Swimlanes as Data: Normalize Function Data (Post 2)
Swimlanes as data: normalize function data (post 2) Swimlane diagrams do not usually “break” because the process changes. They break because lane names drift. It starts small: • “Operations” becomes “Ops” • “Customer Success” becomes “CS” • someone adds a trailing space • a contractor types “Finance / AP” for the same lane the team calls “Accounts Payable” Then Visio Data Visualizer imports the dataset and suddenly: • 2 lanes appear where there should be 1 • steps move into the wrong lane • handoff counts become meaningless • the diagram looks “wrong” and people stop trusting it This is why the Function column (the swimlane owner field) needs normalization. Normalization is not theory. It is simple hygiene. A reliable workflow: 1. Extract the unique Function values from the dataset 2. Choose 1 canonical name for each lane 3. Build a mapping table (alias → canonical) 4. Replace aliases in the dataset using a lookup 5. Lock it down with data validation (drop-down list) 6. Add short definitions so classification stays consistent 7. Repeat the same method for every future lens dataset Practical Excel moves that make this fast: • TRIM to remove invisible spaces • CLEAN to remove non-printing characters • XLOOKUP to map aliases to canonical lane names • Data Validation to prevent new drift Example: Ops | Operations | Operations Team | Ops (with a trailing space) All become: Operations Now every count, pivot, and diagram import becomes predictable. A decision that matters: What does Function represent? It can be: • Department lanes (good for accountability and handoffs) • Role lanes (good for staffing and approvals) • System lanes (good for automation and integration) The mistake is mixing these in one dataset. The clean approach: Keep 1 canonical dataset, then create derived lens datasets. Same Step IDs. Same Next Step IDs. Same flow. Only Function (and optionally Phase) changes to create the view needed for the decision. The payoff is immediate. Once Function is normalized, the dataset becomes a reliable system of record: • lane changes are intentional, not accidental • re-orgs are handled with a mapping update, not diagram surgery • handoffs can be counted (Function changes across connected steps) A governance rule that saves weeks: Keep a “Lane Dictionary” sheet. Lane (canonical) | Aliases | Definition | Owner | Last reviewed date Quick test: Import 20 steps, then deliberately change 1 Function value and re-import. If the lane moves exactly as expected, normalization is working. Normalize Function first. Everything else gets easier after that. #Visio #SwimlaneDiagram #ProcessMapping #DataQuality #BusinessAnalysis #DataVisualizer #Operations process improvement, process mapping, operations, business analysis, workflow, visio, swimlane, automation, lean, standard work
