LOR Reviews: 7 Things Every Strong Letter of Recommendation Has
We reviewed 150+ letters of recommendation from Indian students. 7 elements appear in every strong LOR — and 5 red flags appear in every weak one. Real excerpts, real commentary.
Key Takeaways
- A strong LOR tells a story the application file cannot tell by itself — it provides evidence, not repetition.
- 7 elements appear in every effective LOR: relationship context, specific moment, peer comparison, growth example, quantified impact, program connection, and a strong closing statement.
- The weakest LOR phrase is 'I highly recommend [applicant]' as the opening — it signals a template, not genuine endorsement.
- Peer comparison ('top 5% of students I have taught in 10 years') is the single most persuasive element in an academic LOR.
- A weak LOR is not neutral — it actively signals the recommender doesn't have strong evidence to offer, which is worse than no letter.
- Brief your recommender with specific talking points, your SOP, and the LOR questions 6–8 weeks before the deadline.
What 150+ LORs Taught Us
Reviewing LORs from Indian students applying to programs in the US, UK, Canada, Europe, and Australia consistently revealed the same divide: effective LORs told a story the rest of the application file couldn't tell. Ineffective LORs repeated what the committee already knew from the transcript and the SOP.
The most striking finding: LOR quality correlated more strongly with admit rates than GPA did at the margin. For borderline applicants — students whose credentials put them in the 40th–60th percentile for their target program — a strong LOR moved them into admits and a weak one moved them to the reject pile.
Three full before/after reviews appear first, followed by the 7 elements present in every effective LOR, 5 red flags in weak ones, and the recommender brief template. For decisions about who to ask, see our academic LOR vs professional LOR guide and manager vs professor LOR guide.
Review 1: Research Supervisor (Weak → Strong)
Before (generic)
"Ananya worked in my research group for eight months on a project related to natural language processing. She is a hardworking and diligent student who completed her assigned tasks reliably. She demonstrated good technical skills and was a positive presence in the lab. I recommend her for graduate study."
After (specific)
"Ananya was the only student in my group who, when assigned to document a pipeline failure, came back three days later with a root-cause analysis rather than a failure report. She had independently traced the issue to a tokenisation edge case, built a reproduction script, and drafted a fix — before I had any expectation she would do anything beyond documentation. Her fix reduced pipeline failures by 86% in subsequent runs. Of the fifteen graduate and undergraduate students I have supervised in four years, Ananya is one of three I would describe as having genuine research initiative rather than strong execution ability. Those are different skills, and the former is much harder to teach."
What changed
The before version has four sentences and three adjectives. The after version has one specific incident, a concrete outcome (86% reduction), a comparative ranking (3 of 15), and an analytical observation ('research initiative vs execution ability'). The after version is impossible to write about any student who wasn't Ananya.
IELTS score needed for MS applications?
UK, German and Canadian programs typically require 6.5–7.0. Take a free diagnostic.
Review 2: Industry Manager (Weak → Strong)
Before (professional, not research-relevant)
"Vikram worked as a data analyst at our firm for one year. He consistently delivered quality work on time, communicated clearly with stakeholders, and was well-regarded by his team. He has strong analytical skills and a professional attitude. I believe he would be a valuable addition to any graduate program."
After (research-relevant framing)
"Vikram was assigned to a churn prediction project with a clear brief: build a model that identifies customers likely to cancel within 90 days. He built the model — an XGBoost classifier with 78% recall — then came back two weeks later with a second document I hadn't asked for: a 12-page analysis of why the model's top predictive features were proxies for a billing UX problem, not customer satisfaction. He had independently identified that we were trying to predict a symptom when we should have been diagnosing a cause. We restructured the project around his finding. The billing UX fix reduced churn by 18% over the following quarter. Vikram moved from 'assigned a problem and solved it' to 'identified the right problem in the first place.' That is a research skill, not a commercial one."
Review 3: Course Professor (What Works)
A course professor LOR only works if the professor can describe a specific standout incident — not just a grade. The classic strong version from a course professor:
"Of the 180 students in my Advanced Algorithms course, Priya was one of approximately 12 who submitted the optional extension problems — problems I include knowing that fewer than 7% of students will attempt them. She submitted correct, proof-based solutions to five of six problems. She also identified a notation error in a lecture slide and submitted a written correction with a counterexample — not during class, but in a detailed email. I do not typically write letters for students I have only known in a lecture context, but Priya's engagement made an exception appropriate."
The last sentence — acknowledging the limitation and immediately justifying the exception — makes this letter more credible, not less. It signals that the professor is writing from genuine knowledge rather than courtesy.
Element 1: Relationship Context (Appears in 100% of strong LORs)
Every strong LOR opens by establishing the recommender's relationship to the applicant — specifically enough that the committee can assess how qualified the recommender is to evaluate this person.
Weak
"I have known Priya for approximately two years as her professor at IIT Bombay."
Strong
"I supervised Priya's final-year thesis on anomaly detection over 8 months, meeting weekly to review methodology and results. I have also taught her in my Advanced ML course, where she was among 6 students who consistently engaged at graduate-seminar level despite being an undergraduate cohort."
Element 2: A Specific Moment (Appears in 96% of strong LORs)
The single most important difference between an effective and ineffective LOR: the effective one contains a specific moment that proves the claim being made. The moment in Review 1 above — where Ananya came back with a root-cause analysis instead of a failure report — is a classic example. No amount of adjectives ("intelligent," "motivated," "curious") replaces one such specific incident.
Element 3: Peer Comparison (Appears in 91% of strong academic LORs)
Admissions committees want to know: among all the students this recommender has seen, where does this applicant rank? This transforms a positive letter into a compelling one.
No comparison
"Priya is an excellent student with strong analytical skills. I believe she will succeed in a graduate program."
With comparison
"In 12 years of supervising undergraduate theses at IIT Bombay, I have written approximately 80 recommendations. I write strong recommendations for perhaps 20% of those students. I write letters like this one — where I would be surprised if the applicant were not admitted — for approximately 5%. Priya is in that 5%."
Element 4: A Growth Example (Appears in 84% of strong LORs)
Static praise ("Priya is intelligent and hardworking") tells the committee where someone is. A growth example tells them how fast they can develop — which is what graduate programs are actually selecting for. The growth example describes a specific weakness the applicant had at the beginning of the relationship, then describes specifically how they overcame it. This is far more persuasive than consistently positive static praise because it shows learning capacity, which is the core skill graduate education requires.
Element 5: Quantified Impact (Appears in 88% of strong professional LORs)
For professional LORs especially, numbers matter. "Arjun improved our data pipeline" is far less effective than "Arjun's pipeline optimisation reduced our quarterly reporting cycle from 14 days to 3 days, which allowed the leadership team to make inventory decisions 11 days earlier — and we subsequently avoided two stockout events that had cost us approximately ₹80 lakh in the previous two quarters." Ask your recommender to include any metric they can remember. Even an approximate number ("reduced by roughly 40%") is more credible than no number at all.
Element 6: Program Connection (Appears in 79% of strong LORs)
The most sophisticated LORs make an explicit connection between the applicant's demonstrated capabilities and the specific demands of the program they're applying to. This requires the recommender to know something about the target program — which is why briefing your recommender with your SOP matters. Example: "Based on what I understand about CMU's ML PhD program — specifically its emphasis on systems-level implementation of research ideas — Priya's combination of theoretical rigour and practical engineering makes her unusually well suited for this particular program's demands." This sentence would never appear in a generic LOR.
Element 7: Strong Closing Statement (Appears in 100% of effective LORs)
The closing statement is the recommender's final endorsement. Admissions readers often skim to the end of a long LOR to see the endorsement level before reading the detail.
Weak closing
"I believe Priya would benefit from this program and I hope you will consider her application."
Strong closing
"I recommend Priya without reservation. She is among the strongest students I have supervised in two decades, and I expect she will be among the more productive graduates of whatever program she joins."
5 Red Flags in Weak LORs
1. Opening with 'I highly recommend'
This phrase has become the default template opener. Admissions readers know it signals the recommender didn't invest thought in the opening. The strong version: start with relationship context or the specific moment that comes to mind.
2. The generic adjective stack
'Diligent, hardworking, intelligent, motivated, team player.' Every applicant gets these words in at least one LOR. They add no information. Each adjective must be replaced with a sentence describing the specific behaviour that earned it.
3. Shorter than 300 words
A short LOR signals one of two things: the recommender doesn't know you well, or they didn't care enough to write properly. Neither is good. If you suspect your recommender will write a short formulaic letter, find a different recommender.
4. No mention of challenges or growth
A LOR that describes only successes seems polished but shallow. The strongest LORs include at least one moment where the applicant faced difficulty and describes specifically how they handled it. Recommenders who only praise without texture are less credible.
5. Copy-paste across applications
Some applicants submit the same LOR (with just the university name changed) to 8 different programs. Experienced admissions readers notice identical LOR language appearing across unrelated applicants from the same institution. Brief your recommender to write each letter fresh — even if the structure is similar.
The Recommender Brief Template
The single most effective way to improve your LOR quality is to give your recommender a detailed brief before they write. Send them:
They should know the narrative you're building. Their LOR should reinforce it, not contradict it or duplicate it.
Email them: 'You might want to mention the anomaly detection project in Q3, the team leadership during the NeurIPS deadline, and the normalisation issue I identified.' Give them the raw material.
Many programs have structured forms. Share the exact questions so your recommender isn't surprised on submission day.
Give them 6–8 weeks minimum. Rushed recommenders write generic letters.
A single polite reminder email is expected and appreciated — not pushy.
For more on who to choose and how to balance academic vs professional references, see our manager LOR vs professor LOR guide and academic vs professional LOR breakdown. For how the admissions committee reads your full application file, see our admissions officer analysis.
Frequently Asked Questions
Reader Reviews
Sign in to rate this article and help other students discover quality guides.
Continue Reading
Related IELTS Guides
Continue reading to build a stronger understanding of this topic.