A predicted period date is a calculation, not an observation of what your body is doing today. A tracker can calculate correctly from the dates you enter and still miss the next period. The useful question is: which dates and assumptions produced this estimate?
What Bloom’s web tracker actually calculates
Bloom’s web period tracker adds the cycle length you select to the first day of your last period. It starts with a 28-day selection; that default is not a personal assessment. The bleeding-length setting controls the displayed bleeding days, rather than the date of the next period. The tool also estimates ovulation by subtracting 14 days from the predicted next period. That is a calendar assumption, not confirmation of ovulation. These details describe the website calculator, not a clinical validation or every feature of the separate app.
A worked example: one date, three estimates
Imagine your last period began on 4 October. Selecting a 28-day cycle gives 1 November as the next estimated start. Selecting 31 days moves it to 4 November. If your next period actually begins on 6 November, the interval was 33 days. Record 6 November as the observed start; do not move it back to make the prediction look right. If earlier cycles were 27, 29 and 34 days, their average is 30 days. That average describes those three intervals; it does not promise that the next cycle will last 30 days. These are fictional arithmetic examples, not normal-range thresholds or personal forecasts.
Why two trackers can disagree
Two tools can start from different inputs: the most recent cycle, an average of several cycles, or a default value you never changed. They may round an average differently or use different assumptions for ovulation. A missing start date can make one interval look unusually long. Counting spotting as a period in one tool but not the other can also shift the result. Compare the same recorded starts and cycle-length setting before comparing the predicted dates. Neither a precise-looking calendar nor agreement between two apps proves that an estimate is correct.
Check the record before changing the estimate
Keep observed starts separate from predicted starts. Check that you entered the first day, not the last day, of the previous period. Compare several complete intervals in the cycle-length calculator, and note the shortest and longest as well as the average. Mark a forgotten date as uncertain instead of inventing one. Record spotting separately when possible; our cycle-day-one guide explains that convention. If a clinician has given you a specific logging rule for treatment, follow it. Do not change a contraception schedule to match an app calendar.
What a prediction cannot tell you
Cycle-length information alone cannot reliably identify the exact ovulation day, as a 2018 study explains. It did not evaluate Bloom. A calendar prediction is not contraception or a pregnancy test. NHS fertility-awareness guidance describes a taught method using more than dates. If pregnancy is possible, use a test according to its instructions; see NHS testing guidance. A predicted date alone cannot establish the cause of a late period.
A useful record for your next step
A short record can contain the actual start date, how certain you are, bleeding duration, the prediction shown beforehand, and any questions you want to discuss. Keep personal details in a place you trust; you do not need to send a symptom diary to us to read this guide. Seek medical advice for a persistent change or symptoms that concern you; NHS guidance on missed or late periods explains when to get help. Use the tools below to organize dates, while keeping the estimate separate from the observation.