Airtable's Record Limit Turned an $11K Bill Into a $50K Renewal Quote

Texas Flange was paying Airtable $11,000 a year. Their 2026 renewal came back projected at roughly $50,000. Nothing about the business had changed by a factor of four and a half.

What had changed is that they were out of records.

Most writing on the Airtable record limit files it under capacity and "Airtable is too expensive" under pricing, as though they were separate complaints. They are one mechanism with two symptoms, and the ceiling triggers the price.

How the record limit becomes an invoice

Airtable's plan documentation puts the Business ceiling at 125,000 records per base. Team is 50,000. Texas Flange was sitting north of 125,000.

The ceiling is cumulative across every table in a base, not per table. Airtable's FAQ gives the worked example: one table of 50,000 records and two tables of 25,000 both sit exactly at the Team cap. So splitting orders into orders and line items, the correct relational instinct and the thing linked records encourage, consumes headroom rather than creating it. Activity logs, sync staging and webhook receipts all draw on the same pool.

Above Business there is exactly one door: Enterprise Scale, custom-priced, with no published floor. A capacity problem arrives as a sales call. Add AI credits shifting from optional to required, and an $11,000 bill becomes a projected $50,000 one. That $50,000 is a 2026 projection, not an invoice Texas Flange paid. They left before it landed.

Paying up does not fix throughput. The 5 requests per second per-base limit is enforced on every tier, Business included. "Unlimited API calls" lifts the monthly meter, not the per-second one. The glue layer never gets cheaper as you grow into it: Texas Flange was spending another $3,600 a year on Zapier to shuttle data between systems that should have been one system. That is Integration Debt: the unowned, compounding cost of wiring together tools never meant to work together.

What they did instead

$30,000, one time. A custom ERP that replaced Airtable outright and took Zapier with it, because a consolidated system has nothing left to sync. We wrote up the three-week build separately if you want the mechanics.

Xero stayed, at $720 a year. That was deliberate: the accounting worked, the team knew it, and replacing everything at once is how migrations fail.

Efficiency gains returned the full $30,000 in under three months.

That $30,000 was scoped to Texas Flange's workflows, not read off a price list. Build cost tracks scope. More processes, more integrations, or a compliance surface to satisfy, and it goes up.

Where Airtable is still the right answer

Often. Airtable is a good product being used past its design point, which is a completely different problem from a bad product.

If your record count grows with headcount rather than transaction volume, you are probably fine indefinitely. Small team, a few thousand records, a schema you want to change on a Tuesday without filing a ticket: stay. Airtable sells flexibility, and paying for flexibility is rational right up until what you actually need is control.

The real trigger to leave is not a record number. It is the point where records start growing with revenue, because then capacity cost scales with sales while the tool's value scales with team size.

Do this next

Not a demo. A number. Tell us what all you're replacing and we will map your Airtable stack's five-year cost, seats plus glue plus add-ons, against what a one-time owned build would run for your actual workflows. No obligation. If the answer is stay on Airtable for now, we will say so.