Data fields

The Org Chart is driven entirely by the fields you drag into its wells. This page explains every field: what it does, whether it's required, how many columns it accepts, and how to shape your data for it.

Field summary

Field	Required	Accepts	Max columns	What it does
Entity ID	Yes	Column	1	The unique ID of each card.
Parent ID	Recommended	Column	1	The ID of the card's parent.
Information Fields	No	Columns or measures	3	Text shown on the card. The first is the title.
KPI values	No	Measure	1	The card's main number.
KPI Target	No	Measure	1	The benchmark the KPI is compared against.
KPI Period	No	Column	1	Orders the KPI over time; draws the sparkline.
Image (Base-64)	No	Column	1	A Base64 image on the card.
Card URL	No	Column	1	An http(s) link opened by the card's link icon.
Accent Color	No	Column	1	Categorical field that colours the card's accent border.
Swimlane	No	Column	1	Categorical field that groups cards into labelled lanes.
Tooltips	No	Columns or measures	Unlimited	Values shown on hover only.

Row limit

The visual receives up to 30,000 rows from Power BI. Beyond that, Power BI stops sending rows and the chart is drawn from the first 30,000 it received. For very large hierarchies, filter to a division or a level range before the visual sees the data.

One row per card

The Org Chart deduplicates on Entity ID. How it picks the row for a card depends on whether KPI Period is bound:

• Without KPI Period — one row per Entity ID is expected. If several rows share an ID, the first row encountered wins and the rest are ignored.
• With KPI Period — one row per Entity ID per period is expected. The card is drawn from the latest period's row, and the whole series becomes the card's sparkline.

warning

If you bind KPI Period but your data has only one row per card, the sparkline won't draw (it needs at least two points) and change vs. previous has nothing to compare against. That's expected — bind a target instead, or extend the data over time.

Entity ID

The only field the visual truly requires. Each value becomes one card.

• Text or whole-number columns both work. Numbers are converted to text internally, so 1 and "1" match.
• Blank, null, NaN and infinite values are dropped — those rows produce no card.
• Leading and trailing spaces are trimmed before matching, so "A1 " and "A1" are the same card.
• If nothing else is bound, the Entity ID is shown as the card's title.

Parent ID

Points each card at its parent. Values must match values in Entity ID.

The visual is deliberately forgiving about imperfect hierarchies:

Situation	What happens
Parent ID is blank	The card is top-level.
Parent ID matches no Entity ID	The card is promoted to top-level (an "orphan").
Parent ID equals the card's own Entity ID	The self-reference is dropped; the card becomes top-level.
A longer loop (A → B → C → A)	The loop is detected and broken by promoting one of its cards to top-level.
No card has a blank or unmatched parent	The first card is promoted to top-level so the chart can still draw.
Several top-level cards	They are drawn side by side under an invisible shared root.

tip

Because unmatched parents become roots rather than errors, you can filter the chart to a sub-branch (say, one division) and it will render cleanly with that division's head as the top card.

Information Fields

Up to three columns or measures displayed on the card, in the order you drop them:

1. First field → the card's title, in bold.
2. Second and third fields → smaller lines underneath.

Values are formatted using the column's model format string, and truncated with an ellipsis if they don't fit the card's width. If you leave Information Fields empty, the card shows the Entity ID instead.

tip

Need more than three lines? Put the extras in Tooltips — they show on hover without making every card taller.

KPI values

The measure shown as the card's main number, on the left of the KPI row.

• Must be numeric. Text values are ignored and the KPI row simply doesn't appear.
• The measure's model format string is always respected — currency symbols, percentages, decimals.
• The measure name is shown as a prefix, e.g. Revenue: 1.24M.
• Display units (K/M/bn) and decimal places for the pill are set in Format → KPI Configuration. The tooltip always shows the full, unrounded value.
• Blank or non-numeric results leave the card without a KPI row — the rest of the card still renders.

See KPIs, targets & trends for the comparison logic.

KPI Target

A second measure used as the benchmark for the KPI.

• When the target is greater than zero, the card shows attainment: KPI ÷ target, e.g. 112.0%.
• When the target is zero, blank, negative or non-numeric, it can't produce a meaningful attainment; the card falls back to comparing against the previous period instead (if KPI Period is bound).
• Targets can be a fixed measure, a DAX calculation, or a what-if parameter — anything that resolves to a number per card.

KPI Period

A date or period column that puts the KPI on a timeline. Bind it and every card gains a sparkline.

• Expect one row per card per period.
• The card's displayed KPI comes from the latest period.
• Rows with a blank period are skipped in the sparkline (and can't act as the "previous" comparison point).
• Sorting: dates and numbers sort naturally. Text labels sort alphabetically, so "Q1 2025", "Q2 2025" … works but "April", "May" does not. For text period labels, set Sort by column on the dimension in your model.

tip

A date column from a proper date table is the safest choice — it sorts correctly, formats nicely in the tooltip, and lets report filters slice the trend.

Image (Base-64)

A column containing a Base64 data URI for the card's image.

• Must start with data:image/ and one of png, jpeg, jpg, gif, webp, svg+xml, followed by ;base64,.
• External http(s) links are not accepted — this is a requirement of Microsoft certification.
• Maximum size is roughly 500 KB per image once decoded. Larger values are ignored and the card renders without an image.
• Invalid values are silently skipped, so a partially-populated column is fine.

See Images on cards for how to produce the Base64 column in Power Query.

Card URL

An absolute http(s) address. When present, a small link icon appears in the top-right of the card; clicking it opens the address in a new tab.

• Only http:// and https:// are accepted. Relative paths, mailto:, and any other scheme are ignored.
• Toggle the icon off entirely with Format → Card Configuration → Show Card URL Link.
• Keyboard equivalent: focus the card and press Alt+Enter.

Accent Color

A categorical column (department, region, status…). Each distinct value gets a colour, drawn as a border on the card.

• Colours are picked per group under Format → Accent Color → Colors — one swatch per value found in the data.
• Border width (1–10 px) and position (all sides, left, right, top, bottom) are set in the same card.
• Up to 30 distinct groups get their own picker. Beyond 30, extra groups are still coloured, using the report theme's palette, but have no swatch in the pane.

See Accent colours.

Swimlane

A categorical column that groups cards into labelled lanes. Lanes are drawn as alternating shaded bands with a title.

• Lane direction follows the layout orientation: vertical columns in Top-Down, horizontal rows in Left-to-Right.
• Rows with a blank swimlane value are placed in a lane called Default.
• Cards are ordered by swimlane value within each level.

See Swimlanes.

Tooltips

Any number of extra columns or measures, shown only on hover — never on the card itself.

The tooltip is assembled in a fixed order:

1. The Entity ID.
2. Every field you put in Tooltips, in order.
3. The KPI value (full precision, no display units).
4. The target — or the previous period's value, labelled with that period, e.g. Previous (May 2026).
5. The comparison result, labelled % of target or Variance.

Turn tooltips off entirely with Format → Tooltips. Report-page tooltips are supported too — see Report page tooltips.

Next steps

• Hierarchy & layout — orientation, levels, swimlanes and card sizing.
• KPIs, targets & trends — comparison modes and thresholds.
• Use cases — worked examples with the data shape for each.