Skip to content

What is Airtable?

Established in 2015 and headquartered in San Francisco, Airtable is a venture capital funded cloud collaboration company. Airtable describes itself as a "connected apps platform" for team-based collaboration, allowing its users to create both "back-end" relational databases and "front-end" interactive applications for data filtering and search, customized workflows, data entry, and reporting with little to no coding required.

While Airtable's "freemium" subscription service, primarily marketed as a business tool, been also adopted by Digital Humanities (DH) projects as a data repository and project management tool. Airtable's familiar user interface makes the learning curve potentially lower than other database software. However, for proper use as a DH tool, projects will still need to engage with the intellectual the work of data modeling, documentation, and research data management.

How Does Airtable Work?

Airtable's central feature is its ability to build relational databases with little to no coding. Bases, as Airtable calls them, are created through a point-and-click user-interface that is reminiscent of a spreadsheet, providing functionality for filtering data, linking between records, and sharing information with the public.

Airtable bases, themselves, are flexible and can support many use cases, including:

Airtable is also designed with team collaboration in mind, and can support a range of needs that arise for digital humanities teams:

How is Airtable Different from a Spreadsheet?

Spreadsheets, sometimes called tabular data, are one of the most common form of data storage. Standard spreadsheet formats like Google Sheets, Microsoft Excel, Apple Numbers, and plain-text CSV file, organize data into rows and columns.

In general, spreadsheet software is agnostic about the data within each cell. It does not care if you input text from a book, numerical observations, or links to websites – just that every value is stored in a particular location on the grid.

In a spreadsheet, the relationship between two values is implicit. For example, we might assume that because two values are in the same row, they are observations about the same object or event. Or that because two values are in the same column, they are about same characteristics. However, this meaning is only implied. To a computer, the decision to put data in the same row or column is an organizational decision. It does not have any explicit semantic meaning.

On the other hand, a relational database gives data additional structure by making these relationships explicit. Instead of adding information to cells in a grid, databases work with records. Each record in a database has a unique identifier (its name) and a set of fields that describe attributes of the record. A database table contains only records of a single type, and all records of the same type have the same fields.

Databases are picky about what type of information goes into each field. For example, you can specify that name must contain alphanumeric characters, or that US_zip_code must contain exactly five digits. This enforces a level of uniformity on data that spreadsheets do not.

If we think of a relational database like a spreadsheet, the structure is similar to how we often think about rows and columns: each row is a single object, and each column describes a different aspect of that object. The difference here is that this relationship is made explicit.

Airtable functions like a relational database. Each table of an Airtable base contains one record type and a set of specified fields where each field has a particular data type. Possible data types, among others, include plain text, formatted text, dates, URL, single- and multi-select dropdowns, and formulas.

Another central database feature that is incorporated into Airtable is the ability to create relationships between records. Whether records are in the same or different tables, Airtable provides a "linked record" field type that allows you to connect them. For example, you can have an authors table and books table, with a field in each that relates each book to its authors. This process of establishing relationships between objects opens up a world of possibilities for humanists creating data about complex objects.

How is Airtable Different from a Relational Database?

SQL-based relational database management systems (RDBMSs), like MySQL, PostgreSQL, and Microsoft SQL Server, can have a high learning curve. They often require knowledge of server hosting, the command line, and SQL in order to create a database. Depending on your institution's IT and digital scholarship services, the level of support required to host a shared database may be beyond its scope.

While Airtable functions like a relational database, it differs in key ways from a traditional RDBMSs. Airtable attempts to reduce the difficulty of creating a database by providing built-in hosting, offering a user-friendly interface, and reducing the required perquisite programming and data modeling knowledge. This makes Airtable worth consideration for Digital Humanities teams with various degrees of technical background.

Cloud-Based Workspaces vs. Hosting

Unlike most RDBMSs, Airtable does not need to be independently hosted on a server in order for multiple people to access the data. As a cloud-based tool, Airtable bases can be created and accessed directly in a web-browser (or desktop app). Bases are assigned to a workspace, which allows managers to share access and control user permissions. All pricing tiers can create an unlimited number of bases, and free workspaces can have up to 5 editors.

Spreasheet-Like Interface

Part of what makes Airtable inviting is that it's interface is modeled after a spreadsheet. Tables are organized as separate sheets, records as rows, and fields as columns. New tables, fields and records can be created directly through the UI with no coding required.

Data entry, as well, works like a spreadsheet, with users clicking-into a cell to add information. Airtable also offers an "expanded view" of each record, that mirrors a form, allowing for multiple input options. Tools like DB Browser and DBeaver have brought similar functionality to SQL databases in the past, but their learning curve is still much steeper.

Many-to-Many Relationships

Airtable is very permissive about the relationships it allows users to create in its bases. Notably, Airtable directly supports the creation create many-to-many relationships without a junction table. For those with a background in database design, this may seem like a red flag! However, it means that Airtable users do not need to be familiar with rules around modeling relationships to start building a relational database. Airtable does the work of breaking down these relationships on their backend

This is not to say that Airtable bases can't benefit from entity-relationship modeling and database normalization (especially if the Airtable base will be part of a larger tech stack). However, this is not requisite information for creating simple Airtable bases. We'll return to this topic in Creating Relationships.

Not Just a Relational Databases

While Airtable's features lend itself to creating relational databases, its has many features that make it a useful tool for both relational and non-relational datasets.

Further Reading

This tutorial is licensed under Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).