How these numbers are calculated

The one thing to know

A median “interview to approval” time can only be computed from cases that reached an approval. A case that is still waiting has no approval date, so it contributes nothing — no matter how long it has been going on. The longer someone waits, the less likely they are to appear in the very statistic meant to describe waiting.

So the timing figures elsewhere on this site describe roughly 15% of what we track. They are accurate about cases that finished. They are not a forecast, and they systematically understate the experience of anyone in a long wait.

What that means when you compare yourself

• Being past the median is common and means less than it feels like. Roughly half of finished cases sat above it, and the currently-stuck are not in the figure at all.
• A published maximum is not a ceiling. It is the longest wait that has ended and been reported. Longer waits exist and are, by definition, still running.
• Small consulate samples move a lot. Where a page shows a handful of resolved cases, treat the median as a rough indication rather than a number.
• Pending counts are the honest half of the picture. Where we show them, read them alongside the median rather than after it.

Where the data comes from

• Timelines submitted directly on this site, plus anonymized case reports compiled from publicly shared accounts.
• Every record is stripped of identifying details before storage — no names, usernames, employers, universities, or passport, DS-160 or case numbers. Source links are never displayed.
• Records are deduplicated per person-timeline, so one applicant posting several updates is one case, not several.
• Dates given as a month only are stored as a month, and never counted in day-level calculations.

Known limits

• Reporting is voluntary and self-selected. People are likelier to post an outcome than a continuing silence, which biases visible outcomes toward resolution and toward the unusual.
• Approval is over-represented against refusal for the same reason, so nothing here should be read as an approval rate.
• Dates are as reported. We reject impossible ones (future dates, negative gaps) but cannot verify the rest.
• This is observational community data, not government statistics, and not legal advice.