Why a spreadsheet drifts and a purpose built log does not
Why a spreadsheet drifts and a purpose built log does not
Blog Article
Almost everyone starts with a spreadsheet, and for the first month it is genuinely fine. The drift starts quietly and is usually invisible until you try to analyze what you have.
The first drift is column meaning. Column D was "units" for six weeks, then you started occasionally putting mL in it because that was what you had handy, and now column D contains two incompatible quantities with no marker separating them. Nothing in a spreadsheet objects to this. A field with a declared type does.
The second is the header that stopped being true. You wrote the concentration in a merged cell at the top: 2.5 mg/mL. Then you opened a vial reconstituted differently and kept typing rows underneath the old header. Every row after that point is mislabeled, and there is no line on the sheet marking where the change happened. This is the most damaging spreadsheet failure because it is silent and it corrupts a range of rows rather than one.
The third is formula rot. You set up a column computing mg from units and concentration. Later you paste a block of rows and the formula does not extend, or it extends but references the wrong concentration cell. Some rows are computed, some are typed, and they look identical. A computed column that is only sometimes computed is worse than no computed column, because it invites trust it has not earned.
The fourth is date formats. Manual entry produces a mix of 03/04, 2026-04-03, "Apr 3" and the spreadsheet's own guess at what you meant, and sorting on that column returns an order that is not chronological. You usually discover this at exactly the moment you need the log to be right.
None of this is a spreadsheet defect. Spreadsheets are deliberately unconstrained, which is why they are useful for exploration and unreliable for records maintained by a tired person at 7am. The value of a purpose-built log is entirely in the constraints: a concentration field that belongs to the vial and cannot go stale, a dose that is computed the same way every time, dates that are dates. That is what Peptide Tracker & Calculator is doing that a blank grid is not.
If you are staying on a spreadsheet, mitigate the three worst drifts. Put concentration in a per-row column and fill it every time. Use data validation on the date column. Never merge cells. Those three changes remove most of the corruption, though they do not remove the possibility of it. If you are switching, do it at a vial boundary rather than mid-vial, and carry the old rows over as-is rather than reinterpreting them. Importing an ambiguous history into a structured dose log and marking the ambiguous rows is better than cleaning them from memory.
Report this page