The recommendation letters that helped me were specific, not glowing
When I first thought about EB-1A recommendation letters, I assumed the goal was praise. Strong praise.
When I first thought about EB-1A recommendation letters, I assumed the goal was praise.
Strong praise.
Impressive praise.
The kind of language that makes a person sound exceptional.
But the letters that helped my petition were not useful because they were flattering.
They were useful because they were specific.
A vague letter can sound warm and still do very little.
"Pierre is outstanding" is nice.
"Pierre is a visionary leader" is nice.
"Pierre has made major contributions" is nice.
But those sentences do not prove much by themselves.
The better letter answers a different set of questions:
- How does this person know the work?
- What did they personally see?
- What role did I have?
- What happened because of the work?
- Which part of the petition can this letter help verify?
That is why I started thinking of recommendation letters less like endorsements and more like witness evidence.
The writer's relationship matters
The first thing a useful letter has to establish is the writer's relationship to the work.
Not just the writer's title.
Not just their reputation.
Their relationship to the facts.
In my petition, some letter writers could speak to Founders of the Future. Some could speak to Startup Dream Team. Some could speak to NoCode Drinks or the broader community-building work around technology and entrepreneurship.
The important question was not:
Is this person impressive?
The important question was:
What can this person verify that another document cannot fully show?
That kept the letters grounded.
If someone had a prestigious title but only knew the work from a distance, their letter could easily become decorative.
If someone had direct knowledge of the work, their letter could make the evidence clearer even if the sentence was less flashy.
Specific beats enthusiastic
The strongest parts of a letter usually do not sound like marketing.
They sound like a witness explaining what happened.
For example:
- the role I held
- the period when I held it
- the initiative the writer observed
- the audience or community involved
- the outcomes the writer can personally connect to the work
- why that work mattered in the field
Those details are harder to fake and easier to evaluate.
They also help the petition avoid a common weakness: unsupported adjectives.
"Major" is an adjective.
"80+ events impacting 4,000+ aspiring entrepreneurs" is evidence.
"Influential" is an adjective.
"An initiative launched at 10 Downing Street and continued through a multi-year program" gives the reader something to inspect.
The letter does not need to shout.
It needs to locate the fact.
Different letters can have different jobs
Not every letter needs to prove the whole case.
That was a useful realization.
One letter might help establish the importance of a role.
Another might explain the significance of a program.
Another might describe direct participation in an initiative.
Another might connect the work to the broader technology or entrepreneurship ecosystem.
Trying to make every letter say everything creates repetition.
It also makes the letters feel generic.
I wanted each letter to carry the facts that writer was actually positioned to verify.
That meant some letters were narrower than I expected.
Narrow was fine.
Narrow can be strong when the facts are concrete.
I had to protect privacy and accuracy
There is another reason to be careful with letters.
These are real people, real relationships, and sometimes private details.
In public content, I avoid naming recommenders unless I have already named them publicly or have explicit permission. The same principle shaped how I think about the petition material itself.
The letter should not expose unnecessary private information.
It should not inflate a relationship.
It should not ask someone to claim knowledge they do not have.
The cleanest version is simple:
This person knows this work because of this relationship.
They observed these facts.
Those facts support this part of the case.
That is enough.
The drafting mistake I would avoid
I would not start a recommendation letter draft with compliments.
I would start with facts.
Before writing prose, I would make a small fact sheet for each writer:
- How the writer knows me.
- Which initiative or period they can speak to.
- Which facts they can personally verify.
- Which criterion or final-merits point those facts support.
- Which details should stay private or anonymized.
Only after that would I draft the letter.
That order matters.
If you start with praise, you end up hunting for facts later.
If you start with facts, the praise becomes more credible because it is attached to something real.
The letter should make the exhibit packet easier to read
A good letter does not replace documentary evidence.
It connects it.
Media coverage can show that something was publicly visible.
An event page can show that something happened.
An award page can show recognition.
A letter can explain the human context around those documents: what the work involved, what responsibility looked like, and why the outcome mattered.
That is the job I needed letters to do.
Not decorate the petition.
Not make broad claims that no exhibit supported.
Make the existing record easier to understand.
The test
If you are thinking about recommendation letters, list your possible writers.
Next to each name, write one sentence:
"This person can verify ______ because ______."
If the blank is vague, the letter will probably be vague too.
If the blank is specific, you have the beginning of a useful letter.
The goal is not to find the person who can praise you the loudest.
The goal is to find the person who can describe the work most clearly.
Specific is stronger than glowing.