Skip to main content
Goal: let a submitter enter a value that no admin has pre-approved, without losing consistency. A Text field with Suggest from past entries turned on shows an autocomplete of the values already saved on approved records for that field, so people converge on the same spellings while staying free to type something new.

When to reach for this

A dropdown field backed by a picklist is a closed list: submitters pick from what the published model offers, and only an admin can add options. That is the right shape for a vocabulary you want reviewed before it changes. Some values churn too fast for that loop. Partner names, event titles, one-off placements: by the time a ticket reaches an admin, the campaign has shipped. For those, keep the field as Text and turn on Suggest from past entries. You trade the hard guarantee of a closed list for suggestions that make the consistent choice the easy one.
Earlier versions of Terminus Hub had an “Allow new values” toggle on static picklists that let submitters propose options from inside the form. That capability was removed: a picklist’s option set is now fixed at publish time, and only an admin changes it. This recipe is the replacement. If a dropdown in your model relied on submitter-added values, convert it to a Text field as shown here; nothing converts automatically, because a closed list and free text with suggestions are different shapes.

Prerequisites

  • A governance model you can edit and permission to publish it.
  • Some approved records, if you want to see suggestions right away. Suggestions come from values already saved on approved records, so a brand-new field starts with none.

Steps

1

Create or convert the field

In Fields, create a new field (or open the one to convert). Name: partner_name. Type: Text. If you are converting an existing dropdown, change its type to Text; the field keeps its name and placement in taxonomies.
2

Turn on Suggest from past entries

In the field’s settings, turn on Suggest from past entries. It is off by default. Save the field.
3

Add guardrails with text validation

Suggestions nudge, they do not enforce. Use the Text field’s own settings to rule out the messiest variants: set Transform case to lowercase so Acme and acme collapse into one value, and Transform spaces to replace_underscore if the value ends up in URLs. Add allowed character rules if the value feeds a strict downstream system.
4

Publish the governance model

Publish so the change reaches submitters. Field setting changes, like every model edit, take effect when published.
5

Fill a submission and watch the suggestions

Start a submission on a taxonomy that includes the field. As you type in partner_name, previously approved values that match appear as suggestions. Pick one to reuse it, or keep typing to enter something new. Either way the form accepts the value.

How suggestions behave

  • They come from approved records. A value starts appearing as a suggestion once a record carrying it is approved. Draft submissions do not feed the list, so a typo in an unapproved draft never spreads.
  • They are account-wide. Suggestions are drawn from the field’s values across your whole account, not just the current workspace. Everyone is nudged toward one shared vocabulary, which is the point.
  • They are non-binding. A submitter can always ignore the suggestions and type a new value. If you need hard enforcement, that is a dropdown with a picklist, not this.

Verify

  • With at least one approved record carrying partner_name: acme, start a new submission and type ac in the field. acme appears as a suggestion.
  • Type a brand-new value instead and save. The row validates; free text is always accepted.
  • Approve that submission, then start another one. The new value now shows up as a suggestion too.

Gotchas

  • Reviewers are the quality gate. Nothing stops a submitter from typing a near-duplicate (Acme Corp next to acme). Transforms and character rules catch the mechanical variants; reviewers should still scan new values before approving. See review and approve submissions.
  • Suggestions lag approval. A value entered this morning does not suggest this afternoon unless its submission was approved in between.
  • Converting a dropdown loses the closed-list guarantee. After conversion, past records keep their values, but new submissions accept any text the validation settings allow. Convert only fields where that trade is deliberate.
  • This is a per-field setting. Each Text field opts in on its own; there is no account-wide switch.