Writing Recommendation Letters That Actually Carry Weight
A recommendation letter is more useful when it does something the rest of the application cannot. This is preparation guidance, not legal advice.
It shouldn't simply confirm that someone is talented, hardworking or pleasant to work with. Those statements are easy to make and difficult for a reviewer to evaluate.
A good letter gives the reviewer another person's informed account of the applicant's work: what the referee saw, what the applicant actually did, what happened as a result, and why the work was significant.
That distinction matters in both UK endorsement applications and US immigration cases. The precise criteria differ between routes, but the underlying problem is similar. Generic praise is easy to discount. Specific, credible testimony is much harder to ignore.
"I have known [Applicant] for three years and can confidently say they are one of the most talented, dedicated, and hardworking professionals I have worked with. Their contributions to our team have been invaluable, and I have no doubt they will continue to excel."
There is nothing necessarily false about this. The problem is that it gives the reviewer very little to work with. What contribution was invaluable? What did the applicant actually do? What was the result? Why is the referee qualified to make that assessment? How does the referee know the work was significant? The same paragraph could be used for hundreds of applicants with almost no changes.
"I have worked alongside [Applicant] for three years as a fellow senior engineer at [Company]. I was not their manager and have no financial interest in this endorsement. In 2025, I reviewed the architecture for the redesign of the company's payments infrastructure, which [Applicant] led. The new system now processes transactions for more than 300,000 active customers and, according to the team's post-launch reporting, reduced failed-payment incidents by 40% during its first year. The approach [Applicant] developed for idempotent transaction handling was subsequently adopted by two other engineering teams as their reference implementation. I consider this one of the strongest examples I have seen of [Applicant] combining technical judgement with practical, organisation-wide impact."
This version gives the reviewer several things the first one does not: who the referee is and why they are qualified to comment, how the referee knows the applicant, what work the referee personally observed or reviewed, the scale of the work, measurable results, evidence that the work was adopted beyond the original project, and a professional judgement grounded in the facts that came before it. The difference is not the amount of praise. It is the amount of useful information.
See how recommendation letters can support specific criteria in the real examples library, or work through the referee-planning worksheet in the Case-Building Toolkit.