I have a theory that product managers often inversely estimate the difficulty of feature requests. I call it the "It Shouldn't Be Too Hard" fallacy.

Many times I've had a PM cautiously explain a needed feature and almost apologize for how hard it will be, while I know it will actually be simple to implement. I'll later experience the exact opposite. A manager casually commits us to something that seems simple, but is, in fact, exceedingly difficult to accomplish.

I was working on research for my web apis series and was keeping some notes in a spreadsheet. I had an idea for something I wanted to do which I don't think exists. It occurred to me, this was a perfect example of me being in the PM's shoes, committing this fallacy.

I was using Google Sheets and had an idea that I wanted to split a cell in half while keeping both of these cells in the same column. Surely this functionality must exist. I mean, merging cells is a thing. Why can't I just do the opposite. "It Shouldn't Be Too Hard", right?

I gave it some thought and the idea of this feature started snowballing in complexity.

  • How will you address these two sub-cells in formulas?
  • Will conditional formatting apply to each sub-cell or would it affect the whole cell?
  • What happens if you split a splitted cell?
  • What if you split one of splitted cells, how do you address those?
  • How does resizing affect the sub-cells?
  • What about backwards compatibility?
  • What about performance?

These are all questions that can have answers, but they're not obvious and will definitely upset some users.

A simpler solution occurred to me. I really only wanted this feature for formatting. I could accomplish the same thing using a tab.