Custom list fields
Capture what this process needs, with field meanings your team shares.
Create lists with supported text, date, amount and other fields so a custom register reflects the information your team actually maintains.
A column label is only useful when its meaning is agreed.
A general spreadsheet may grow columns without explaining what belongs in them. Custom list fields give the process a structure inside Tsumira. The team still decides which information is required, how it is maintained and what a value means. A field called approved, amount or due date does not by itself implement a formal approval, accounting rule or contractual obligation.
Choose the data the record needs.
Use supported field types such as text, dates and amounts to describe the work. Separate the record's main identity from explanatory detail. A material checklist, for example, needs an identifiable item and useful preparation information rather than every conceivable field. Begin with a small dataset so colleagues can point out ambiguous labels before the register grows.
- Supported field types for custom registers
Distinguish a custom register from a standard business record.
Custom lists can organize work that does not fit a standard module. They should not silently become a second source of truth for customers, time or invoices. Decide which information belongs in its native business record and which is specific to the custom process. An amount in a free list is not automatically an invoice line or tax-aware financial entry.
- A clear boundary between custom and native records
Review the fields through actual work.
A field is worth keeping when it supports a decision or required action. Inspect the records through a useful view and ask another colleague to update a sample. If people interpret a field differently, clarify the process before collecting more data. AI can help discuss a structure, but proposed changes still require review and user permissions apply.
- A small sample before broad data collection
A concrete workflow
Define the row
Explain what one record represents and why it is maintained.
Choose fields
Add supported fields with unambiguous labels and meanings.
Test updates
Ask a colleague to maintain a real sample without verbal coaching.
Refine
Clarify ambiguity before expanding the list or importing more data.
An illustrative working example
Illustrative scenario: a launch-material register
A team records material needed for a launch. Text explains the item and a date identifies when the team expects it. Before adding further detail, another colleague updates several records and points out whether expected date means requested, promised or received. The team agrees the meaning and keeps customer identity in CRM rather than recreating a parallel customer master in the list.
Questions about this topic
Can custom fields replace financial records?
No. A custom amount field does not establish invoice, payment, currency or tax behavior. Keep authoritative financial processing in its proper module.
Can we use text and dates?
Supported fields include text, dates and amounts. Define the meaning of each field with the people who will maintain the records.
Should we import every spreadsheet column?
Start with the information the workflow actually needs. Test a small sample and remove ambiguity before expanding the register.
Can AI decide the structure for us?
Use it to explore options. People still review proposed fields and actions and agree the business meaning before execution.
Explore connections around this workflow.
Ideas for intended connections. Check the actual availability, permissions and supported data for each app before using a scenario.
Microsoft Excel
Application concept: Compare Excel columns with explicit text, date and amount fields and review units before creating the schema.
Google Sheets
Application concept: Identify inconsistent Google Sheet values before choosing field names and required information in the List.
Notion
Application concept: Review the meaning of Notion properties before mapping them to the supported field types rather than copying labels blindly.