Airtable Limitations That Hit Growing Businesses Hard

I’ll say it upfront: Airtable is a genuinely great product. For a small team tracking projects, managing a content calendar, or running a basic CRM, it’s hard to beat. You get a database that looks like a spreadsheet, with just enough structure to keep things tidy.

But there’s a pattern we see constantly. A business starts with Airtable when they’re small. It works brilliantly. They build more bases, more automations, more views. And then somewhere around 15-30 employees, or a few thousand records, things start to crack. Not dramatically — just enough friction that people start working around the tool instead of with it.

Here’s where those cracks appear, and what to do about each one.

The 100,000 Record Limit

Airtable caps each table at 100,000 records on the Business plan. On the free and Plus plans, it’s 50,000. That sounds like a lot until you realise how quickly records accumulate.

A services business logging 50 jobs per week hits 2,600 records per year in their jobs table. Add line items, communications, and linked records, and you’re at 10,000+ records within a year. An e-commerce company tracking orders, SKUs, and customer interactions can blow past 100,000 in under two years.

When you hit the ceiling, there’s no graceful upgrade path. You can’t just pay more for more records. You have to archive data out of the table, split tables across bases, or fundamentally restructure how you store information. All of which break your existing views, automations, and integrations.

Performance Degrades Long Before the Limit

You won’t make it anywhere near 100,000 records before Airtable gets slow. Most users report noticeable lag at 10,000-20,000 records. Views with complex filters, sorts, or lookups take several seconds to load. Grouped views with rollup calculations can take 10+ seconds.

For a team that lives in Airtable all day, those seconds add up to minutes, and those minutes add up to hours. More importantly, slow tools change behaviour. People stop checking the system as often. They keep their own side-spreadsheets for quick reference. Now you’ve lost the single source of truth that was the whole point of Airtable in the first place.

Relational Data, Sort Of

Airtable markets itself as a relational database, and technically it is — you can link records between tables. But it’s relational in the way a bicycle is a vehicle. True, but limited.

The gaps show up fast:

• No many-to-many relationships without creating a junction table manually, which Airtable doesn’t natively support well
• Lookup and rollup fields are the only way to pull data across links, and they’re slow, limited to basic aggregations, and can’t do anything conditional
• No JOINs, no subqueries, no computed views — if you need “show me all customers who ordered product X but not product Y in the last 90 days,” you’re either writing a convoluted formula or exporting to a real database to run the query
• Linked record fields break easily when records are deleted or tables are restructured

For simple one-to-many relationships — a project has many tasks, a customer has many orders — Airtable is fine. The moment your data model has any real complexity, you’re fighting the tool.

Automations That Break Quietly

Airtable’s built-in automations are useful for simple triggers — when a record is created, send an email; when a status changes, update a field. But they have real constraints that bite as you scale:

• 25 automations per base on the Business plan (and they share a pool of runs per month)
• No error handling — if an automation fails mid-way through a multi-step sequence, there’s no rollback, no retry logic, no way to pick up where it left off
• No conditional branching — you can’t say “if the order value is over $5,000 AND the customer is in Queensland, route to the regional manager; otherwise, auto-approve.” You need separate automations for each path.
• Rate limits and timeouts — automations that call external APIs or process many records hit Airtable’s rate limits, causing silent failures

Most growing businesses end up connecting Airtable to Zapier or Make for their more complex automations. Now you’ve got two systems to debug when something breaks, and your costs have doubled.

Permissions Are All or Nothing

Airtable’s permission model is base-level, not record-level. You can give someone access to a base as a creator, editor, commenter, or viewer. But you can’t say:

• “Sales reps can only see their own deals”
• “The finance team can see cost fields but the sales team can’t”
• “Contractors can submit records via a form but can’t see any existing data”

There are workarounds — you can create filtered views and share those, or use Airtable Interfaces to control what people see. But these are cosmetic restrictions, not real security. Anyone with editor access to the base can see the underlying data. For businesses handling sensitive pricing, employee data, or customer financials, this is a genuine problem.

Reporting That Doesn’t Scale

Airtable can show you summary bars, simple charts, and grouped views. For basic “how many deals did we close this month?” reporting, it’s adequate. For anything more, it falls short:

• No cross-base reporting — if your data spans multiple bases (which it will, once you hit record limits), you can’t report across them natively
• No time-series analysis — tracking trends over months or years is manual
• No calculated metrics — things like “average deal cycle time” or “revenue per customer segment” require exporting data and analysing it elsewhere
• Dashboards are static — no drill-down, no dynamic filtering, no “click this bar to see the underlying records”

Most businesses end up exporting Airtable data to Google Sheets or a BI tool for real reporting. At that point, you’re maintaining two systems — Airtable for operations and something else for analytics.

When Airtable Is Still the Right Call

Despite everything above, Airtable remains excellent for:

• Early-stage businesses figuring out their processes — it’s cheap, fast to set up, and easy to restructure as you learn what you actually need
• Small teams (under 10 people) with straightforward data models and low record volumes
• Content and marketing teams managing editorial calendars, campaign trackers, and asset libraries
• Prototyping — building a working version of a process before investing in custom software

What to Do When You’re Hitting the Walls

First, optimise what you have. Archive old records, simplify views, reduce the number of lookup and rollup fields. This can buy you months.

Second, evaluate whether a higher tier solves your problems. Often it doesn’t — the Enterprise plan raises some limits but doesn’t fix the fundamental performance and relational constraints.

Third, be honest about your trajectory. If you’re growing 20-30% year-over-year and already bumping into these limits, you’ll be in real pain within 12-18 months. Planning a migration now — while things still work — is dramatically less stressful than doing it in crisis mode.

The businesses that handle this transition best are the ones that treat Airtable as a prototype. It taught them what their processes look like. It showed them what data they need and how it connects. That knowledge makes the next system — whatever it is — faster and cheaper to build, because you’re not guessing at requirements. You’ve lived them.