This is a frustrating problem, and we understand why it feels like a dead end. The user has built a dynamic chart that depends on query data, and the moment a null value appears, the whole visualization resets. That is not a user error. It is a design limitation in how traditional charting engines handle gaps in data. The instinct to avoid restructuring the entire data model is correct, modifying source queries just to make a chart behave is a workaround, not a solution. The user should not have to choose between clean data for other processes and a chart that actually works.
What this reveals is a deeper issue with legacy spreadsheet tools. They treat null values as failures rather than information. A null is not nothing. It is a valid state that means "no data at this point." But most chart engines see a blank cell and panic, breaking the axis or resetting the series. The user already tried the "Properties follow data points" setting with no luck, which tells us the tool simply lacks the logic to interpret nulls as intentional gaps. The workaround of filling zeros or placeholder values introduces noise into the model, distorting trends and creating more work downstream.
The practical lesson here is that your data model should drive the chart, not the other way around. If you are spending time adjusting source queries just to keep a graph stable, you are fighting the tool instead of letting it work for you. The better approach is to use a system that understands null values natively, one that can interpolate, skip, or display gaps without breaking the visualization. AI-native spreadsheets are starting to handle this intelligently, treating data models as living structures rather than rigid grids. They can recognize when a null belongs in the chart and when it should be ignored, all without requiring you to rewrite your queries.
Do not waste energy patching around a chart that insists on breaking. Instead, look for a tool that respects your data as it is. If your current spreadsheet cannot handle a null value without resetting, that is not your problem to solve. It is the tool's limitation. The right solution will let you keep your data model clean and your charts smooth, no workarounds required.