Measure custom-field adoption and coverage

Analytics → Custom fields answers one question: is the structure we designed actually being filled in?

Custom-field analytics showing coverage, label use, and field activity — Coverage reveals whether a field is useful in daily work or has become empty schema that nobody maintains.

What follows the time range, and what doesn’t

The view says this up front, because it matters:

Ignores the time range: Field Coverage, Chats tagged, Untagged, Top column, always your whole inbox as of now. Follows the time range: Changes, Top taggers.

So a “0 changes” week doesn’t mean coverage dropped; it means nobody tagged anything.

Headline stats

Stat	Meaning
Total Fields	Number of custom columns in the current view
Average Coverage	Average percent of chats with a value set, across all fields
High Coverage	Fields filled in at least 75% of chats
Needs Attention	Fields filled in fewer than 25% of chats
Members	Workspace members who contributed field changes

Needs Attention is the one to act on. A field under 25% is either unnecessary or unenforced, decide which and either delete it or make someone responsible for it.

Field Coverage

Per column: how often it’s filled, and its value distribution.

“{filled} of {total} chats” with a tooltip explaining the count.
Multi-select columns get a longer tooltip: because a chat can hold several values, the per-value counts overlap and sum to more than the number of tagged chats. That’s expected, not a bug.
Top values per field, expandable with Show {n} more.

Label usage

A donut plus supporting counts:

Chats tagged / Untagged, “{pct}% of chats tagged”
Top column, the field applied to the most chats
Changes - custom-field values added, edited or cleared this period
{used} of {total} columns carry data, how much of your schema is dead weight
Top taggers - per member: {created} new · {edits} edits · {n} cleared, with Last active

Field Update Activity

A heatmap of when custom field values are updated. Darker cells mean more updates; click a cell for the detail. Useful for spotting whether tagging happens as work flows or in a guilty Friday-afternoon batch.

Top Chats by Field Completion

Chats with the most fields filled, by completion rate. Duplicates across members are deduped.

Recent Changes

The last n field value updates - “Set to: {value}”, “Removed: {value}”, with the chat, the field and who did it.

In workspace mode this feed widens beyond field edits to include Telegram and workspace actions: messages sent/edited/deleted, reactions, forwards, media, member joins/leaves, role changes, account shared/unshared/reconnected, broadcasts sent, comments, files, column create/delete, reminders, ticket events and API key events.

Member Activity

Who’s updating custom fields - counts of custom-field changes, Telegram actions and workspace actions per member.

Empty states

Message	Meaning
No Workspace Custom Fields	No workspace-scoped columns exist yet, create them in the Chats view
No Custom Fields	You haven’t created any personal custom fields
Select an account to view custom fields analytics	Pick an account in the filter

A practical routine

Once a month:

Sort fields by coverage. Anything under 25% → delete it or assign an owner.
Check {used} of {total} columns carry data. If it’s 4 of 11, your table is mostly noise.
Look at Top taggers. If one person does 80% of the tagging, the process isn’t a process, it’s that person.
Build an Untagged view from the No custom fields filter and work it down.