What the question asks, and what the Criteria page says
The prompt has three tasks, and a strong answer does all three:
- Show it. An example or examples of how you have demonstrated a commitment to improving the lives of others.
- Explain it. Why this is important to you.
- Extend it. How it is relevant to your future plans.
Most weak answers do the first task badly and skip the second and third. Plan the word budget so all three are visible.
What the Criteria page says
Gates reads this criterion broadly. Applicants define it themselves, through evidence of past, current and future commitment to the societies in which they will live and work. So there is no approved list. There is one test: is the commitment real, sustained and aimed at other people's lives?
Gates also selects needs-blind. Your background can explain why you act. It is not the answer.
Six forms the evidence can take, and the four tests of commitment
| Form | What it looks like | What to prove |
|---|---|---|
| Direct service | Teaching, care, health work, legal aid, relief | Who you served, how often, what changed for them |
| System or policy change | A rule, budget or process that now works differently | What changed, for how many, and what you did to cause it |
| Knowledge that changes practice | Research, data or a tool others now use | Who uses it and what they do differently |
| Voice for the excluded | Advocacy, representation, bringing people to a table | Who was heard that was not before, and the result |
| Sustained informal commitment | Caring for family, mentoring, supporting neighbours | Duration, cost to you, what went uncredited |
| Building opportunity | Scholarships, jobs, training, access | Who now has a chance they did not, and what they did with it |
Choose the form where your evidence is strongest, not the one that sounds most impressive. A modest, sustained commitment you can prove beats a large claim you cannot.
The four tests of commitment
A single event is activity. Commitment shows up in:
- Duration. More than a season. Say how long.
- Cost. Time, money, comfort or opportunity you gave up.
- Continuity. It carried on without applause, or after the first result.
- Direction. It points at the same people or problem now, and in your plan.
Score your candidate stories on these four. A story that passes three is strong. One that passes one is a line on a CV.
One example or several, and the structure and word budget
| Choose | When | How |
|---|---|---|
| One deep example | You have one commitment with real duration and measurable change | About 200 words on it, then the "why" and the future |
| Two examples as a pattern | A long commitment has taken more than one form | One main example (about 170 words), one short (about 70), and a sentence naming the pattern |
| Three or more | Almost never | A list reads as a CV and dilutes the evidence |
The structure and the word budget
| Move | Words (of 300) | What it does |
|---|---|---|
| 1. The person and the problem | About 40 | Start with a specific person, place or moment, not a statistic |
| 2. What you did | About 80 | Actions, duration, cost, who worked with you |
| 3. What changed | About 50 | The outcome for people, with exact numbers, and what they do now |
| 4. Why it matters to you | About 45 | A specific root: a moment, a person, a loss |
| 5. Where it stands now | About 30 | Current commitment, so it is not only a story from the past |
| 6. The future | About 55 | The link to your course and a dated plan, consistent with Question 2 |
Writing the "why it matters to me," about people, and if your work is not obviously charitable
This is the part applicants most often skip or fake. Three rules:
- Use a specific root. A memory, a person, a loss, a moment you saw something. "I am passionate about helping others" is empty.
- Keep it short. Two or three plain sentences. If it runs longer it becomes a confession and pushes out the evidence.
- Link it to action. The root should explain the choice you then made. "I chose plant pathology so that no one has to guess" connects the loss to the work.
If your root is painful, you can state it simply. The panel does not need detail. It needs to believe the commitment is real.
Writing about people with respect
- Give people agency. Write "we" for work done with others, and show who made decisions. The people you worked for are partners, not recipients.
- Credit others and be exact about yourself. Say what the team did, then what you did.
- Name roles, not people. Do not identify minors, patients or anyone who could be harmed by being named.
- Avoid rescue language. Phrases like "I saved" or "I gave a voice to" position others as passive.
- Show that the people shaped the work. A commitment that began because residents asked is stronger than one imposed on them.
If your work is not obviously charitable
Many strong applicants work in industry, research or government and fear they have no "service" to describe. The test is traceability: can you follow your work to a person in two steps? A software engineer who rebuilt a clinic's records system can say which nurses now spend an hour less on paperwork and what they do with it. A researcher whose data changed a regulator's standard can say who is safer. If you cannot name the person in two steps, the story is not yet evidence of this criterion.
The future clause
The prompt asks how this is relevant to your future plans. A good future clause has three parts: what you are doing now, what Cambridge adds, and the dated aim.
| Part | Example |
|---|---|
| Now | "I am training 12 extension officers to run the test without me." |
| What Cambridge adds | "A PhD would give me the virology to build a test that anticipates new strains." |
| The dated aim | "By 2032, a county-funded service reaching 40,000 households." |
Keep the aim honest. A role, a year and a scale you could defend. Do not promise to change a country.
Three worked examples (all invented)
Example A: direct service built from research (about 285 words)
In 2021 I met a farmer in Siaya who had pulled up her whole cassava field, a season's food for six people, after a neighbour said the leaves looked sick. They were healthy. The virus test that would have told her cost KES 4,500, more than she earned in a month. I decided that a test no farmer can afford is not a test.
For three seasons I ran the Siaya Cassava Testing Project from my own savings and weekends, then with a small grant. I adapted my thesis assay to cost KES 300 a sample, trained 14 volunteers from farmer groups to run it, and tested the fields of 1,420 farmers across 31 villages. Where the test found virus, we helped farmers pull only the infected plants and replant with clean cuttings from a certified nursery 40 kilometres away. In the first season 63 per cent of the farms we tested had no infection, so those families kept fields they would otherwise have destroyed. Over three seasons, participating farms reported harvests 31 per cent higher than their neighbours'. Several farmers now run the testing day themselves.
This matters to me because my family lost a harvest to disease when I was nine, and I remember what a missing meal looks like. I did not choose plant pathology to publish. I chose it so that no one has to guess.
I am now training 12 extension officers at the Lake Region Agricultural Research Institute to run the test without me. A PhD would give me the virology to build a test that anticipates new strains, and by 2032 I plan for a county-funded testing service reaching 40,000 households.
Why it works. It opens with a person and a decision, not a statistic. Duration, cost and continuity are all visible. The outcome is a change in households with exact numbers. The "why" is a specific memory in two plain sentences. The future is a role, place, year and scale, and it links to the PhD without repeating Question 2.
Example B: giving voice to a community (about 290 words)
After the 2020 port explosion, the families of my neighbourhood were told their homes would be rebuilt and their stories forgotten. Within a fortnight I started recording what residents wanted preserved: a bakery, a stairway where children played, a shrine at a corner. I did it because older residents asked me to, and I could speak Arabic and Armenian, the two languages of the street.
Over four years I led a volunteer team of nine students and residents to record 41 oral histories and map 112 sites of daily life. We sat with families as they told their stories, returned transcripts for correction, and left each family the right to withdraw anything. In 2023 the municipality adopted our map in its reconstruction plan, and 17 of the sites we documented were protected from demolition. Three elders who had been excluded from earlier consultations were invited to the planning table, and they remain on its advisory group.
This matters to me because my grandmother lost her street in 1975 and never again had anyone ask her what it was like. I saw that being asked is itself a form of repair.
I am now developing a training programme so residents of other districts can document their own streets without outside researchers. A degree at Cambridge would give me the comparative method to understand how displaced communities in other cities keep their memory. Afterwards I will found a documentation centre at a Beirut university that trains community historians and lends them equipment, with a target of 10 neighbourhoods by 2031.
Why it works. The commitment began because residents asked, which gives them agency from the first paragraph. Care shows in practice: transcripts returned for correction, a right to withdraw. The outcome is physical (17 sites protected) and human (three elders at the table). The future turns a personal project into a method others can run without her.
Example C: a quieter, sustained commitment (about 265 words)
My commitment to the people of my neighbourhood began when I was sixteen. A fire destroyed 400 homes on the hillside above Valparaíso, including my cousins', and for six weeks I carried bricks and fitted roofs with a volunteer brigade. I realised that much of what we were rebuilding had been unsafe before the fire.
At twenty-two I began something smaller and more regular. Every second Saturday for five years I visited homes in my district, checked walls and roofs against a simple structural list, and told families plainly what was and was not safe. I did 310 checks. In 38 houses I found a defect likely to fail in an earthquake, and for 29 of those families I helped find the money or volunteers to fix it. Neighbours I had never met began to call me before they built, and three of them now do structural checks themselves.
Nobody thanked me for most of those Saturdays, and that is part of why I did them. I do not believe in help that depends on being noticed.
These checks are why I became an engineer. They also showed me my limit: I could make 310 houses safer, but there are 60,000 in my city. What I need is research that proves a cheap retrofit works, and a national standard that does the work of 60,000 Saturdays. That is the PhD I am proposing, and the unit within Chile's housing ministry I will build afterwards.
Why it works. It is the pattern version: two forms of the same commitment, six years apart, tied by a sentence of reflection. Duration and cost are the evidence: five years of Saturdays without recognition. Honest scale (310 of 60,000) makes the PhD sound necessary rather than ambitious.
Weak versus strong, the test suite and common mistakes
| Weak | Why it fails | Strong |
|---|---|---|
| "I volunteered at an orphanage for a summer and learned the value of compassion." | One short event, and the lesson is about the applicant | "I tutored the same eight children every Saturday for three years. Six passed their national exams, and four now tutor younger pupils." |
| "I am passionate about empowering youth." | A value, not evidence | "After my father's business closed, I taught myself accounting and ran the household books at seventeen. Now I run free bookkeeping sessions for 40 market traders." |
| "My startup impacted thousands of lives." | Unprovable and untraceable | "Our app cut waiting time at 12 clinics from 90 to 35 minutes. Nurses told us they now finish by their shift end." |
The test suite
- The outcome test. Is the result a change in people's lives, not a count of events?
- The duration test. Does it last longer than a season?
- The agency test. Do the people you worked with make decisions in the story?
- The root test. Could someone else have written your "why"?
- The two-step test. Can you trace your work to a named role of person in two steps?
- The referee test. Can your Gates referee confirm the central story first-hand?
- The thread test. Is the future the same one as in Questions 2 and 4?
Common mistakes
- A list of activities with no outcome.
- A single short event described as a commitment.
- Rescue language that makes others passive.
- A "why" that is a value statement, not a root.
- No current or future commitment.
- Need as the argument, when Gates is needs-blind.
- Numbers you cannot explain.
- Identifying a beneficiary who could be harmed by it.
Checklist
I opened with a person, place or moment, not a statistic
I state duration, cost and what changed, with exact figures
The people I worked with have agency in the story
"Why it matters to me" is a specific root, not a value statement
It includes something current, not only the past
The future links to my course and matches Question 2
My Gates referee can confirm the central story first-hand
Every number is one I can explain at interview
It stays under 2,100 characters including spaces