Google Spreadsheet vs Hikoutei

A plain Google Spreadsheet is a great collaboration surface. Hikoutei is a typed repository and safe write layer for Google Sheets-backed MVPs — it keeps the spreadsheet as the human-facing view while moving application reads and writes to local SQLite. The two are not competitors; they serve different sides of the same workflow.

Start with a plain spreadsheet when

• People are the only readers and writers, and the data is small.
• There is no application logic over the rows yet.
• You need ad-hoc filtering, formatting, and comments as the primary interface.

A raw spreadsheet gives you instant human collaboration with zero infrastructure. If that covers the whole workflow, there is nothing to build.

Add Hikoutei when the spreadsheet becomes an application

The moment application code needs to read or write the same rows, a spreadsheet becomes a poor database: reads are slow and quota-bound, writes race with human edits, and there is no schema or transaction boundary.

Hikoutei keeps the spreadsheet for people and moves the application to SQLite:

Concern	Plain spreadsheet	Hikoutei
Human review and editing	Native	Native (async projection + User_Input observation)
Typed application reads	Manual row conversion	Typed entity API over local SQLite
Application writes	Direct API calls, quota-bound	Local commit + durable outbox
Schema drift	Undetected until parse errors	Header validation and clear failures
Overwriting human edits	Easy to do by accident	Conflict-aware updates with compare-and-set evidence

What Hikoutei is not

• Not a database replacement. SQLite is the authority; Sheets stays a projection and input surface.
• Not a raw Sheets API wrapper. Direct row and cell manipulation remains limited; if that is all you need, use google-spreadsheet or @googleapis/sheets directly.
• Not a transaction-safe database on top of Sheets. Sheet updates are asynchronous; a successful flush() never means the Sheet write has completed. In a sync-enabled runtime it commits entity tables plus the durable outbox in one local SQLite transaction; in a local-only runtime (no sync env) it commits entity tables only.

Decision guide

• Team edits the rows directly, no application involved → spreadsheet.
• Application reads/writes, humans review in Sheets → Hikoutei.
• High write throughput, complex queries, or multi-region coordination → a conventional database; Hikoutei and spreadsheets are the wrong tool.

Related: Google Sheets as an ops UI.