Airtable

For the right use case, yes — Airtable is the best no-code relational database for ops, marketing, and content teams. It earns a 6.3/10 here, with an 8.5 on ecosystem integrations and an 8.0 on AI features. The AI suite (Omni, Superagent, Hyperagent) is the most mature in this tool category. The honest caveats: three independent billing meters can each force an upgrade at different times, external client write access costs extra, and the Team→Business per-seat jump is 125% with no intermediate tier. If the database is the product and most users are viewers rather than editors, Airtable delivers strong value.

No. After two rounds of layoffs in December 2022 (~254 people) and September 2023 (~237 people, per Forbes), Airtable rebuilt to approximately 900-947 employees. $478M ARR in 2024 with 27% year-over-year growth, approximately 90% gross margins, and $100M+ free cash flow (Sacra, August 2025) are the operational data. The secondary-market valuation reset from $11.7B (2021 peak) to approximately $4B (CEO Liu, January 2026) is real and worth knowing — but it reflects ZIRP-era excess in the 2021 figure, not operational decline. Airtable is a financially healthy, cash-generating business executing an aggressive AI pivot.

Start on Free (5 editors, 1,000 records, 100 automation runs/mo) — it's genuinely functional for small teams. Upgrade to Team ($20/seat/mo annual) when you cross 1,000 records or need 25K automation runs. The cheapest sustainable working configuration is Team at $100/mo for 5 editors with unlimited free viewers. The trap to avoid: giving editor access to people who only need to view data. Every person with edit permissions is a monthly seat — viewers are free on all plans.

Airtable requires an upgrade to Business ($45/seat/mo annual vs $20/seat on Team) — a 125% per-seat increase with no intermediate tier. If you have 5 editors, you go from $100/mo to $225/mo. There are two alternatives: (1) split the base into two linked bases of 25,000 records each, keeping both on Team tier (works architecturally, but complicates cross-base rollups); or (2) upgrade to Business and gain 125K records, SSO, audit log, and admin panel. Teams that project record growth before committing avoid discovering this mid-project.

Read-only, yes — shared view links are free on all plans including Free, and unlimited external viewers are free. Write access (creating or editing records) requires either Airtable Portals (~$120-150/mo on Team/Business tiers for 15 guests) or a third-party portal builder: Softr, Noloco, or Stacker (all add $50-200+/mo). Many teams discover this cost after building the internal base and showing clients a prototype — which is why external write access requirements should be the first question in any Airtable project scoping.

Airtable has the most mature AI suite in this tool category as of mid-2026: Omni (June 2025) for natural-language interface building and data analysis, Superagent (January 2026) for agentic workflow automation, and Hyperagent (2026) extending this further. The DeepSky acquisition (October 2025, ex-OpenAI leadership) signals long-term structural commitment. AI credits are included in all plans. The honest weakness: Omni-generated interfaces are block rearrangements of pre-built components — they cannot produce custom-designed branded portals. For internal data work and analysis, Omni is impressive. For client-facing designed experiences, a separate front-end is still required.

Structured relational data management where the database is the deliverable — CRMs, editorial calendars, project trackers, content pipelines, event management, HR ops, and ops tracking. Airtable excels when: most users are viewers (not editors), data needs linked-record relationships across tables, and multiple team members collaborate on the same dataset simultaneously. It is not the right tool for polished branded external apps (those need a front-end layer), native mobile apps, full-stack web applications with complex logic, or SEO-driven public websites.

Airtable is a relational database first — linked records, rollups, 40+ field types, complex formulas, and automation rules are the core. Notion is a document tool with database views bolted on — it works for simple lists and tables but doesn't handle true many-to-many relationships or formula-driven rollups as cleanly. For a team whose primary output is structured data (not documents with embedded data), Airtable's relational model is meaningfully more capable. For a team that primarily writes and occasionally tracks data, Notion's unified document+database is simpler. There is no Notion review currently in our manifest, so we can't link to a direct comparison.

Three structural risks: (1) Vendor lock-in — no SQL dump, no automation export, no schema export; leaving requires rebuilding from CSV exports + documentation. (2) Unilateral pricing changes — the February 2026 Free-tier record cap cut (1,200→1,000, no notice, bases went read-only) is the clearest demonstration that Airtable controls the terms of what you've built on the platform. (3) Per-seat pricing at scale — the three billing meters (seat + records + automation runs) become increasingly costly as teams and data volumes grow. These risks don't make Airtable a bad choice for many use cases — they make the migration path and platform dependency worth thinking about before committing.

If your base has grown beyond 50,000 records and the Business-tier cost is significant, or if you need a branded external portal that Airtable Interfaces can't deliver, a migration makes financial sense to evaluate. The data migration (CSV → Supabase/PostgreSQL, modeling linked records as proper foreign keys) is typically 1-2 weeks of data engineering; the front-end rebuild is the larger scope. RapidDev handles Airtable → Supabase + custom Next.js portal migrations and offers a free scoping call to size the project at rapidevelopers.com/contact. The data modeling phase is almost always the critical path — Airtable schemas that look simple in the UI often have surprising complexity when mapped to relational SQL.