My Airtable, Softr or Glide app is hitting limits: when do I move to a custom app?
Short answer: Move when two or more of five measurable triggers are true: you are past about 70 percent of a record or user limit, your monthly bill per person keeps rising, screens are slow for daily users, you run on workarounds to get around limits, or the tool cannot do something your business needs. Airtable is an online database that looks like a spreadsheet, Softr builds apps and portals on top of data, and Glide builds apps from spreadsheets. Do not rebuild everything at once. Move the one workflow that hurts most, first.
You built something useful in a no-code tool, it now runs part of your business, and the limits are starting to cost you time or money.
I'm Ranjit, and I run Blur Studio: 15+ years building software, 120 apps taken live, and clients who have gone on to raise $10M+. I build AI systems and processes that help businesses do a lot more with a lot less time. I also work differently from most of what else is out there, more on that further down this page. Say hello.
No-code tools are a good way to start. Most owners who outgrow them did nothing wrong. The tool did its job until the business got bigger than the plan. This page gives you five triggers you can measure, a plan for moving one workflow at a time, and a three-year cost comparison for a team of 10 to 30 people. The vendors' own pages will not tell you when to leave, so I will.
What the limits actually are
Numbers first, because "we hit a limit" is vague. These come from each vendor's own pricing or help pages, as of November 2026. Check the current page before you decide, since plans change. For the move to a custom app, see what to do when you outgrow Airtable.
| Tool and plan | Records or rows | Other limits that bite | Price (list) |
|---|---|---|---|
| Airtable Free | 1,000 per base | 1,000 API calls a month per workspace, 5 editors | $0 |
| Airtable Team | 50,000 per base | 100,000 API calls a month, 25,000 automation runs a month | $24 per collaborator monthly, $20 annual |
| Airtable Business | 125,000 per base | Unlimited API calls, 100,000 automation runs a month | $54 per collaborator monthly, $45 annual |
| Softr Basic | 50,000 records | 2,500 workflow actions a month, 5 team plus 5 client app users | $19 a month |
| Softr Pro | 500,000 records | 10,000 workflow actions a month, 10 team plus 50 client app users | $99 a month |
| Softr Business | 1,000,000 records | 25,000 workflow actions a month, 30 team plus 100 client app users | $329 a month |
| Glide (classic) Business | 25,000 rows from spreadsheets, 100,000 on high-scale tables | 30 users included, 5,000 updates a month, then 2 cents per update | From $199 a month billed yearly |
Two notes. Airtable bills collaborators, not just editors: on Team, anyone with Commenter permission or higher is billable; on Business, Editor or higher. Softr's pricing page has a monthly and yearly toggle, so confirm which price you are looking at; the Softr prices here are as listed on that page in November 2026. And Glide now sells two products: the original Glide ("Classic") and a newer AI app builder called GlideOS with separate plans from $25 a month. Glide says Classic apps cannot be seamlessly moved to GlideOS, which is its own reason to stop and think.
Five triggers that mean it is time
These are my own rules of thumb, built from the limits above. Two or more true at once means start planning.
1. You are past 70 percent of a hard limit
If your Airtable base holds 35,000 of Team's 50,000 records, or your Glide sheet holds 18,000 of 25,000 rows, you will hit the wall sooner than you think. Records grow faster than plans. Hitting the limit in the middle of a busy week is worse than planning a move now.
2. The price per person keeps climbing
Airtable charges per collaborator. Add five more staff and the bill goes up by five times $20 to $54 a month. Glide charges per extra user beyond 30 and per update beyond 5,000. Softr's plans step up in tiers, with extra app users sold as add-ons on some plans, so check what your next ten users would cost on its pricing page. When the next hire adds a visible chunk to a software bill, count it.
3. Screens are slow for people who use them all day
If staff wait several seconds for a list to load, and they do it dozens of times a day, the hidden cost is bigger than the subscription. Ask them how many minutes a day they lose. Five minutes a day across 20 people is over 400 hours a year.
4. You run on workarounds
Signs: a second spreadsheet to hold the overflow, a Zapier chain (Zapier connects apps so one event triggers another) to copy data between the tool and something else, a person who "knows how it works", or monthly automation runs that keep running out. Each workaround is a place where things silently break.
5. The tool cannot do what the business now needs
Examples: different customers must see only their own records with detailed rules, pricing logic that depends on ten conditions, or a connection to a system that has no ready-made connector. When you spend more time bending the tool than using it, you are building custom software the hard way.
If only one trigger is true, you may just need a better plan or a cleaner base. My guide on how to know you have outgrown spreadsheets covers the earlier stage of the same problem.
A three-year cost comparison (20 people)
Here is a simple model for a team of 20 people who all need to edit data. Tool prices are the list prices above. The custom app figures are made up for the example: replace them with real quotes.
| Option | Monthly | Three years |
|---|---|---|
| Airtable Team, annual billing, 20 people | $400 | $14,400 |
| Airtable Business, annual billing, 20 people | $900 | $32,400 |
| Softr Business (30 team users included, price as shown on the page) | $329 | $11,844 |
| Glide Business, 30 users included, no extra updates | about $199 | about $7,164 |
| Custom app (made up): $12,000 to build, $200 a month to host and look after | $200 | $19,200 |
What this shows. Custom is not automatically cheaper. Against Airtable Team, Softr or Glide at this size, the subscription is cheaper over three years. Against Airtable Business, custom wins. The point is that the tool prices keep climbing with headcount and records, while the custom running cost stays fairly flat.
The break-even formula: build cost divided by monthly savings. If you pay $900 a month for Airtable Business and a custom app costs $200 a month to run, you save $700 a month. A $12,000 build pays back in about 17 months. If your tool costs $329, you save $129 a month and the same build takes 93 months, so staying put is right. For real build costs, see my guide to AI app development cost. For the broader decision, build versus buy software goes into more detail.
Add the costs this table leaves out: time spent on workarounds, mistakes from manual copying, and the risk of a plan change you cannot control.
Move one workflow first
The biggest mistake is a full rewrite. It takes months, you lose your working system while you wait, and you find out about the surprises at the end. Do this instead.
- List your workflows. Write each one as "who does what, with which data". A typical Airtable base hides five or six.
- Score each from 1 to 5 on pain (slow, error-prone, hits limits) and on value (money or time it touches).
- Pick the one with the highest pain and clear edges. Good first picks: order intake, a client portal, a booking or quote flow. Bad first picks: anything that touches every other table.
- Export the data. Check how your tool lets you get data out (for most, a CSV file is the starting point), and do a trial export now so you know what you have.
- Build the new version beside the old one. Keep the old tool running.
- Run both for two to four weeks. Staff use the new one, and you compare numbers with the old one.
- Cut over, then archive. Keep a read-only copy of the old data for a few months.
Then repeat for the next workflow, or stop. It is a perfectly good outcome to end up with one or two custom apps for the core processes and leave the rest in the tool you already have.
What to ask a developer before you commit
- Who owns the code and the database, and can I get both if we part ways?
- How will you move my existing data, and how will we check nothing was lost?
- What will it cost each month to host and look after, in plain numbers?
- What happens when I want to change a form or add a field? Who does it, and how long?
- Can you show me the finished version of the first workflow before I pay for the second?
For who to hire and in what shape, see fractional CTO versus development agency. If you want to compare how other no-code tools handle the same limits, my guides to Airtable alternatives, Softr alternatives and Bubble alternatives (Bubble is a no-code builder for web apps) are a good start. And no-code versus AI app builders explains the newer option of describing an app in chat.
Who should not move yet
- Fewer than about 10 regular users and under half of every limit: stay.
- Your process is still changing every month: the flexibility of a no-code tool is worth the cost. Custom software is best once the process is stable.
- You have no one to look after a custom app afterwards: budget for that, or stay.
How I'd approach this
I would not start with a quote. I would start with a one-page limits check: records used against the limit, monthly bill against headcount, and a list of the workarounds. That takes an hour and tells you which triggers are true.
If one or none is true, I would stay and tidy the base. If two or more are true, I would pick the single most painful workflow and price a custom version of just that. Compare that price to the break-even formula above. If the payback is under about two years, it is usually worth doing. If it is over three years, stay on the tool and revisit when your team grows.
The most common mistake I see is moving too early because the tool feels messy. Messy is a data problem, and a custom app will not fix messy data. Clean first, then decide.
Questions people ask
What are the Airtable record limits?
As of November 2026, Airtable's help center lists 1,000 records per base on Free, 50,000 on Team and 125,000 on Business, with the per-table maximum also counting toward the base limit. Check Airtable's plans page for the current numbers before you plan.
What are the Softr database and user limits?
Softr's pricing page lists 50,000 records on Basic, 500,000 on Pro and 1,000,000 on Business, with app user limits of 5 team plus 5 client users, 10 plus 50, and 30 plus 100. Team users share your business email domain; the rest count as client users. Check the page for the monthly or yearly price toggle.
How many users can a Glide app have?
Glide's classic Business plan includes 30 users, with extra users at $5 each on annual billing or $6 monthly, plus 5,000 updates a month and 2 cents for each update after that. Glide also has a newer product, GlideOS, with separate plans and no seamless way to move classic apps across.
Can I keep Airtable and add a custom app?
Yes, and it is often the best first step. Keep Airtable for the parts that still work and build a custom app for the one workflow that hurts. Airtable has an API, but its call limits depend on your plan, so check them before syncing both ways. See my Airtable alternatives guide for other options.
How long does moving one workflow take?
It depends on the workflow. As a planning guide, allow time for a data export check, a build beside the old tool, and two to four weeks running both. A single clear workflow is a much smaller job than a whole base. Ask any developer for a written plan and a fixed price for that first piece.
Will a custom app cost less than my subscription?
Sometimes. It depends on headcount and plan. In the made-up example above, custom beat Airtable Business at 20 people but not Softr or Glide. Use the break-even formula and real quotes. My AI app development cost guide explains what drives the price.
More in this topic: Problem solving: AI-built and no-code apps (26 guides) →