Part one: choosing the episode
Selection matters more than writing. A well-chosen story writes itself in an afternoon.
Four filters
1. Can I name an exact number that moved? Not the size of the problem, the size of the change.
2. Did I decide things, or execute someone else's decisions?
3. Is it from work or voluntary activity? Chevening's committees warn specifically against school and university examples.
4. Did I have to bring people with me who did not have to follow me? Influence without authority outscores authority, because authority explains the outcome without reference to you.
Leadership does not require a leadership title
You do not need to have been a chief executive, director, manager or elected official. Leadership shows wherever you took responsibility, exercised judgement, moved people and produced a result: employment, research, public service, healthcare, technology, business, community work, civil society, volunteering, professional associations, creative industries.
Chevening is looking for a force multiplier: somebody who identified a problem nobody had claimed, mobilised whatever was available, and delivered something measurable.
Choose evidence, not titles
Weak: "I became president of X, team leader of Y and project manager of Z."
Strong: "When X emerged, I recognised that Y was blocking progress. I took responsibility for fixing it by doing Z, which meant persuading A and mentoring B. As a result, C changed."
The job title test. Remove your job title from the finished essay. Do your actions still demonstrate leadership? If not, rebuild around what you decided rather than what you were called.
Strong and weak candidates
| Strong | Weak |
|---|---|
| You spotted a problem nobody had claimed | You delivered a project you were assigned |
| You persuaded people outside your reporting line | Everyone involved reported to you |
| Something failed partway and you changed course | Everything went to plan |
| You can name what continued after you left | The work ended with the funding |
| The outcome was measured by somebody else | You are the only source for the result |
Size is not the filter
A well-evidenced pilot in forty villages beats a vaguely described national programme. Assessors are looking for causation they can trace to you, and small projects make that easier.
If your best example is imperfect
Use it, and say what did not work. An honest account of something that half-succeeded beats a polished account of something unverifiable. What you must not do is inflate a marginal role, because the interview will find it.
Part two: the three failures that cost the most marks
Treating a position as proof. Listing roles and responsibilities as though seniority were evidence. The question is how you gained influence over people who did not have to listen, how you kept it, and what came of it.
Defining leadership instead of demonstrating it. Many applicants explain what leadership means and never show themselves doing it. Spend no words on the concept.
Stopping at A. A competent situation, task and action, then a closing line about a great success that taught a lot about teamwork. Half the marks live in R and L.
Part three: naming, in this essay
Naming is not just a course-essay technique. It belongs here, and most applicants use none of it.
- Name the institution you worked through, however small. "The Katiba Midwives' Association" beats "a local partner".
- Name the official you had to move, with their title, because the title shows where in the system they sit.
- Name the thing you founded.
- Name the artefact that changed: the form, the guideline, the standard operating procedure, the dataset, by what it is actually called.
The rule: the name and what you did with it go in the same sentence. A name with nothing attached is a contact list.
| Unattached | Attached |
|---|---|
| "I worked with the Ministry of Health." | "I persuaded the Ministry's Quality Assurance Unit to adopt the form as its default." |
| "I have met the Registrar General." | "The Registrar General appointed me to the technical advisory group after the study." |
| "We partnered with a midwives' body." | "The Katiba Midwives' Association agreed to issue notification slips at delivery, which is how we reached 3,142 children." |
Do not name beneficiaries, patients, clients or minors. Use their role.
Aim for four to eight proper nouns in 300 words.
Part four: writing it, beat by beat
S. Situation, around 40 words
Open on the problem, sized with an exact number. Not on your biography, not on a quotation, not on a definition.
Variant one, the contradiction.
"A third of clinics in Molo District ran out of first-line antimalarials at least once last quarter, while the Central Medical Store held nine months of cover. This was never a supply problem."
Variant two, the failed assumption.
"Katiba Province distributes certified seed to 12,400 smallholders each season. Our own follow-up survey found that 5,900 of them never planted it. Nobody had asked why before buying another season's worth."
Variant three, the unowned gap.
"Molo District lost roughly one in five referred maternity patients between referral and arrival. The clinics assumed St Anselm's Hospital was tracking them; St Anselm's assumed the ambulance service was."
Variant four, lived experience converted immediately.
"I grew up in Sisan District, where fewer than half of children hold a birth certificate. I now run the research that measures why, and the answer is not that families do not want one. It is that the nearest registrar is 26km away and opens three days a week."
Variant four is the strongest opening available if it is true of you. It gives you standing to define the problem and it answers, without a word spent on it, why you will return home.
A stance is fine. A definition is not. If you open by describing a problem nobody had claimed and then claiming it, the reader already knows your philosophy of leadership without your having spent forty words announcing it.
Openings to avoid: "Leadership has always been my passion." "Since childhood I have wanted to serve my community." "According to the World Bank, developing countries face many challenges."
T. Task, around 30 words
What became your job, under what constraint, with what at stake. This is the most frequently omitted section, and without it every later achievement is ambiguously credited.
"Dr Kwame Antwi, our Director of Pharmacy Services, asked me to find the cause before the next procurement round: eight weeks, no budget for new systems, and a quarter of the district drug budget at risk if we simply reordered."
Stronger still when the constraint is human:
"As the only member of the unit with field experience, I volunteered to lead the diagnostic, over the objection of a manager who wanted us simply to order more."
Naming resistance here makes everything after it read as influence rather than administration, and it sets up the learning paragraph.
A. Action, around 110 words
The longest section and the one applicants underwrite. The test for every sentence: does it contain a decision?
Use the obstacle chain.
"Growers told me the Riverine Cooperative Union had abandoned the scheme twice over unpaid invoices, so I dealt with payment first, negotiating a fourteen-day settlement window with Sahel Agro Processing and putting it in writing. I piloted the new collection route with two cooperatives, weighing every delivery myself for the first month so disputes had an evidence trail. When the route held, I trained 22 cooperative treasurers to keep the same records, and linked the group to Molo Foods so that nobody depended on a single buyer."
Four decisions, each caused by a named constraint, each with a real institution attached. Compare "I coordinated with stakeholders, conducted capacity-building activities and facilitated market linkages", which contains no decisions, no constraints and no nouns.
Isolate the conflict. Zero budget. A manager who disagreed. A community surveyed four times with nothing returned. Incentives running the other way. An impossible deadline. Without a named obstacle, your result has nothing to be measured against.
Name the influence, not just the activity. Somewhere here you should have persuaded, negotiated, mobilised, mentored, convened, brokered, secured cooperation, resolved a disagreement, brought conflicting groups together or built consensus around a decision people disliked. Say who they were and what changed their minds.
Prioritisation under constraint.
"We had resources for a fifth of the district, so I targeted the twenty villages with the lowest baseline coverage rather than the ones easiest to reach."
The adaptation beat. The highest-value single sentence available to you.
"Our first training ran as a written manual across four evening sessions, and attendance collapsed after week two because the evenings clashed with market days. We rebuilt it as twenty-minute demonstrations at the collection point on non-market mornings, and completion went from a third to 21 of 22."
Never use the passive voice for your own work. "A project was launched" hides the only thing being assessed.
Keep the team visible while staying in "I". Attribute precisely: "I negotiated the settlement terms; the cooperative chairs ran the member meetings; between us we had a working route by March."
R. Result, around 70 words
Exact numbers first, then reach. Use the impact ladder.
"Stockouts in the pilot clinics fell from eleven to two the following quarter. The Central Medical Store adopted the reorder table across all 43 facilities, and it is now written into the district supply procedure."
Three rungs: the direct metric, the institutional adoption, the durable artefact.
Name the artefact. Not "our advocacy led to change" but "the late-registration procedure I drafted was issued as national guidance in 2025". The named instrument is the highest-value form of result there is, because it is the one an assessor could look up.
The micro-case. One sentence about one person, after the numbers, never before.
"One trader, refused three times for paperwork he had in fact submitted, was licensed within a fortnight of the new process."
If you genuinely have no numbers, never invent one. Use a change of state: a process changed, an institution adopted a practice, a collaboration was established, a template you wrote is still in use, a service reached somewhere it had not.
If the outcome was mixed, quantify what moved and state plainly what did not.
L. Learning, around 50 words
Two or three sentences answering both remaining sub-questions. A working principle you could apply tomorrow, drawn from the specific thing that went wrong earlier in your own essay.
Variant one, the operating rule.
"I learned that a shortage is often a signalling problem wearing the costume of a supply problem. I now interrogate the data a system already holds before anyone proposes buying anything."
Variant two, the reversal.
"My first design failed because I had solved for my own convenience rather than the nurses'. I now test every change with whoever will carry its cost before it goes near a policy document."
Variant three, the identity claim.
"I stopped thinking of myself as an engineer who writes reports and started thinking of myself as the person answerable for whether the water arrives. That changes which meetings I insist on being in."
What a leadership lesson can be about. Listening before deciding. Building trust before asking for anything. Adapting how you communicate. Delegating something you wanted to keep. Managing disagreement rather than avoiding it. Taking a calculated risk. Using evidence to settle an argument. Creating ownership in others. Changing strategy when the situation changed. Pick the one your story actually demonstrates.
Optional tie-in. One clause forward to the course or career plan: "That gap, being able to see the failure but not to measure it, is what took me toward this course."
Endings to avoid: "Leadership is not about titles but about service." "I learned the importance of teamwork and communication."
Part five: two full sample answers
Sample one, civil registration (299 words)
Bracketed labels show structure. Do not include them. All names anonymised and reinvented.
I grew up in Sisan District, where fewer than half of children hold a birth certificate and therefore cannot sit national exams. Nationally coverage is 38 per cent; in Katiba Province it is 30. The barrier is not demand. The nearest registrar is 26km away and opens three days a week. [S]
After my three-province study for the Institute for Public Administration, I decided to test whether registration could be brought to families rather than the reverse, with no budget and no mandate. [T]
I founded Kalima Registration Outreach and negotiated with the Katiba Midwives' Association to have their members issue notification slips at delivery, which removed the document families most often lacked. Our first attempt used a printed guide, and uptake stalled because it assumed literacy the mothers did not have, so I rebuilt it as a spoken checklist delivered at the six-week clinic. I trained 34 midwives, and because we could reach only a fifth of the province, I targeted the twenty wards with the lowest baseline coverage rather than the accessible ones. I then briefed Dr Amina Sesay, the Registrar General, on the results. [A]
We registered 3,142 children across 40 villages and cut median distance to a registrar from 26km to 4km. Dr Sesay's office appointed me to the Civil Registration Technical Advisory Group, and the late-registration procedure I drafted was issued as national guidance in 2025. One boy, refused twice because his mother's own certificate had lapsed, sat his exams last November. [R]
I learned that the families closest to a broken process know precisely why it breaks, and that my first design failed because I built for my own convenience rather than theirs. I now test every change with whoever carries its cost before it reaches a ministry. [L]
Sample two, financial services (297 words)
Sahel Trade Finance declined four of every five loan applications from women running market businesses, while our default rate on those we approved was 2.1 percentage points below the portfolio average. We were not managing risk. We were mismeasuring it. [S]
As a credit analyst with no authority over lending policy, I asked the finance director for three months and access to the archive to establish what our scoring model was actually rejecting. [T]
I re-scored 1,214 declined applications against subsequent repayment behaviour, traced through the Molo Market Traders' Association ledgers. The model was weighting registered collateral, which women in our market rarely hold in their own name. I rebuilt the score around twelve months of cash-flow data and proposed a pilot. The credit committee refused, so I ran it as a bounded exception on 200 loans with a lower ceiling, which they accepted because the downside was contained. When the cohort outperformed, I brought the arrears data to the committee rather than the argument, and I asked the Association's chair, Mrs Fatou Njie, to speak to the committee directly. [A]
Approvals for women-owned businesses rose 34 per cent across 8,140 active loans with no measurable increase in default. The cash-flow score is now our default model for unsecured lending below 500,000, and I was invited to present the method to the Bankers' Association working group on responsible lending, whose 2025 code adopted it as recommended practice. [R]
I learned that a committee which will not accept an argument will often accept a small, bounded experiment, and that the experiment then does the arguing for you. I now design every proposal with a version small enough to be approved by people who disagree with me. [L]
What both samples do. They open on a contradiction or a lived stake converted immediately into a statistic. They give the national figure and then their own province. They use exact numbers, not round ones. They name the institutions they worked through, the officials they moved and the artefact that changed. They contain a visible failure and a redesign. They record an appointment or invitation somebody else gave them. They close on a principle that could not have been predicted from the opening.
Part six: dos and don'ts
Do
- Choose one episode and go deep
- Open on the problem, sized with an exact number
- Convert lived experience into evidence in the same breath, if you have it
- State what became your task, the constraint, and what was at stake
- Name the institutions, officials and artefacts, with what you did attached
- Show at least three decisions that were yours
- Name whom you influenced and what changed their minds
- Include one failure and the redesign it forced
- Say what you could not afford to do and how you chose
- Put two exact figures at the result, plus what continued after you
- Record any appointment or selection somebody else gave you
- Close on a principle drawn from what went wrong
Don't
- Define leadership, quote anyone, or open with a proverb
- Use a school, university or student society example
- List the responsibilities your role carried
- Name people or institutions with nothing attached
- Name beneficiaries, patients, clients or minors
- Describe activity without decisions
- Write "we" where you mean "I"
- Use the passive voice for your own work
- Round a number you actually know
- Invent a figure because the essay feels bare
- End on a feeling
Part seven: common mistakes, and the fix
| Mistake | Why it costs marks | The fix |
|---|---|---|
| "I have always been a natural leader." | Assertion with no evidence. | Delete. Open on the problem with an exact number. |
| Opening with your upbringing, unconverted. | Spends the most valuable 40 words on unscored material. | Convert it into a statistic in the same sentence, or cut it. |
| A quotation or definition of leadership. | Chevening asks you not to. | Replace with your first decision. |
| No task paragraph. | Your contribution becomes unassessable. | Two sentences: what became yours, the constraint, the stakes. |
| A list of responsibilities. | Position is not influence. | Rebuild around people who did not have to follow you. |
| "I coordinated with various stakeholders." | No decision, no noun. | Name them, name the disagreement, name what you conceded. |
| Names with nothing attached. | A contact list, which committees warn against. | Put the name and what you did with it in one sentence. |
| No institutions named at all. | Nothing is checkable. | Add four to eight proper nouns, each tied to an action. |
| "We achieved" throughout. | Your contribution disappears. | Split it: "I negotiated X; the team then did Y." |
| Nothing goes wrong. | Reads as luck or invention. | Add the adaptation beat. |
| "The project was very successful." | Unverifiable. | Two exact figures, plus what continued afterwards. |
| "Over a thousand people." | Round numbers read as estimates. | 1,214. |
| "Our advocacy led to reform." | The result has no shape. | Name the guideline, form, dataset or procedure that changed. |
| Five examples in 300 words. | None lands. | Cut to one and spend the words on A and R. |
| Ending on "leadership is about service". | Wastes 50 marked words. | Draw the principle from your own adaptation beat. |
Part eight: checklists
While drafting
- One episode, from employment or voluntary work
- Opens on the problem, sized, with an exact number
- Lived experience converted into evidence, if I have it
- Task, constraint and stakes stated
- At least three decisions that were mine
- Whom I influenced, and what changed their minds
- One obstacle and the adaptation it forced
- Prioritisation under constraint, if I have it
- Two exact figures at the result
- The artefact that changed, by its name
- One thing that continued without me
- Any appointment or selection somebody else gave me
- Learning drawn from what went wrong
Before submitting
- All four sub-questions visibly answered
- Four to eight proper nouns, each attached to an action
- No beneficiaries, patients, clients or minors named
- No definition of leadership anywhere
- First person, team work attributed precisely, no passive voice for my own work
- The essay still shows leadership if my job title is removed
- Every figure real and every name spelled correctly
- Dates, employer and title match my work history
- At least one name here recurs in Essay 2 or Essay 4
- The gap this story exposes is the gap Essay 3 closes
- Within the word limit shown next to the box