Data Visualizer template walkthrough – TSV columns, examples, and a validation checklist
Data visualizer template walkthrough – tsv columns, examples, and a validation checklist Visio Data Visualizer is strict for a reason. It is not importing a “spreadsheet”. It is importing a data model. That means small mistakes that look harmless in Excel can break the import. The fix is to treat the template like code: exact headers, exact values, no extras. The biggest gotcha is separators: Data Visualizer expects TSV (tab separated values). If the file is CSV, the importer often reads the whole line as 1 field and fails. The columns that matter most: 1. Process Step ID Must be unique. Must be stable. Treat it like a primary key. Many teams use 010,020,030 so inserts are easy later. 2. Next Step ID This is where the flow lives. Every value must point to an existing Step ID. Multiple next steps go in 1 cell as comma separated IDs with no spaces: 030,040 3. Shape Type Must be one of the allowed values (Start, End, Process, Decision, etc.). Typos here create hard-to-debug failures. 4. Function If a swimlane diagram is expected, Function cannot be blank. Also: keep lane names consistent. “HR” and “Human Resources” are different lanes. 5. Phase (optional) Use it only if phases are meaningful. Keep values consistent. 6. Connector Label (optional) Leave it blank unless the arrow label is truly needed. If a decision branches to 2 paths, the same label may show on both if it is filled. Import failure usually means 1 of these: • separators are wrong (CSV instead of TSV) • a header is off by 1 character • a blank line exists • a Step ID is duplicated or missing • a Next Step points to an ID that does not exist • a cell contains a hidden line break A simple validation checklist before import: • TSV, not CSV • No blank lines • Header row matches the template exactly • No duplicate Step IDs • No Next Step IDs that do not exist • End shapes have blank Next Step ID • Comma lists have no spaces • No leading/trailing spaces in cells • No hidden characters from copy/paste The best confidence test: Import a 10-step dataset first. Then change 1 Function value (lane name) and re-import. If it round trips cleanly, scaling becomes mechanical, not emotional. A clean template plus 1 known-good example dataset removes most guesswork. That is the fastest way to get to a repeatable process diagram workflow. It also makes team reviews easier. Every time. #Visio #DataVisualizer #ProcessMapping #Excel #BusinessAnalysis #Operations If you want the validation checklist as a copy/paste pre-flight step, comment “template” and it will be sent. process improvement, process mapping, operations, business analysis, workflow, visio, swimlane, automation, lean, standard work
