Airtable vs. Custom Software: When You've Outgrown the Super-Spreadsheet

Airtable might be the best onboarding ramp in software. A spreadsheet anyone can use, with just enough database underneath to feel powerful. Most teams that come to us started their operations in Airtable, and they were right to.

This post is about the other end of the ramp: how to tell when you've driven off it.

//What Airtable is genuinely great at

Credit first, because the tool earns it:

• Speed to working system. An intake tracker or content calendar exists by lunch.
• Zero training. If your team can use a spreadsheet, they can use Airtable.
• Cheap experiments. Testing whether a process even needs software? Airtable is the right lab.

If that's where you are, stay. Nothing below applies to a five-person team tracking projects in two tables.

//The ceiling, and how teams hit it

The pattern we see is consistent. Somewhere between year one and year three, the helpful super-spreadsheet becomes the thing everyone works around.

The automation ceiling. Airtable automations handle "when record changes, send email." They struggle with real logic: multi-step branching, retries when an API call fails, anything conditional on data living in another system. So teams bolt on Zapier, and now the process lives in two subscriptions and breaks in the seam between them. We've seen a single intake workflow spread across Airtable, Zapier, and three email templates nobody could find.

The permissions ceiling. Airtable's sharing model is base-level. The moment you need "sales sees contacts but not costs, contractors see their jobs only," you're either paying for Enterprise or building parallel bases with synced tables, which is how data starts diverging.

The interface ceiling. Interface Designer is fine for internal dashboards. It is not a client portal. If customers or vendors need to log in, upload files, and see only their own records, you're outside what the tool wants to do.

The record-count wall. Bases cap out (50k records on Business plans), and performance degrades well before the cap. A growing business generates records forever. The wall isn't hypothetical, it's scheduled.

The per-seat math. Airtable charges per collaborator. At $20-45 per seat per month, a 15-person team runs $3,600-8,100 a year, every year, for what is often one core workflow. Add the Zapier tier and the portal tool you bolted on for the gaps, and the "cheap" system costs five figures annually. The math mirrors the SaaS cost spiral we've broken down before.

//The tell: when the workarounds become the job

Record counts and seat prices are measurable, but the real signal is behavioral. You've outgrown Airtable when:

• Someone on your team is the designated "Airtable person" and process questions route through them
• New workflows get designed around what Airtable can do, not what the business needs
• You maintain a doc explaining which base is the real source of truth
• The phrase "don't touch that view" gets said in meetings

That's your team adapting to the tool. Software is supposed to adapt to you. It's the same trap as fighting a CRM whose data is always wrong, the tool stops being leverage and starts being overhead.

//What replacing it actually looks like

The good news: an Airtable base is the best spec document we ever receive. Your tables are the data model. Your views are the screens. Your automations (and the Zapier duct tape) are the requirements list. Teams that arrive with a mature Airtable setup get scoped faster than anyone else, and a build like that ships in days, not months, here's the process.

A custom replacement typically means:

• Your workflow, not a workaround. The branching logic Zapier couldn't do is just code.
• Real permissions. Roles, client portals, contractor views, table stakes in a custom app.
• No record limits, no per-seat pricing. Headcount growth stops being a software line item.
• One system instead of a stack. The Airtable + Zapier + portal-tool + form-tool bundle collapses into one thing you own.

//The honest decision framework

Stay on Airtable if the spreadsheet model still fits, headcount is small, and you're still discovering your process. Move when the subscription stack costs more per year than a build, when workarounds are consuming real hours, or when customers need to touch the system.

And if you're not sure, run the ninety-second math: seats times price times twelve, plus every tool taped to the side. Compare that annual number, forever, against building the workflow once and owning it. For teams past a certain size, the spreadsheet-on-steroids quietly became the most expensive tool in the stack.