Good program fit is not name-dropping. It is a grounded match between the applicant's preparation, future research questions, and the intellectual environment of a specific programme, track, faculty, or lab.
Fit has three independent axes, and a target can pass one while failing another:
- Structural — how does this target admit people, and what is a shortlist entry here? (step 1)
- Thematic — is the applicant's actual work inside the programme's declared theme, specialisation, or track scope? (step 2)
- Intellectual — which specific people or groups does the applicant connect to? (steps 3–5)
A fourth check, goal fit, compares the applicant's stated post-degree goal with the graduate the programme intends to produce (step 2b). It flags and reweights; it never drops a target.
Check theme before people. Where a themed programme has no single advisor to map — a consortium master's with specialisation pathways, a thematic call, a mission-scoped track — thematic fit is the fit assessment.
Use this when
- The user is choosing PhD programmes, research master's, departments, labs, or advisors.
- The user asks whether their profile matches a programme's declared theme, specialisation, track, or application focus.
- The user wants to map their interests to faculty or research groups.
- The user asks whether a professor is a strong, medium, or weak fit.
- The user needs target-specific fit notes before writing an SOP.
- The user wants a balanced shortlist based on research fit rather than prestige alone.
- The user is deciding whether a given programme is worth applying to at all.
Do not use this when
- The user wants to write the SOP itself. Use
apply-sop-writer. - The user wants a broad literature review not tied to application fit. Use
literature-triangulation. - The user wants to audit a finished application package. Use
apply-package-auditor. - The user wants CV structure or bullet rewriting. Use
apply-cv-builder. - The question is only what a programme, track, or call selects for, with no applicant evidence in play — including characterising a whole candidate list before comparing any of it to a profile. Use
apply-program-theme-extractor, then return here with its dossier.
Workflow
1. Extract the applicant's research direction
Identify:
- target field and subfield
- concrete research questions or themes
- prior projects and methods
- preferred theory/empirical/system-building balance
- constraints: country, funding, geography, deadlines, degree type
- target programs or faculty already under consideration
If interests are vague, turn them into several searchable themes instead of forcing one narrow topic.
Establish each target's application regime before shortlisting — it decides what a shortlist entry even is. references/application-regimes.md carries the four regimes and their consequences:
- Supervisor-first (France, French-speaking Belgium, many project-funded posts) — the unit is a person and a project; outreach is the critical path; depth beats breadth.
- Vacancy-first (Sweden, Norway, MSCA/EURAXESS) — the unit is an advertised position and the applicant becomes an employee; you monitor vacancy boards rather than shortlist departments.
- Call/competition-first (Italian bandi) — the unit is a specific call with ranked selection and a fixed window.
- Department-first (typical US admissions) — the unit is a department with several plausible advisors; pre-application contact is optional and sometimes discouraged.
Identify the funding regime separately, since it may have a different selector — programme-attached, a standalone scholarship, project or employment funding, or a nominated scheme screened by the host institution or a national sending partner. Also check for the two edge cases: funding that is eligibility-assessed rather than competitive (Italian DSU is income-based), and levels where a country simply has no general government scholarship, so there is nothing to apply to.
Ask which regime applies per target rather than assuming from country — most countries contain several, and a single university can differ by department or funding source.
2. Screen thematic fit — before mapping any faculty
Many programmes declare a theme, specialisation, track, or application focus that scopes what work belongs there: an Erasmus Mundus joint master's with named specialisation pathways, a doctoral school organised around a mission area, a call restricted to listed topics, a centre funded for one agenda. That declared scope is published selection content, not marketing. references/application-regimes.md carries the theme axis and the extraction method; the short form:
If a theme dossier from apply-program-theme-extractor already exists for this target, consume it — take its declared theme, decomposition, pathway list, and status label as given and start at classification below. Otherwise run the extraction inline. Where the dossier's status is inferred, the operative theme may reorder which real evidence leads, but it cannot produce an "outside scope" verdict — only a declared theme can drop a target. Where the status is genuinely open, skip this step and record it as such; where it is link needed, leave the target unclassified.
- Extract the declared theme from the live official pages — the programme's own words, plus the pathway/track list, the stated objectives, and any funder mission behind it. Never infer a theme from the programme's name or your prior knowledge of the field. Search the web for every target you do not already have page text for, including programmes you have never encountered; a programme absent from your prior knowledge is the normal case, not a reason to skip the screen. Fetch the programme/consortium page, the current cycle's call, and each pathway page separately, and record the URL and access date. Where the official page cannot be located or the cycle's call is unpublished, ask the user for the link and leave the target unclassified rather than screening against a guess.
- Decompose it into the concrete research activities it actually funds. A theme like "digital and green transition" is a container: read the pathway descriptions, module lists, and example projects to see which mathematics or engineering it means, and note where the container is genuinely broad rather than narrowing it for the user.
- Map the applicant's real evidence onto it — projects, methods, data, application domains, publications. Method-level matches (optimisation, PDEs, statistical learning) count only where the programme's own materials show that method serving the theme.
- Classify, and let the verdict decide whether the target stays on the list.
Classify each target's thematic fit as:
- Core — the applicant's existing work sits inside the declared scope on its own terms. State which pathway or track it maps to.
- Adjacent, bridged — the work is outside the stated theme but connects through a named, checkable bridge: a specific method the applicant has used that the programme's own materials apply to this theme, a shared problem structure, or a domain the applicant has actually worked in. Write the bridge down as one sentence and record what evidence supports it.
- Stretch — a plausible connection exists but rests on a project the applicant has not done, an interest they cannot evidence, or a claim that would need the programme to take their word for it. Treat as a weak target and say why.
- Outside scope — no honest bridge exists. Deprioritize or drop the target. This is a legitimate, expected output. Do not manufacture a connection to keep a target alive; a fabricated theme bridge is the single most detectable weakness in a themed application and it costs a wasted cycle.
Two guards:
- Facts fixed, weights vary. A thematic bridge changes which of the applicant's real work leads, how deeply it is described, and which published criterion it is framed against. It never changes what they did, never adds an interest they do not have, and never converts an unrelated project into a themed one by relabelling it.
- The swap test. If the bridging sentence still reads fine after substituting a different specialisation, it is not evidence of thematic fit — it is filler. Rewrite it or downgrade the classification.
Where thematic fit is core or adjacent, bridged, hand the bridge sentence and its supporting evidence to apply-sop-writer. Where it is stretch or outside scope, say so before the user invests in the application.
2b. Check goal fit — flag, never drop
Where the applicant has a stated post-degree goal and a programme profile exists (from apply-program-theme-extractor, pass C) — or the user asks — compare that goal with the programme's intended graduate: the declared graduate profile and learning outcomes, and any published destination data. Fetch the graduate-profile and career-prospects pages live if no profile exists, recording URL and access date; where none is published, record goal fit: unassessed — no graduate profile published. Extract that profile first and applicant-free — its content must be unchanged for a different applicant, so do not let the applicant's goal, draft, or background decide what is read, kept, or phrased — and only then relate it to the applicant. Label each piece of evidence: formal graduate profile or learning outcomes = declared; destination or alumni data = evidenced; career-prospects or brochure copy = advertised unless it is itself a formal profile or is backed by destination data, which cannot support an aligned or divergent verdict on its own.
Report one of:
- Aligned — the declared profile or destination data points where the applicant wants to go.
- Divergent — for example a PhD-pipeline goal against an industry-oriented professional master's, or an academic-career goal against a development scholarship built to return graduates into home-country policy roles. Name the divergence and what it means for the application ("what after" answers, which evidence leads).
- Unassessed — no goal stated, or no profile published.
Keep the labels apart: a declared graduate profile and a programme's advertised selling points are different evidence, and a brochure line is never a graduate profile. Goal fit never produces "outside scope" and never drops a target — only a declared theme can. Where the funding scheme's own published criteria score post-degree plans (development, bilateral, or leadership missions), a divergence is a real risk; say so, and let the user decide.
3. Define fit criteria
Use criteria such as:
- topic overlap with the applicant's future questions
- methodological overlap with prior preparation
- active recent work by the faculty or lab
- room for the applicant to contribute something specific
- departmental ecosystem: multiple relevant faculty, seminars, centers, collaborators
- evidence quality: papers, lab pages, grants, student placements, recent projects
- risk factors: inactive advising, weak overlap, only one possible advisor, outdated webpages
Separate strong evidence from assumptions that need verification.
4. Classify fit
For each professor, lab, or program, classify:
- Strong fit: clear topic and method overlap, active work, multiple concrete bridges to applicant's background.
- Medium fit: plausible overlap but needs more evidence or has partial mismatch.
- Weak fit: mostly keyword overlap, prestige-driven, inactive, or unrelated to the applicant's actual direction.
- Unknown: insufficient information; list what to check.
5. Produce target-specific fit notes
For each target, produce notes that can feed later SOP tailoring:
- why this faculty/lab fits
- which applicant experience connects to the fit
- what future question could be pursued there
- what evidence still needs checking
- what not to say because it is shallow or unsupported
Do not write polished SOP paragraphs unless the user asks; hand off to apply-sop-writer for prose drafting.
6. Balance the shortlist
If the user is building a list, group programs by:
- thematic fit level, with outside scope targets moved off the list rather than ranked last
- research fit strength
- application competitiveness if known
- advisor depth and single-advisor risk
- funding or structural constraints
- deadline priority
Avoid presenting prestige as a substitute for fit.
Output format
Use:
## Applicant research themes
- [Theme]: [why it matters for search]
## Fit criteria
- [Criterion]: [how to judge it]
## Thematic fit
| Target | Declared theme / track (as published) | Thematic fit | Bridge (one sentence, or "none") | Evidence behind the bridge |
|---|---|---|---|---|
<!-- Thematic fit: core / adjacent, bridged / stretch / outside scope -->
## Goal fit
| Target | Applicant goal | Intended graduate (declared / evidenced) | Goal fit | Source |
|---|---|---|---|---|
<!-- Goal fit: aligned / divergent / unassessed — never a drop -->
## Faculty/program fit table
| Program/lab/faculty | Regime | Fit level | Evidence | Risks / unknowns | Follow-up |
|---|---|---|---|---|---|
## Target-specific fit notes
### [Programme, track, or professor]
- Thematic fit and pathway: ...
- Fit bridge: ...
- Applicant evidence to mention: ...
- Future question: ...
- Verify before applying: ...
## Shortlist priorities
1. [Target to investigate/apply to first]
## Deprioritized targets
- [Target] — [why the theme, regime, or evidence does not support applying]
Quality bar
A strong output distinguishes real intellectual fit from shallow keyword overlap, and real thematic fit from a relabelled project. It should help the applicant investigate programmes, produce credible target-specific notes, and avoid wasting application effort on poor-fit targets. An output that finds every target to be a good fit has failed: naming at least what would have to be true for a weak target to become viable — or dropping it — is part of the job.