From Excel to an org chart in ten minutes: how to prepare your staff file
A practical guide for HR: which columns your spreadsheet needs, the five mistakes that break an import, and what to do when your HR system exports no Manager column at all.
Every HR team has the file. It is called something like "Staff_list_FINAL_v3.xlsx". It has a few hundred rows, a title line on top, merged cells somewhere in the middle, and it is the most accurate picture of the company that exists anywhere.
The good news: that file is already 90% of an org chart. The structure is in there. It just has not been drawn yet.
This guide explains how to get from the spreadsheet to a working chart quickly, which columns matter, and how to fix the problems that most often get in the way. The examples use orgcanvas, but the principles apply to any tool that builds a structure from data.
The one rule: one row, one position
Before anything else, check this. Each row of the file should be one position. If the position is filled, the row holds the person's name. If it is vacant, the name cell stays empty and the row is still there.
This matters more than it looks. A file that lists only people hides every open seat, and a chart built from it will show a company with no vacancies, which is never true. A row for "Credit Analyst, vacant" tells the chart where you are hiring and what the budget assumes.
The columns you need
The two you cannot skip
- Full name. One column, or separate Last name / First name / Patronymic columns: they are joined automatically.
- Position. The job title as people use it: "Chief Accountant", "Branch Manager", "Teller".
With only these two you will get cards, but no lines between them. Every position will stand at the top.
The one that draws the lines
- Manager. Who each position reports to.
This can be the manager's name or the manager's Employee ID. Use the ID if you can. Names repeat: a company of five hundred people almost always has two Armen Sargsyans, and a manager column with names cannot tell them apart. With an Employee ID column, the lines are drawn without a single question.
The ones that make the chart useful
- Department, as one column or several for each level: Division, Department, Unit, Section. Up to five levels are understood.
- Employee ID. Needed for the manager link above, and later for updating the chart from next month's file.
- Grade, salary and currency, if you plan to look at pay structure.
- Hire date, employment type, location or branch, and contract end date, if you have them.
Column names can be in English, Russian or Armenian ("Position", "Должность", "Պաշտոն"). They are recognised automatically, and you can correct the match by hand before anything is imported.
Five mistakes that break an import
1. Titles and notes above the header
Many exports start with a company name, a date and an empty line before the real header. That is fine: title lines are skipped. What is not fine is a second header halfway down the sheet, or a "Total" row at the bottom. Delete those.
2. Merged cells
Merged cells look tidy and read badly. A department name merged across twenty rows usually arrives as one filled cell and nineteen empty ones. Unmerge and fill the value on every row.
3. A manager who is not in the file
If row 47 reports to "Gevorgyan" and there is no Gevorgyan in the file, the line cannot be drawn. This usually means the manager left, or is listed under a different spelling. The review step shows each case and offers a fix: point the row to the top, or skip it.
4. Loops
A reports to B, B reports to A. It happens more often than you would think, typically after a manual edit. A loop cannot be drawn as a tree, so it is flagged and must be fixed before the import.
5. One person, two rows with different spellings
"Petrosyan Anna" and "Anna Petrosyan" are the same person to you and two people to a computer. An Employee ID column removes this problem entirely.
When your file has no Manager column
This is the most common situation with exports from HR and payroll systems. They know who works in which department, but not who reports to whom. Without the lines, the whole company would stand in one long row.
There is a practical way out: build the lines from the departments. In every department the head is found from the title (Head, Director, Manager, Начальник, Ղեկավար), or, if no title stands out, from the highest grade. A deputy or an assistant is never taken as the head. People report to the head of their department, heads report to the head of the department above, and the top heads report to the CEO or General Director.
You see the result before anything is saved: who was taken as the head of each department, and how sure that is. Where the titles do not make it clear, you choose the head yourself, or, if the AI assistant is turned on, ask it to suggest one. Only the department names, the job titles and the grade levels are sent to it, never names, IDs or salaries.
When branches or stores flood the top level
Banks and retail chains often export every branch as its own department, next to Finance and HR. Forty branches at the top level make a chart nobody can read. When the import sees three or more departments named like branches or stores, it offers to put them under one parent, "Branch network", and to record each as a branch. One click, and the chart has a shape again.
Nothing is saved until you say so
The import is a review, not a leap. You see how many positions, levels and departments will be created, every error with its fix, and every warning. Nothing is written until you press Import, and the result goes into the official structure of the company, where it can be undone.
Next month: update, do not rebuild
The first import is the hard part. After that, the same file from the HR system becomes an update. Upload next month's export, and you see what changed: new hires, people who left, transfers, new titles. Tick what to apply. The structure stays alive without anybody redrawing it.
A checklist before you upload
- One row per position; vacancies have an empty name.
- A Position column with real titles.
- A Manager column, ideally by Employee ID. If you have none, departments will do.
- Department columns, one per level.
- No merged cells, no totals, no second header.
- Employee IDs unique and filled for every person.
That is it. Ten minutes of tidying the file usually saves days of drawing boxes by hand, and you end up with a structure that can be updated instead of redrawn.
Want to try it on a ready-made company first? Open the demo of a 1,000-person bank, no sign-up needed.