What 10,000 UCAS personal statements for 2027 entry look like
- 36% share of statements at or within 100 characters under the 4,000-character limit (about 10,100 UCAS personal statements reviewed on MyRevisionAgent)
- 12% share of statements using fewer than 3,000 of the 4,000 characters (about 10,100 UCAS personal statements reviewed on MyRevisionAgent)
- 39% share of the notes from the automated statement review that are "Weak claim", its most common type (about 90,500 notes on the latest version of each statement reviewed on MyRevisionAgent)
- 19:00 start of the busiest hour of the day for reviews, London time (about 10,800 reviews on MyRevisionAgent)
about 10,100 statements · Published 12 November 2026 · Free to reuse with a link (CC BY 4.0) · Analysis: Nic Bamberger, founder
The median UCAS personal statement for 2027 entry reviewed on this site is 3,958 characters long at the time of review, out of the 4,000 the three answers share (n = about 10,100 statements). 36% are at or within 100 characters under the 4,000-character limit, 12% use fewer than 3,000 characters and 27% were 4,000 characters or more when they were reviewed, a draft of exactly 4,000 characters counting in both this group and the first; these are drafts, and a draft can be longer than the 4,000 UCAS accepts when it is reviewed.
Based on version 1 of How long are UCAS personal statements, and where do they fall short? (12 November 2026). For the latest figures see the dataset.
No student work, names or schools are ever published. How we protect students.
Key findings
Median length at review
At the time of review, UCAS personal statements reviewed on MyRevisionAgent run from 2,726 characters at the 10th percentile to 4,396 at the 90th, around a median of 3,958; these are drafts, and a draft can be longer than the 4,000 UCAS accepts when it is reviewed.
Median share of characters per question
In UCAS personal statements reviewed on MyRevisionAgent, the median share of a statement's characters is 31% for why the course, 35% for qualifications and studies and 31% for preparation outside education; each is a separate median, so they need not add up to 100%.
Most common note type: Weak claim
The most common type of note from the automated statement review on MyRevisionAgent is "Weak claim", 39% of about 90,500 notes on the latest version of each statement.
Share reviewed again as a linked redraft
6.4% of UCAS personal statements reviewed on MyRevisionAgent were reviewed again as a linked redraft; those had a median of 2 reviews, and the median change in their score from the automated statement review, from first review to latest, was +2 points out of 100.
Busiest hour of the day
Across all reviews on MyRevisionAgent the busiest hour of the day starts at 19:00 London time, and among reviews by applicants with a 15 October course among their choices, 2% of those made before the 15 October 2026 deadline came in its final 24 hours.
Figures
Figure 1: Personal statement length, characters across all three answers
Latest reviewed version of each statement, at the time of review. Drafts can be longer than the 4,000 characters UCAS accepts.
| Characters | n | Share |
|---|---|---|
| 500-999 | 120 | 1.2% |
| 1,000-1,499 | 320 | 3.2% |
| 1,500-1,999 | 240 | 2.4% |
| 2,000-2,499 | 180 | 1.8% |
| 2,500-2,999 | 340 | 3.3% |
| 3,000-3,499 | 750 | 7.4% |
| 3,500-3,999 | 5,400 | 53.6% |
| 4,000+ | 2,700 | 27.1% |
Figure 2: What the review notes flag, share of all notes
| Type | Share |
|---|---|
| Strength | 25.7% |
| Weak claim | 38.6% |
| Generic | 18.1% |
| Unevidenced | 2.2% |
| Off target | 1.9% |
| Unreflected | 13.3% |
| Other (small groups merged) | 0.1% |
Figure 3: Reviews before the 15 October 2026 deadline, by hours left
Reviews by applicants with a 15 October course among their choices.
| Hours before | n | Share |
|---|---|---|
| 168+ hours | 3,100 | 81% |
| 72-168 hours | 400 | 10% |
| 24-72 hours | 270 | 7% |
| Final 24 hours | 90 | 2% |
The three questions in practice
UCAS introduced the three-question format for 2026 entry, in place of a single essay, and the three answers share one 4,000-character limit. Of about 90,500 notes the automated statement review left on the latest version of each statement, 34.3% are on why the course, 35.5% are on qualifications and studies and 30.1% are on preparation outside education. Notes marking a strength make up 27.4% of the notes on why the course, 27.6% of the notes on qualifications and studies and 21.7% of the notes on preparation outside education.
Redrafting and timing
The redraft share counts only linked redrafts: a redraft reviewed without a link to the earlier version counts as a new statement, so this is a lower bound. The deadline figures cover applicants with a 15 October course among their choices (Oxford, Cambridge, medicine, dentistry or veterinary science): of their reviews made before the 15 October 2026 deadline, 81% came a week or more before it.
Download and cite
Based on version 1 of How long are UCAS personal statements, and where do they fall short? (12 November 2026). For the latest figures see the dataset.
CC BY 4.0: free to reuse with a link to this page.
Cite as: MyRevisionAgent Research (2026-11-12). What 10,000 UCAS personal statements for 2027 entry look like. https://myrevisionagent.com/research/ucas-personal-statements-2027-entry
Analysis: Nic Bamberger, founder
This analysis is not updated. For the latest figures, see How long are UCAS personal statements, and where do they fall short?.
Questions
Which UCAS personal statement question gets the most feedback?
In the automated statement review on MyRevisionAgent, qualifications and studies gets the most notes: 35.5% of the notes on the latest version of each statement, against 34.3% for why the course and 30.1% for preparation outside education.
What is the most common feedback on a UCAS personal statement?
In the automated statement review on MyRevisionAgent, the most common type of note is "Weak claim" (39% of notes), followed by "Strength" (26%) and "Generic" (18%).
How much does a UCAS personal statement change between drafts?
Among UCAS personal statements reviewed on MyRevisionAgent that were reviewed again as a linked redraft, the median change in length from first review to latest was -7 characters, and the median change in the score from the automated statement review was +2 points out of 100.
Is my statement or essay in this data?
If your personal statement was reviewed on MyRevisionAgent in the period this page covers, it may be counted here, only ever as part of totals of at least 50. No statement text, name, school or account is published, and nothing on this page can be traced to one student. Admin accounts and reviews paid for by a school are left out altogether. To ask about the data, use the press and data requests form on this page.
About this data
Sample: about 10,800 reviews, about 10,100 statements, 8 July 2026 to 12 November 2026
Unit: One personal statement: every review of one statement editor document, or a single review made outside the editor
Excluded:
- Admin accounts
- Reviews paid for by a school
- Refused duplicate submissions
- Failed runs
- Course-planning reviews with no statement
- Statements under 500 characters across the three answers
Method
Source: UCAS personal statement reviews on this site
Suppression: Groups under 50 are merged into Other, and a larger group may be merged in with them so that no small group can be worked out. On an ordered scale (scores, grades, lengths, hours) a small band is joined to the band next to it instead, and the row names the joined range. Units with no value for a table's column are a Not recorded row when there are 50 or more of them. When there are fewer, a table on an ordered scale leaves them out and shows only shares of the units that have a value, rounded to whole percent, with no counts; the table's caption, its chart note and every sentence using those shares say so. Counts are rounded (to the nearest 100 above 1,000, otherwise to the nearest 10) so exact volumes are not published; shares and medians are calculated from exact counts, so rounded counts may not add up exactly.
What changed: Version 1, 12 November 2026: First published version
Definitions
- Review
- One completed review of all three answers of a personal statement.
- Statement
- The reviews of one statement editor document, or a single review made outside the editor. Length, score, notes and course fit use its latest review.
- Characters
- Characters across the three answers of the draft as it was sent for review, counted as the statement editor's counter counts them: every character including spaces, line breaks and leading or trailing spaces, most emoji as two. UCAS allows 4,000. These are drafts at the time of review, not submitted applications, and the review accepts drafts up to 8,000 characters, so some are longer than UCAS allows.
- Near the limit
- 3,900 to 4,000 characters, both included: at or within 100 characters under the 4,000-character limit.
- Question share
- Each answer's share of a statement's characters. The three medians are taken separately, so they need not add up to 100%.
- Review score
- The review's overall score for the statement, 0 to 100, and the band word it gives.
- Note
- One passage of a statement that the automated review marks, on the latest version of the statement, with one of the types below. A note type with any small group is merged whole into Other, in both note tables.
- Strength
- A note marking a passage that works.
- Weak claim
- A note on a line that works but could do more (shown to the applicant as "Could be stronger").
- Generic
- A note on a line that is true but says nothing particular to the applicant (shown as "Name the specific thing").
- Unevidenced
- A note on a claim with nothing behind it (shown as "Back this up").
- Off target
- A note on a passage that answers something other than the question it sits under.
- AI tone
- A note that a passage reads as generic or machine-written to a reader (shown as "Reads as AI-written"). It describes how the line reads, not how it was written: it is not an AI-detector result and not a finding that a statement was written by AI. (Currently merged into Other.)
- Unreflected
- A note on a passage that names a real experience, such as a book, a placement or a project, but not what it changed in the applicant's thinking (shown as "Say what it changed").
- Impact
- How much the automated review judges that acting on a note would move the statement's mark: high, medium or low. Counted over notes that ask for a change, not strengths. Not rated: a note from a review made before 29 September 2026, when impact was first recorded.
- Course fit
- The automated review's judgement of how well the statement supports each course choice: strong, moderate or weak. There is one per course choice, so its n counts choices, not statements.
- Redraft
- A statement with a second linked review. Change is measured from its first review to its latest.
- Course area
- Inferred from the course titles of the applicant's choices by keyword rules. Course fit is shown only for areas where each fit level has at least 50 choices; the other areas are merged into Other, and a larger area may be merged in with them so that no small one can be worked out.
- Hours before deadline
- Time from a review to 18:00 UK time on 15 October 2026. Only reviews made before it by applicants with a 15 October course among their choices (Oxford, Cambridge, medicine, dentistry or veterinary science).
- Hour and weekday
- London time, all reviews.
More data
- How long are UCAS personal statements, and where do they fall short?
- Which TOK prescribed title do students choose, and where do essays lose marks?
- Where do Extended Essays lose marks, by criterion and subject?
- All research - Every dataset and analysis in one place.
- How the graders are tested - The test sets, cross-validation and error ranges behind the essay graders.
- The UK course data behind the UCAS tools - Coverage and sources for the course and university data.
This data comes from the free statement review on MyRevisionAgent.
Press and data requests
Journalists, schools and researchers can ask for a custom cut of the data or a comment.