This is the kind of request that reveals how much the spreadsheet world still asks us to work around its limitations. A user needs to visualize water meter error boundaries using stacked columns as permissible limits and scatter points as actual computed errors. The idea is elegant: the columns define zones of tolerance, and the points show where real-world measurements fall within them. But getting a standard tool to execute this combo chart requires explaining the concept, hoping someone has a workaround, and then manually wrestling axis scales and series overlaps until the visual finally matches the logic. That is not data analysis. That is puzzle-solving.
We believe this should be simpler. The user is not asking for anything exotic. They have a clear mental model: a set of stacked boundaries, -5% to 5%, -2% to 2%, -1% to 1%, -0.5% to 0.5%, and a set of points that must never exceed those boundaries. The columns are not decorative. They are the rules of the system, the maximum permissible error. The points are the evidence. The chart is a compliance check, a metrological curve that answers a concrete question: do my measurements stay inside the allowed range? That is a fundamental job of a spreadsheet, yet the process of building it feels like a detour.
What this tells us is that too many spreadsheet tools still treat charts as afterthoughts, static decorations rather than dynamic expressions of your data logic. The user should be able to say "stack these ranges as boundaries" and "plot these points against them" without needing to figure out which series type to assign, how to manage axis alignment, or how to force a column to span negative and positive values symmetrically. The tool should interpret intent. Instead, the tool asks the user to translate their thinking into workarounds.
For anyone building similar visualizations, error bands, tolerance zones, target ranges, the lesson here is that you are not alone in this friction. The solution exists, but it should not require a forum post. An AI-native spreadsheet would recognize the pattern: you have boundary data and observation data, and you want to overlay them. It would suggest the chart type, set the axes, and let you adjust the styling. The technology exists to make this a one-click operation. Until then, know that the request itself is sound, the logic is clear, and the frustration is a sign that the tools have not caught up to how people actually think about their data. Build the chart anyway. The insight it gives you is worth the effort.