated Python/R-first analysts and tooling integrations; while Notebooks 2.0 shifts to vanilla JavaScript, migration complexity and potential fragmentation across old/new formats remains to be seen.
- Recent moves
- Observable Canvases reaches general availability with AI integration, new chart types, and version history
Hex
- Positioning
- Unified SQL, Python, and no-code notebook workspace that publishes interactive apps, targeting team-wide analytics rather than solo developer visualization work.
- Customer segments
- Data analysts and analytics engineers at mid-to-large tech companies. Secondary buyers are non-technical stakeholders consuming published apps.
- Strengths
- Notebook-to-publishable-app path in a single environment, with collaborative editing for both SQL and Python cells without context switching.
- Watch for
- Sharing permissions are coarse, link sharing exposes projects workspace-wide with no password protection, creating governance risk for sensitive analyses.
- Recent moves
- May 2025: raised $70M Series C led by Avra, total funding $172M, with Snowflake Ventures, a16z, and Sequoia participating.
Mode Analytics
- Positioning
- SQL-notebook-first BI platform, now folded into ThoughtSpot Analyst Studio. Targets code-first analysts, not JavaScript-based visualization builders.
- Customer segments
- Analytics engineers and growth analysts at mid-to-large enterprises. Buyer is data team lead or analytics manager inside ThoughtSpot Cloud accounts.
- Strengths
- SQL, Python, and R notebooks on a single canvas with shareable, embeddable output for non-technical stakeholders.
- Watch for
- Former Mode-only customers mid-migration to Analyst Studio are actively evaluating Hex, Metabase, and Lightdash instead of completing the move.
- Recent moves
- January 2025: ThoughtSpot Analyst Studio reached general availability, formally absorbing Mode SQL notebooks and Python workbench as a paid add-on.
Deepnote
- Positioning
- Cloud Jupyter replacement with real-time collaboration and AI code assist, targeting data teams rather than visualization-first JavaScript developers.
- Customer segments
- Data scientists, analysts, ML engineers at startups and mid-market SaaS companies. Some academic and research use.
- Strengths
- Simultaneous multi-user notebook editing with Google Docs-style co-presence, lowering coordination friction for distributed data teams.
- Watch for
- Revenue at $3.9M with 22 staff and no funding since 2022 signals runway pressure and limited capacity to match better-funded rivals.
- Recent moves
- February 2026: headcount contracted to 22 employees (down from 35), following July 2024 Hyperquery acquisition, suggesting post-integration restructuring.
Sources
- observablehq.com — Product overview, customer list (NYT, Washington Post, NBC News, MIT, HuggingFace, Getty Images, The Economist), core features
- observablehq.com — 2025 product launches, Canvases early access in April, Notebooks 2.0 summer release, AI Assist integration timeline
- www.finsmes.com — Series B funding: $35.6M in January 2022 led by Menlo Ventures with participation from Sequoia Capital and Acrew Capital
- www.businesswire.com — Series B details, CEO Melody Meckfessel confirmation, customer count at time of funding
- observablehq.com — Stitch Fix use case: data scientists and decision makers using Observable to connect APIs, iterate rapidly, and document logic
- en.wikipedia.org — Mike Bostock background: D3.js creator, graphics editor at New York Times, co-founder of Observable with Melody Meckfessel
- www.melodymeckfessel.com — Melody Meckfessel background: VP of Engineering at Google managing 1000+ engineers across 6 offices, co-founder and CEO of Observable
- news.ycombinator.com — Community reception of Notebooks 2.0: feedback on JavaScript syntax improvements, web editor timeline, AI limitations, social features removal
- observablehq.com — Current pricing: Notebook Free tier, Notebook Pro at $22/month per editor, Teams at $90/month
- observablehq.com — Observable Canvases announcement (April 15, 2025): computational whiteboards, AI-assisted analysis, spatial organization of queries and charts