By Cyril KimbiA recommendation letter is the only part of your application written by someone else, and the only part an admissions committee treats as independent evidence. That is exactly why a weak one does more damage than students realise — and why most weak letters are the applicant’s fault, not the referee’s.
What committees are reading for
Not praise. Praise is assumed. They are reading for three things:
- Evidence — a specific instance, not an adjective. “Diligent” tells them nothing; “reworked her project after the first result contradicted her hypothesis, which nobody else in the cohort did” tells them a great deal.
- Comparison — where you sit relative to others the referee has taught. “Among the strongest three students I have supervised in six years” is the single most weighted sentence in most letters.
- Credibility of the writer — do they actually know you, and in what capacity?
A short letter that names one concrete thing you did beats a long one full of warm generalities.
Choosing referees
The instinct is to ask the most senior person available. That is usually the wrong instinct.
Who knows your work beats who has the biggest title. A letter from a professor who taught you and supervised your project carries far more weight than one from a dean who met you twice. Committees can tell the difference immediately, because one letter contains detail and the other does not.
Prefer, in order: someone who taught and assessed you in the relevant subject; someone who supervised a project or dissertation; a work or internship supervisor for professional programmes. Avoid family friends, religious leaders and political figures unless the scheme explicitly asks for a character reference — in an academic file they signal that no academic was willing.
Where a scheme asks for three letters, aim for a spread: two academic, one supervisory or professional.
How to ask
The request itself affects the quality of what you get.
- Ask early — three to four weeks minimum. A rushed letter reads as rushed.
- Ask in person where possible, then follow up in writing with everything they need.
- Ask whether they can write a strong letter. This phrasing matters. It gives a hesitant referee a graceful way to decline, and declining is far better for you than a lukewarm letter you never see.
- Make declining easy. “I completely understand if you are too pressed for time” costs you nothing and protects you from a bad letter.
The pack to send
This is where most applicants lose marks they never know about. A busy academic writing from memory produces a vague letter. Send, in one email:
- Your CV.
- The statement or essay you are submitting, so the letter reinforces rather than contradicts it.
- The programme and institution, with a line on why you are applying.
- The deadline and the submission method — portal link, upload, or posted letter.
- A short reminder of your work with them: the course, the year, your grade, the project title, and one or two things you did that they might remember.
- Anything specific you hope they can speak to, framed as a suggestion, not an instruction.
That last item is legitimate and expected. What is not legitimate is drafting the letter yourself for them to sign — and if a referee asks you to, treat it as a signal to find another referee, because a self-written letter reads unmistakably like one.
Asking from Cameroon: two practical problems
Institutional email. Many portals require the referee to submit from an institutional address, and a letter arriving from a personal webmail account is sometimes rejected outright. Confirm your referee has access to an institutional address before naming them.
Reachability after graduation. Lecturers move, and a referee you cannot contact eighteen months later is a referee you do not have. Before you leave an institution, get a working personal email for anyone likely to write for you, and stay in occasional contact. This costs nothing at the time and is unrecoverable later.
A third, smaller point: if your referee is unfamiliar with the destination system, tell them the grading scale. A letter saying a student scored 15/20 means little to a committee that does not know 15/20 is strong. One line — “placed 4th of 96” — solves it.
Waiving your right to see the letter
Many American applications ask whether you waive the right to read the letter. Waive it. Confidential letters are taken more seriously, because a referee writing in confidence is assumed to be candid. Declining to waive suggests you do not trust your own referee.
Following up
A single polite reminder a week before the deadline is appropriate and generally welcomed. Confirm submission through the portal rather than asking again. Afterwards, tell your referees the outcome — accepted or not. It takes two lines, it is the decent thing to do, and it is why they will write for you again.
Frequently asked questions
How many letters do I need?
Usually two or three. Send exactly the number requested; extra letters are rarely read and can read as trying too hard.
How far in advance should I ask?
Three to four weeks at minimum, more during examination periods when academics are least available.
Can a work supervisor write it?
Yes, especially for professional and taught programmes. For research programmes, academic referees carry more weight.
What if my referee asks me to draft it?
Decline politely and offer a detailed summary of your work instead. A self-drafted letter is recognisable and undermines the file.
Should I read the letter before it is sent?
No. Confidential letters are given more weight, and waiving the right to read them is the stronger signal.
