Prompt Bank collection and the script check
We read each provider's own application or rules page and recorded the prompt word for word, with the word or character limit, the deadline, the award amount if stated, the source link and the read date. Prompts from a PDF that the provider hosts count, and the PDF link is the source. A prompt was never written from memory, and a limit or deadline the page did not state was left empty.
On Nov. 18, 2026 a script opened every source page again and tested whether the recorded quote appears in the page text. Each record carries the result. The result for all 120 records is printed in the Prompt Bank. More is on the Prompt Bank method page.
Classification into nine types
Each prompt was put in one of nine types by what it asks: why you deserve the award (2 prompts), career goals (22), financial need (6), a challenge you overcame (9), community service (8), leadership (6), tell us about yourself (29), why this field (10) and unusual prompts (28). A prompt that touches two themes was put under the one it asks for first. The counts table is below.
| Prompt type | Prompts | Median word limit | No limit stated |
|---|---|---|---|
| Tell us about yourself | 29 | 500 | 5 |
| Career goals | 22 | 550 | 1 |
| A challenge you overcame | 9 | 250 | 2 |
| Why you deserve it | 2 | 250 | 1 |
| Community service | 8 | 550 | 1 |
| Leadership | 6 | 400 | 1 |
| Financial need | 6 | 250 | 0 |
| Why this field | 10 | 325 | 3 |
| Unusual prompts and contest themes | 28 | 500 | 4 |
| All prompts | 120 | 500 | 18 |
The Reuse Map numbers
The Reuse Map is built from the same records. For each type it shows how many prompts there are, the median word limit and how many state no limit. Across all 120 prompts, 87 state a word limit, 15 state a character limit and 18 state none. The median word limit is 500, and 24 prompts are set at exactly 250 words and 16 at exactly 500. The Reuse Map method page shows how to use these numbers.
The ranked view: criteria and weights
Two kinds of pages rank application platforms: the page on applying to many scholarships at once and five pages for different readers. Each platform is scored from 0 to 10 on seven criteria. The weights, which add to 100, are below.
| Criterion | What is scored | Weight |
|---|---|---|
| One profile, many applications | Apply to several scholarships from one profile, including automatic re-application to recurring awards | 25% |
| Matching to eligibility | Shows the student only what they qualify for, on stated inputs | 20% |
| Vetted listings | A stated process for screening listings, and what it excludes | 15% |
| Tracking | Status of every application in one place | 10% |
| Essay reuse | One essay adapted to several applications | 10% |
| Cost to the student | Free scores 10 of 10; a paid membership scores lower, and the price is printed in the table | 10% |
| Trust signals | Public review scores quoted with source and date, years operating, a named legal entity | 10% |
Weights committed on Nov. 18, 2026, before any platform was scored. They have not changed since.
Each criterion has written anchors for 0, 3, 6, 9 and 10. A fact that only the platform claims, with no second source, caps its criterion at 7. If no evidence was found, the score is 0. Breadth facts such as database size, number of listings, user counts and funding are not scored.
Five situation weight sets
A situation page ranks on that situation's own weights, not merged with the main weights.
First scholarship essay: cost 20, vetted 20, matching 20, one profile 15, essay reuse 10, tracking 10, trust 5. A first-timer is the easiest reader to mislead and has the least reason to pay: screening, matching and cost come first; volume tools matter less until the first essay exists.
Ten applications in one month: one profile 30, essay reuse 20, tracking 15, matching 15, vetted 10, cost 5, trust 5. Volume against a deadline: submitting many from one profile, reusing one essay and tracking status carry the most weight.
Strong story and weak writing: matching 25, vetted 20, essay reuse 15, one profile 15, cost 10, tracking 10, trust 5. Each essay costs this reader a lot of effort, so it should go only to awards they can win and that are real: matching and screening first.
Writing in a second language: essay reuse 25, one profile 20, matching 20, vetted 15, cost 10, tracking 5, trust 5. Writing each new essay takes longest for this reader, so reusing one well-edited essay and applying from one profile carry the most weight, with matching close behind.
Adult returning to study: matching 25, cost 20, vetted 15, one profile 15, tracking 10, essay reuse 10, trust 5. Few awards fit adult learners, so matching on age, part-time study and field matters most; adults pay their own way, so cost weighs more.
Entry condition
A platform is ranked only if its own pages show it serves any scholarship applicant. Otherwise it is listed as "not for this reader on its pages, read" with the date.
Blind scoring and the weights commit
The weights, the criteria and the entry condition were committed on 2026-11-18, before any platform was scored. The scores are the network score set for these platforms, made blind on 2026-11-18 from merged, dated evidence: a fresh scorer received only the rubric and a file of sourced facts, and two further blind reviewers checked it. None of them was told which site the scores were for, any weight or any view. Every score cites the fact numbers behind it, and every change from this site's first scores is logged in score-changes.jsonl with its reason.
Cost is scored for the plan a platform's other scores depend on: if a feature that earns points elsewhere in the table needs a paid membership, the platform is scored on that membership's price, and if that price is not published before sign-up it gets the score for an unpublished price.
The scores are published as scored, including those that are judgement calls. A dollar price on a platform's own live help center counts as published even when the article is old; read the other way, ScholarshipOwl's cost score would fall from 4 to 1, which moves its total by 0.30 points on the main view and by 0.60 on the first-essay and adult-returning views. The ranking tables show every total.
Questions this page answers
- How can I check that a quote is on its source page?
Open the source link printed under the prompt in the Prompt Bank and search the page for the first few words of the quote.
- When were the weights fixed?
On 2026-11-18, before any platform was scored.