Making GRID's spreadsheet engine 10% faster

Models are free-form sheets with formulas that typically generate insights in the form of output based on a set of inputs.

Tables are the structured sheets that typically have (somewhat) defined columns detailing the attributes of each row of data. Insights from tables generally come from filtering, grouping, aggregating and sorting the row-level data.

As you point out, a lot more attention has been paid to the table use-case where people are essentially using their spreadsheets as small databases. Historically you could say this the usage pattern that gave birth to the modern BI tools (such as Qlik, Tableau and PowerBI). More recently, this is where companies like Airtable and Smartsheet play, but also to some extent what Coda and Notion are trying to achieve with their tables/databases.

The models are where GRID shines. The unique value GRID offers is the ability to easily create and publish an interactive web document on top of spreadsheet calculations, whether the models are built in Excel, Google Sheets or in our own built-in GRID Sheets.

I don't think this specific category has a name, but more broadly we've seen people talk about "next-gen spreadsheets" where - in addition to GRID - Equals, Rows and Spreadsheet.com also play, each with their own unique flavor.

GRID is clearly the solution that leans most heavily into models and interactivity (hence the focus on our engine as evidenced e.g. by alexharri's article).

While all the other solutions have capable calculation engines, I think it's fair to say that they are more focused on table-like use cases. Equals leans heavily into data connectivity, governance and reusability of spreadsheets, while Spreadsheet.com focuses a lot on project management use cases. Both offer an otherwise very traditional spreadsheet user experience. Rows has also focused a lot on connectivity, but lately they have leant heavily into AI to generate insights from data in their spreadsheets.