16 November 2026, 7 min read
Org chart software in 2026: from a drawing to a live map of who does what
Org chart software used to mean a diagram. Today an org chart can be a live view of people data. Here is what to look for and what to avoid.
Org chart software has meant different things over time. At first it was a drawing tool: boxes, lines, a template and an export to PDF. Now the better version is a view of data you already keep, and it matters more than it looks. An org chart is the first thing a new joiner opens, the last thing anyone updates and, increasingly, the thing AI tools consult when they need to know who owns what.
This post covers how org charts got here, what separates a drawing from a live map, and what to check before you buy or build one.
A short history of the box-and-line diagram
The idea is old. According to Wikipedia, the engineer Daniel McCallum is credited with creating the first organisational charts of American business around 1854. The page also describes three common shapes: hierarchical, matrix and flat.
For most of the following century, a chart was a document someone produced on request. It was correct when printed. That limitation has not changed much for charts drawn in slide software today. Every reorganisation produces a new slide, and old slides keep circulating.
What goes wrong with a drawn chart
If you have ever inherited one, the problems are familiar.
- It drifts. Someone leaves, someone moves team, and the slide still shows the old shape.
- It is a picture. You cannot ask it a question. "Who are Priya's direct reports?" means reading boxes.
- It cannot hold detail. There is nowhere to put a start date, a location or a team page without cluttering the box.
- It is all or nothing. You either share the picture with everyone or with nobody.
- It has no owner. Updates happen when someone cares, which is usually just after something goes wrong.
None of these is a design flaw in the drawing. They come from the chart being a copy of the data rather than a view of it.
The live map: a view of records
The shift is simple to state. Instead of drawing the chart, you keep a record for each person, with a manager field, and the chart is generated from those records. Change someone's manager in one place and every chart that includes them changes.
That gives you things a drawing cannot:
- Queries. Ask for everyone under a given manager, everyone on a team or everyone in a location.
- Several views of one dataset. The same people records can appear as a table, a board by team, a timeline of start dates and an org chart.
- A page behind each box. Click a person and see their record, with responsibilities, projects and notes.
- History. You can see who reported to whom last quarter, because a change to a field is an event rather than a new slide.
Here is how the approaches compare.
| Capability | Slide or drawing | Dedicated chart tool | Chart as a view of people data |
|---|---|---|---|
| Update on a change | Redraw by hand | Edit boxes or import a file | Edit one field |
| Answer questions | Read it | Limited to the tool | Query the underlying records |
| Detail behind each person | Cramped boxes | Varies | Full record page |
| Other views of the same data | None | Varies | Table, board, timeline |
| Field-level privacy | None | Varies | Depends on the system |
| Used by AI tools | Not readable as data | Depends on export | Readable as records, under access rules |
"Varies" is honest. Org chart tools differ a lot, and a feature that is absent from one is a headline feature of another. Check the plan you would actually buy.
A chart that is a copy of the data starts going wrong the day it is saved. A chart that is a view of the data cannot.
Matrix and dotted lines
Most real organisations are not a clean tree. Wikipedia defines matrix management as a structure where some staff answer to two or more managers at once, typically a department head and a project lead, with the second link sometimes shown as a dotted line.
A tree-only chart handles this badly. The usual workaround is to put "(dotted to Marcus)" in a job title, which breaks every query and every filter.
A better model stores the two relationships separately. A person has a primary manager and, optionally, a project lead. The chart can draw one as a solid line and the other as dotted, and a query for "everyone Marcus leads" can include or exclude the dotted ones deliberately.
Privacy: the structure is not the same as the data
An org chart looks harmless. In practice the structure is usually fine to share widely, and the details behind it are not. Names, roles and managers are normal company knowledge. Salary, performance rating, compensation band, leave and personal contact details are not.
That means the software needs to separate the record from its fields. If permissions only work per page or per chart, you are forced to choose between a useful chart nobody can see the details behind and a detailed one that exposes things it should not.
Check three things:
- Can a single field have a different audience from the rest of the record?
- Does that rule apply in every view, including exports and API calls?
- If an administrator needs to read a restricted field, is the access logged and is the owner told?
The third point is easy to overlook. Someone always has to be able to fix a problem, but a silent override is a different thing from an auditable one.
Views inside views
Large organisations need more than one picture. A single chart of a thousand people is unreadable, so the useful version lets you open a team and see something different inside it.
The pattern we have found useful is nesting. An org chart shows the departments. Open a department and you can see its team as a board, its projects as a timeline or its members as a table, each a view of a related collection. The chart becomes a map you can zoom into, not a poster you scroll across.
For this to work, teams, people and projects need to be linked records rather than text in a box. That is why the choice of org chart software is really a choice about how you store people data.
Keeping it current
Software does not keep a chart accurate on its own. People do, and they need a reason. Three habits help.
Give the data an owner. Usually that is HR or a people operations lead, with each manager responsible for confirming their own reports. A field with no owner is a field nobody fixes.
Tie changes to events people already run. A new hire, a leaver and a team move all trigger some process. If updating the record is a step in that process, the chart follows for free. If it is a separate chore, it will lag.
Review what looks odd. A person with no manager, a manager with forty reports or a team with no members are all signs of a missing update. A table view sorted by manager makes these obvious in seconds, which is another argument for keeping the chart as one view among several.
What AI changes
AI tools ask org questions constantly. "Who approves this?" "Who is on call for billing?" "Which team owns the onboarding flow?" A tool that can read people and team records can answer these properly. A tool reading a slide deck can guess.
The condition is access. An assistant answering "who reports to Priya" is fine. One that can also quote her salary is not. If your org data feeds AI tools, make sure those tools see only what the person asking could see. We go into how in how to give AI agents your company's knowledge without leaking what they shouldn't see.
A checklist before you pick one
- Is the chart generated from records, or drawn?
- Can you import from a CSV or sync from a database you already have?
- How are matrix and dotted lines represented?
- Can fields have their own audience?
- Can you nest views, or do you just get one big picture?
- Who owns each record, and how are changes reviewed?
- What do readers cost? A chart that only a few people can afford to view defeats the purpose.
How Contextely approaches this
Contextely treats the org chart as one view of a People collection. People have typed properties and a manager relation, and the org chart view is generated from them alongside table, board, calendar and timeline views. Views can be nested, so a department can open into its own board or timeline.
Fields carry their own audience, so permissions can keep salaries private while the structure is shared with the company. You can bring people data in by CSV upload or a read-only Postgres connection, and the pricing page has the plans, with unlimited readers on each.
The short version
An org chart should be a view of data, not a drawing of it. Pick software that stores people and relationships as records, handles matrix lines honestly, keeps sensitive fields private even when the structure is public and lets you move between views of the same data.
Questions
What is org chart software?
Software that shows who reports to whom and how a company is organised. Older tools are drawing programs. Newer ones generate the chart from people data, so it stays correct when someone joins, leaves or changes manager.
Should an org chart be drawn by hand or generated?
Generated, if you can. A hand-drawn chart is accurate on the day it is made and drifts afterwards. A chart generated from records of people and their managers updates whenever the record changes.
How do I show matrix or dotted-line reporting?
Store the second reporting line as its own relation, separate from the primary manager, and let the chart display it as a different kind of line. Do not encode it in a job title.
Can everyone in the company see the whole org chart?
Usually the structure (names, roles, managers) can be open, while details on each person such as salary or performance notes stay restricted. Choose software that lets a field have its own audience.