I created databases in Airtables with info, photos and links to digital music products. I currently have entries in different workspaces, as I use the free tier, and can only have a limited amount of entries per database.

Long term, I'd like to transfer it all over to another platform, and have a site with a database/knowledge base and wiki style layout to it. Wondering if there are current options that could work for this or if it is something I'd need custom developed?

I would love to hear any ideas regarding softwares that create a front-end site, preferably that integrates directly with Airtable, but not 100% necessary. I have tried Softr, but wanted to weigh it against other options. Not super knowledgeable with this area of things, so any tips or info would be greatly appreciated!

Did anyone catch this recent episode of HardFork where they discuss the Airtable acquisition? It’s about 2/3 of the episode in. Should we be nervous?

“Airtable was valued at $11.7 billion when it was acquired, but recently though, Bending Spoons was able to get Airtable for an enterprise value of $1.29 billion. So what do we make of the sharp decline here as we run the numbers?”

Hey guys, 20yo uni student here trying to figure out how to pay my bills.

I’ve been reading up on the Airtable + Make. com + Softr stack. People online claim you can learn this stack in 1–2 months and start charging small businesses $1,500+ per project to build client portals and automate workflows.

To be completely honest, this sounds too good to be true.

Most entry-level jobs require a 4-year degree and pay $15–$20/hr. Why on earth would a business pay $1,500+ for a system built using tools someone learned in a few weeks?

I want a brutal reality check from people actually doing freelance/agency work in this space:

1. Is the money real? Are small businesses actually paying $1k–$2k for basic Airtable/Make builds, or is this just hype sold by course creators?
2. What are you ACTUALLY being paid for? If the tools are easy to learn, why don't business owners just spend a weekend setting it up themselves?
3. What is the hard part no one talks about? Is it landing clients? Scope creep? Edge cases breaking?

I’m ready to put in the work to learn the technical side, but I don't want to chase a fake ghost skill. Tell me like it is.

Was setting up a proper backup of one of my bases and went to check what the export actually contains. Posting because I did not know this.

Attachment URLs expire two hours after the API returns them. It's in the docs (https://airtable.com/developers/web/api/field-model), along with the recommendation to download the files rather than store the links. So the attachment column in a CSV export is a list of links that are dead by tomorrow: the records survive, the files do not.

While I was in there, two more things I had assumed were covered and are not:

- Linked records export as bare rec... IDs, with nothing alongside them to rebuild the relations.

- Field types, select options and formula definitions are not in the export at all.

Not bugs, just what the export is. But if your backup plan is "I can export a CSV whenever I want", it's worth testing that assumption on a base you care about - specifically, open one of the exported attachment links tomorrow and see what happens.

If anyone has found a way to keep attachments through the built-in export I'd genuinely like to hear it.

Don't: think in terms of cells.

Do: think in terms of records and fields.

In Airtable, a record is one line item in a table. It represents one of those "things" that live in your table. So if you have a table called "Projects", each record is a Project.

The fields of that table are the attributes of the project. The primary field, most simply, the name of the project, is like the noun. All the additional fields are like adjectives.

For instance, a record with primary field "Website Upgrade" might have fields like "Due Date" and "Status".

******

Don't: make random tables like "tabs" in a spreadsheet

Do: link related tables to model real-world relationships

In spreadsheets, sometimes tabs are made for new projects, or new years, or basically anything. In Airtable, tables have a real purpose.

They're home for one kind of data. Like a folder in a filing cabinet.

A "Projects" table should have records and fields for attributes on your projects.

A "Tasks" table should have records and fields for attributes of tasks.

Then, use a Linked Record field to link them together. When each Project record is linked to its children Task records, now you've created a relationship between those tables that reflects the actual relationship between projects and tasks.

******

Don't: treat the data layer as a UI to "do the work"

Do: use Interfaces to support your actual workflows

In the world of spreadsheets, there's really no UI on there. You create a table or a list of rows/columns and trust everyone to fill stuff out correctly.

Even with a variety of Views in the Airtable Data layer, we can still do better with Interfaces.

Interfaces is where you can show specific fields, to specific people, in a specific order, to better support the step 1, step 2,...Step n of your actual workflows.

For instance: click on a project, see its children tasks, update the completion status on a task, right there from within the project. This is instead of having to click around from table to table finding the right task. It actually supports the workflow.

******

I hope this quick overview was helpful! Feel free to comment or DM with any questions.

- Jonathan

Hi all,

I have jumped into Airtable over the last few months and it has been great.

I need some assistance with how to structure something specific. I've tried a few solutions but they all have some downsides.

I will keep this generic as it gets complicated pretty fast.

I have a set of "Endpoints", which are physical devices. They have a table of their own with a couple dozen fields of misc. information.

I have another table that is a list of "Steps", essentially a process I follow. This process gets applied to each device. The catch is that not every step is needed, some steps can be performed out of order, and some steps may need to be repeated. It will change relatively frequently, and is different per device. I also need the process to be dynamic enough that I can add/remove/edit steps without making large bulk changes. I have steps in their own table, so they are easy enough to update.

I am needing to track a large amount of Endpoints (200+) and the status of each one by step.

What I have now is 3 fields on the Endpoints table:

1. "Completed Steps" - reference to Steps table, allows multiple selection
2. "Next Step" - reference to Steps table, allows only 1 selection
3. "Remaining Steps" - reference to Steps table, allows multiple selection

This works okay and fulfills some of my requirements. I can see:

• History of what steps were completed
• List of remaining steps
• Next step to perform

What I don't like about this solution is that I have to manually add/remove items from each list. If I complete a step, I have to add it to "Completed", change the "Next Step", and remove it from "Remaining"

Because these are a dynamic link to the Steps table, I'm having trouble getting this done with automations. But I may be missing something.

Is there a way that I can maintain these "Completed" and "Remaining" steps automatically? Or another way to structure this kind of thing?

Thanks in advance

Hi all,

We are exploring a pipeline to have teams submit a word document and a Excel table via a form. The word doc is only about 7 or so pages usually and the Excel has about 400 rows and 5 columns of data.

The expected behavior is that the data entered in specific fields within the documents are to be populated across multiple tables, creating new records where applicable.

I would be curious to hear people's experiences with this, including what processes and tools you have considered, if you had to make a choice and how you juggled tradeoffs.

We have access to Claude code and there is openess to having the submission analyzed in Claude if it improves accuracy and changes pushed through via the mcp. My take is that there are various ways to tackle this kind of problem via AI scripts, ai fields within the base, and Claude code/mcp server but just not familiar with all the detailed tradeoffs to come to a confident decision.

Edit: Thanks all for your guidance! The Word document has fields where expected values are meant to go, but I can't ever guarantee someone will use it in the way that it was designed.The Excel table has about 5 columns, each with drop down options for field validation and a numeric field for the values that the user will enter, so would say it is structured. We have been shying away from interfaces mainly because of the seat cost in getting other teams within our organization to collaborate and use Airtable. I am also exploring Softr and Noloco as possible ways to reduce cost but also wanted to see if we could use what is available in our available tech stack.

I'll do some more research into the recommendation for parsing, as flag the cost variability associated with an AI solution that others on the team have pushed for- appreciate the ideas y'all have shared. Thank you!

I'm looking at using Airtable as a central supplier-facing database rather than maintaining a separate Dropbox/OneDrive-style file sharing system.

We already have a paid Airtable account and my idea is to create supplier-specific Interfaces containing the relevant product information and files/attachments — things like tech packs, artwork, fabric information, approvals, PO information etc.

Our suppliers are mainly based in mainland China.

What I'm trying to establish is:

• Does Airtable work reliably in mainland China without a VPN?
• Do Airtable Interfaces and attachments/files load and download normally?
• Has anyone used publicly shared Interface pages with Chinese factories/suppliers?
• Alternatively, if I invite suppliers as interface-only viewers, can they simply create their own free Airtable account and access the Interface without us needing to purchase additional seats for them?
• Are there any issues with Airtable login/verification in China?
• Do attachments hosted within Airtable work reliably, or are certain file/CDN services blocked or slow?
• Has anyone actually run a supplier portal/product development workflow like this with Chinese factories?

The suppliers would mainly need view/download access rather than editing the underlying Airtable base.

I'm particularly interested in real-world experience from anyone working with factories in China, as technically being accessible and actually being reliable enough for daily production use are two different things.

Thanks!

When: Wednesday, August 26 at 7am PT

Who: Anne Marie Rützou Bruntse, Senior Program Manager at Airtable

Hey everyone, excited to do this!

I've been a borderline fanatic Airtable builder since 2019. I discovered the platform as a customer looking for a better solution than a massive and very complicated spreadsheet I'd created to run our workflows. Now I work on Airtable Academy, helping design the courses, certifications, and learning paths that help other people (hopefully) fall in love with Airtable as well.

One of the things I own is AcademyOS, which is the Airtable app that powers our content workflows for the Academy team. It tracks everything from an incoming course request to the day something goes live: lessons, deliverables, feedback, sprint planning, etc.

Recently I've been experimenting with AI agents to help the team ship content faster and find ways to automate the trickier parts of the workflow, including triaging feedback, content requests and product changes, creating sprint plans, drafting course content and uploading completed content to the Academy. I currently have a fleet of 11 agents, each owning different parts of the workflow.

Drop your questions below — I'm happy to talk app design, content operations, automations, AI features and my experience with implementing AI agents in Airtable. See you at 7am PT on August 26th!

I’ve been working through a workflow that was getting close to Airtable’s 50-automation limit, and the surprising part wasn’t the limit itself. It was realizing how much duplicate logic had accumulated over time.

In a sample setup with about a dozen tables and roughly 18,000 records, the biggest improvements came from:

• Removing old disabled automations.
• Combining status-based workflows that shared the same trigger.
• Adding explicit “Needs Processing” flags instead of triggering on every edit.
• Moving non-urgent batch updates to n8n.
• Logging failures instead of letting them disappear silently.

The first version didn’t work well because every small record update triggered another run. After adding clearer state fields, daily runs dropped from roughly 1,400 to 520.

At some point, adding another automation stops being the solution. How are you handling automation sprawl in larger bases?

Airtable is great, but it would be sooo much better if there was a way to embed external websites in the interfaces. Has anyone figured out a good workaround for this?

I use Airtable for client portals among other things, and I wanted to embed the web page where clients can pay their invoices. I tried the HTML option, which didn't work, so instead I made a button that links to the external web page. I'd just rather have it so clients could pay their invoices without leaving their client portal.

Trying to prefill a classic Airtable form via ?prefill_FieldName=Value. Field name matches exactly, but the value never populates. Tested with a single Text field, signed in as owner, no filters on the field.

This is for a form built in the Forms tab on the website, so it's not an "interface" form as far as I can tell (not sure what that is).

I run a consulting firm and I work with Airtable a lot. Thought I'd share my honest review of the platform.

The easiest way I describe it: what you'd get if a spreadsheet, a database, and an automation engine had a baby.

What it does well:
Airtable feels like a spreadsheet but runs like a database underneath. You get 40+ field types, multiple views (grid, kanban, and more), real-time collaboration, and automated notifications.

AT basically pioneered the no-code automated workflow category. Once the automation setup is dialed in, it runs smoothly. Slack alerts on record updates, new tasks created straight from form submissions, you name it. Forms, project management, a relational database, a lightweight CRM... Airtable bends to whatever your workflow needs.

The AI layer takes it further. Airtable AI can generate content, process info, and build entire applications without code, using models from providers like OpenAI and Anthropic. The Cobuilder feature lets you describe an app in plain language and it generates the tables, fields, and relationships for you (great for fast prototyping).

Where marketing teams really benefit:
This is where Airtable shines. I love recommending Airtable to marketing teams (and go to market teams). My marketing teams use it for campaign tracking, content calendars, and lead management (and now the AI layer drafts campaign copy directly from existing content tables). One tool ends up covering what used to take three or more. I've built entire go to market and marketing engines on Airtable and I can't imagine any other place I'd have built them.

The honest tradeoff:
There's a REAL learning curve. The same flexibility that makes Airtable powerful also makes it confusing at first. It definitely won't hold your hand through it. I love introducing teams to Airtable, but I feel their frustration when they're learning it.

Pricing jumps once you need paid features too, from their (pretty generous) free tier to $20/user/month+. I do hate that you cap out on automations and can't even purchase more, let me add them on, I'll give you my money!

My opinion: If your team needs structure without rigidity (especially for marketing, ops, or project workflows) Airtable is one of the most versatile tools out there.

What's everyone elses thoughts?

(sorry in advance for formatting etc, I'm new to reddit).