Online Tool Store
· 4 min read
How to Turn a Dated CSV Into a Timeline
Manesh Jayawardhana
CIO & Co-founder
The project plan lives in a spreadsheet with 58 phases, each with a start and end date. Rendering it as a timeline should be a format conversion.
Two of those rows have the end date before the start. One has a date typed as text. Nothing in the spreadsheet complained, and a timeline renderer will draw two bars going backwards and silently drop the third.
Validate before converting
The conversion itself is trivial. What matters is what happens to the rows that do not convert cleanly, and there are three categories.
Reversed dates — end before start. Almost always a typing error or a copy-paste that swapped columns. A timeline renderer given a negative duration does something arbitrary, and which arbitrary thing depends on the library.
Unparseable values — a date typed as text, a placeholder like “TBC”, an empty cell where a date was expected.
Ambiguous formats — 03/04/2026, which is two different dates depending on convention.
Flagging all three is more useful than converting the ones that work and dropping the rest, because the rows that fail are frequently the interesting ones. A phase with no end date is often the phase nobody has planned.
| Row problem | If not flagged |
|---|---|
| End before start | Bar renders backwards or vanishes |
| Unparseable date | Row silently dropped |
| Ambiguous format | Wrong date, looks correct |
| No end date | Renders as a point, or breaks |
Ambiguity should stop, not guess
03/04/2026 is 3 April in most of the world and 4 March in the United States. A parser that picks a convention and applies it silently is right roughly half the time, and wrong invisibly the rest.
On a project timeline that means phases shifted by weeks with nothing indicating a problem. The chart looks correct, the dates look plausible, and the plan is wrong.
Dates above the 12th disambiguate themselves — 25/03/2026 can only be day-first. A file containing only early-month dates has no signal at all, and that is exactly the case where guessing is most dangerous.
The correct behaviour is to require the format when the data could be either.
Point events and open-ended spans
Not every row is a span.
A milestone has one date. Most timeline renderers handle a start with no end as a point marker, which is usually right.
An ongoing phase has a start and no end because it has not finished. Rendering it as a point is wrong; rendering it as a span to today, or to the plan’s end, is a choice that should be explicit.
A row with an end and no start is nearly always an error rather than a meaningful case.
Deciding what each means before conversion avoids a timeline where three different situations all render as the same marker.
Decide what a duration means
An ambiguity in date-only data that changes every span by a day.
A phase running from 2026-01-05 to 2026-01-09 is either four days or five, depending on whether the end date is inclusive.
Project plans usually mean inclusive — the phase happens on the 9th. Timestamp ranges in software usually mean exclusive — the span ends at the start of the 9th.
A timeline renderer will pick one, and a five-day phase drawn as four is a visible error on a chart where phases are meant to abut.
Stating the convention and, better, converting date-only values to explicit start and end instants before rendering removes the ambiguity entirely.
Common mistakes to avoid
- Converting without validating, so bad rows surface as rendering oddities.
- Letting a parser guess an ambiguous date format.
- Treating every missing end date as a milestone.
- Uploading a project plan containing commercial dates to a hosted converter.
- Fixing the errors in the converted output rather than in the source spreadsheet, so the next export repeats them.
How to do it with CSV to Timeline Converter
The CSV to Timeline Converter validates as it converts, in the browser.
- Paste the CSV with its header row.
- Name the title, start and end columns.
- Read the flagged rows — reversed dates and unparseable values are data errors, not conversion failures.
- Fix them in the source so the next export is clean.
Other data tools that keep files local are in the tools directory.
Frequently asked questions
What date formats are handled?
ISO and common regional formats where they are unambiguous. Genuinely ambiguous dates are flagged rather than guessed, because guessing produces a plausible timeline that is quietly wrong.
What happens to rows with no end date?
They are treated as point events, which most renderers handle. A row with an end and no start is flagged, since that is nearly always an error.
Is my data uploaded?
No. Conversion runs in the browser, so project plans and anything commercially sensitive stay on your device.
Final thought
Read the flagged rows before rendering anything. A timeline with two reversed dates looks fine at a glance and is wrong in exactly the places nobody checks.