How WaitDelta Measures Visa Wait Estimates
By Smith Shah · Last reviewed October 10, 2026
WaitDelta reads the State Department's monthly Global Visa Wait Times publication. The publication contains two kinds of estimate, a historical average wait and a next-appointment estimate, and WaitDelta is verifying which of the two each stored figure represents. Each figure is shown with the date WaitDelta read it, which is not the date the State Department published it. Reading the same publication again does not create a new observation.
Source
State Department Global Visa Wait Times table
Source updates
Monthly
WaitDelta check schedule
Twice weekly, Monday and Thursday
Last completed check
October 1, 2026
What does WaitDelta measure?
WaitDelta records the State Department's published wait estimates for five U.S. posts in India. It does not observe booking systems, so it cannot show which dates are open to you. Each stored figure is one number of days per post and category, stamped with the date of the check that read it.
| Measure | What it means | Does WaitDelta store it? |
|---|---|---|
| Historical average wait | The State Department's estimate of how long interview appointments took to obtain over a recent period. | The publication contains it. WaitDelta stores one figure per post and category and is verifying whether that figure is this one. |
| Next-appointment estimate | The State Department's separate estimate of the next available interview appointment. | The publication contains it. WaitDelta stores one figure per post and category and is verifying whether that figure is this one. |
| Availability in your booking account | The dates actually offered to you when you book. | No. Only your own booking account shows it. |
| Time after the interview | Administrative processing and passport return, which sit outside the interview queue. | No. The duration varies by case. |
Use WaitDelta figures to compare posts and to see how a published estimate moved. A published estimate is not an appointment date, and it is not a promise.
Where does WaitDelta's data come from?
From one source: the State Department's Global Visa Wait Times table, which the Department updates monthly. The table groups visa classes into shared categories, and WaitDelta reads the rows for its five Indian posts.
WaitDelta shows several labels for each source category, so one number can appear under more than one label. Those labels are not independent datasets: H-1B and L-1 share one source figure, as do F-1, F-2, J-1 and J-2.
| State Department category | Labels WaitDelta shows | Status on WaitDelta |
|---|---|---|
| Visitor (B-1/B-2) | B-1/B-2, B-1, B-2 | Shown with its check date. Mapping is part of the open audit. |
| Student and exchange (F, M, J) | F-1, F-2, J-1, J-2 | Shown with its check date. Mapping is part of the open audit. |
| Petition-based (H, L, O, P, Q) | H-1B, L-1, H-4, L-2, O-1 | Withheld while we verify. |
| Post | Official name |
|---|---|
| Mumbai | U.S. Consulate General Mumbai |
| New Delhi | U.S. Embassy New Delhi |
| Chennai | U.S. Consulate General Chennai |
| Hyderabad | U.S. Consulate General Hyderabad |
| Kolkata | U.S. Consulate General Kolkata |
Each city page reflects one post only, not an India-wide figure. The source is the State Department's Global Visa Wait Times table; see also its wait times and processing page.
How does WaitDelta turn the table into stored numbers?
A scheduled job fetches the State Department page, reads the table rows for the five Indian posts and stores one row per post and category with the time of the check. Values published in months are converted to days at 30 days per month, and blank or NA cells store nothing.
- Fetch. An automated request service retrieves the page, because the State Department site blocks direct automated requests.
- Parse. The job finds each source category's column by its header text, reads the table rows that name one of the five posts and takes three cells from each, one per source category. If a header does not match, or a post appears twice, the job stores nothing and reports the failure. The mapping itself is still the subject of the open audit.
- Convert. A value published in months becomes days at 30 days per month. A value published as less than half a month is stored as 14 days, which is a WaitDelta convention and not a State Department number.
- Skip blanks. A cell marked NA or left empty stores nothing, so WaitDelta holds no value for it.
- Store. One row per post and category is added with the time of the check, the number of days and the source URL. Labels that share a source category receive the same number.
WaitDelta stores the converted number of days, not the text the State Department published. It does not yet archive the raw page or record the date on which the State Department published each table. The open audit adds both.
Every completed check adds a new row even when the State Department has not published a new table, so consecutive rows usually repeat the same value. Row counts measure WaitDelta's checks, not independent observations.
How often does WaitDelta check, and how often has it completed a check?
WaitDelta's check is scheduled twice a week, on Mondays and Thursdays at about 03:00 UTC, with a daily catch-up that repeats a missed or failed check, and the State Department updates its table monthly. The most recent completed check was October 1, 2026, 11 days ago, so a scheduled check is overdue.
| Completed check | Days since the previous check |
|---|---|
| Oct 1, 2026, 08:39 UTC | 3 |
| Sep 28, 2026, 08:32 UTC | 4 |
| Sep 24, 2026, 07:29 UTC | 3 |
| Sep 21, 2026, 07:51 UTC | 4 |
| Sep 17, 2026, 07:28 UTC | 14 |
| Sep 3, 2026, 07:04 UTC | 10 |
| Aug 24, 2026, 03:11 UTC | 4 |
| Aug 20, 2026, 03:03 UTC | 3 |
| Aug 17, 2026, 03:07 UTC | 1 |
| Aug 16, 2026, 00:52 UTC | Earlier check not listed |
The schedule calls for gaps of three or four days between checks. A longer gap is a gap in WaitDelta's checking, not a change at the State Department.
Failed or skipped runs store no figures, so they do not appear in this list. Since October 10, 2026 each of them is recorded in WaitDelta's own pipeline log, a failure raises an alert, and a daily catch-up repeats a check that was missed or failed. An overdue WaitDelta check delays WaitDelta's update; it does not mean the State Department stopped publishing.
What does WaitDelta do when a value is missing or a check is late?
A blank or NA cell is shown as Not published, and WaitDelta does not read it as a closure or as proof that no appointments exist. A late check is flagged with the date of the last completed check. A figure under verification is labeled Withheld, which is a WaitDelta status.
- Not published: the State Department's table had no value for that post and category in the latest completed check, or WaitDelta has not read one in 45 days.
- Check overdue: WaitDelta's last completed check is more than 5 days old. Pages say so and show that date.
- Withheld: WaitDelta has stopped displaying the figure while it verifies the mapping. The State Department did not stop publishing, and visa services were not suspended.
Which figures are withheld, and why?
WaitDelta is withholding petition-based figures (H-1B, L-1, H-4, L-2 and O-1) while it checks the source mapping behind them. The check compares WaitDelta's column mapping with the State Department table's headers. It was not complete when this page was last reviewed on October 9, 2026.
H, L, O, P and Q share one State Department estimate, so one source number feeds five WaitDelta labels. Until the check is complete, no petition-based figure appears on WaitDelta pages, comparisons or the timeline calculator. For booking and interview rules without figures, read H-1B stamping in India: Booking and interview rules.
The audit traces archived source input through parsing, storage and display. It covers seven checks:
- Source snapshot: raw input, URL, publication date, retrieval time and content hash kept together.
- Column mapping: all five posts compared with the exact source headers.
- Metric identity: historical average and next-appointment estimate stored in separate fields.
- Values and units: the original value kept beside any conversion, including NA and inequalities such as less than 0.5 month.
- Independent records: one observation per post, category, metric and publication.
- Historical repair: affected records rebuilt from archived input where possible, and gaps marked where not.
- Presentation: guides, city pages, comparisons and the calculator checked against the repaired output.
The audit covers every displayed category, because the difference between an average and a next-appointment estimate affects B-1/B-2 and F-1 as well as petition-based classes. Until it finishes, B-1/B-2 and F-1 figures appear with their check date, and which of the two estimates each represents is not yet confirmed.
How are trend arrows calculated?
A trend arrow shows the difference, in days, between the current stored figure and the latest figure stored at least 30 days earlier for the same post and category. A down arrow means a lower figure and an up arrow a higher one.
Trend = current stored days − stored days from at least 30 days earlier.
The State Department updates monthly, so a 30-day comparison usually spans at most one new publication, and a repeated reading of the same publication produces no change. An arrow describes a change between two stored figures; it does not explain why the figure changed. As an illustration with invented numbers, 75 days now against 90 days a month earlier is shown as ↓ 15 days.
What can WaitDelta's monitoring show about a policy change?
It can show that a published estimate changed between two readings. It cannot attribute that change to a policy such as the October 8, 2026 PERM action, and it cannot isolate the applicants of a named employer, because petition-based estimates cover every H, L, O, P and Q applicant at a post.
A later reading is not a baseline. A table read on October 9 is a reading made on October 9; it is not a measurement taken on October 8, and it is not a record of the position before an announcement.
WaitDelta publishes a monitoring report only when new, comparable readings show a change worth reporting. The report states the readings, their dates and these limits. For what the October 8 action does and does not cover, see Is H-1B suspended? What the October 8 PERM action changed.
Is WaitDelta affiliated with the U.S. government?
No. WaitDelta is an independent platform and is not affiliated with the U.S. Department of State, any U.S. embassy or consulate, or any other U.S. government body. It reports published estimates and does not set them.
Verify current wait times at travel.state.gov and in your booking account before you make travel or application decisions. For every tracked class and post, see US visa wait times in India.
Changes to this page
- October 10, 2026. Changed how WaitDelta's check handles failure. Every run, including a failed one, is now recorded; the check finds each column by its header text and stores nothing when the headers do not match; and a daily catch-up repeats a missed or failed check. The last completed check before this change was on October 1, 2026.
- October 9, 2026. Rewrote this page to describe the process as it runs. Removed the statement that WaitDelta's twice-weekly check follows the State Department's publication schedule: the State Department publishes monthly, and WaitDelta's check is a separate schedule.
- October 9, 2026. Removed the description of every figure as an average of the previous 30 days, the claim of 20 independent time series, the statement that Not Available means no standard appointments exist, and the universal 3 to 10 business day administrative-processing figure. None could be supported.
- October 9, 2026. Withheld petition-based figures (H-1B, L-1, H-4, L-2, O-1) while the source mapping is verified, and separated an overdue WaitDelta check from whether the State Department still publishes a figure.