Starting something

Most projects die in the head of the person who imagined them, usually because they were imagined too big. This lesson is the recipe for starting something real: small enough to finish, visible enough to matter, and measured well enough to prove it worked.

Pick something visible and finishable

Save the whole school is not a project. Cut the canteen's packaging waste this term is. A good first project has an end date you can see from the start, a result someone could photograph, and a size that survives exam weeks. Finishing matters more than scale: a finished small thing builds the trust and confidence that unlock the bigger one. An abandoned grand plan builds the opposite.

There is a mechanism behind why finishing beats scale, and it is worth naming: trust compounds. Every finished project, however small, becomes part of your record, and the record is what the next decision maker checks before saying yes, so each finish makes the next permission cheaper to get. There is also a mechanism behind why grand plans die, and it is not laziness. Motivation decays on a fairly predictable curve, high in week 1 and thin by week 6, so the honest question when sizing a project is not what could we do at our most excited. It is what will still get done in the flat middle weeks, when the project feels like admin. Size it for week 6 you, not week 1 you.

Find two friends and a number

One person is an opinion. Three people are a group, and as the last lesson showed, groups get answered. Two friends is the minimum viable team, and asking two people is not embarrassing. Then, before changing anything, measure. Count the bins, read the meter, time the thing, survey the year group. A before number is the most persuasive object you will ever hold, because it turns your project from a vibe into evidence, and it is the only way to know afterwards whether anything changed.

The before number does something subtle in the meeting itself. Without one, the conversation is about whether the problem exists, which is the hardest kind of conversation because it is opinion against opinion. With one, existence is settled in the first sentence, and the conversation moves straight to what should be done about it, which is a far easier place to agree from. The number also protects the person you are asking: a principal or a manager who says yes to you has to justify that yes upwards, and your measurement is exactly what they will use to do it. You are not just persuading them. You are arming them.

Ask for the smallest useful permission

Do not ask the principal to transform the school. Ask for the smallest permission that lets you start: may we weigh the canteen bins for one week, may we run a stall at lunch, may we put up three signs. Small asks get fast yeses, and each yes makes the next ask easier. Bring your before number to the meeting, propose something specific with a date on it, and say what you need from them in one sentence.

A worked example: the canteen bins

Here is the whole method on one small project, with illustrative numbers. Week 1: you and two friends get permission to count the canteen's rubbish, and you find it fills roughly three bags a day, 15 a week. That is your before number. Week 2: you take it to the canteen manager with one specific ask, a trial of paper trays instead of plastic ones every Wednesday for a month, which costs the canteen almost nothing to say yes to. Weeks 3 to 6: you keep counting, and Wednesdays now fill two bags instead of three. Week 7: you bring a single page to the manager and the principal: 15 bags a week before, a third less on trial days, and one ask, make it every day. Nothing in that sequence needed a committee, a budget or any permission beyond the smallest useful kind, and every step produced a number the next step could stand on.

Expect friction, celebrate the finish

Something will go wrong. A teacher will not reply, it will rain on the planting day, half the survey answers will be jokes. Friction is not a sign the project is failing. It is what doing anything real feels like, and pushing through one or two snags is exactly what separates projects that finish from projects that fade. When you do finish, measure again, share the before and after, and actually celebrate. The celebration is not vanity. It is how the next project recruits itself.

While you are in the friction, guard against one misconception: that resistance means the idea is bad. Institutions are slow with good ideas too, because the people inside them are busy, cautious and answerable to someone else, not because they are enemies of yours. An unanswered email usually means a full inbox rather than a rejection, so the correct response is a polite follow up a week later, not a retreat. Persistence, done courteously, is not being annoying. It is how the busy people learn you are serious.

Make it outlive you

One more step separates a project from a change: the handover. Most school projects die the day their founder gets busy, graduates or simply moves on, and everything quietly returns to how it was, which is why the final task of any project is to make yourself unnecessary. The fix is boring and it works. Write the job down in a few lines, recruit someone a year or two younger while the project is still visibly succeeding, hand it over publicly so the school knows who owns it now, and then actually let go. The case study at the end of this topic shows exactly this move, and it is the reason the result survived its founder.

Projects that work at school

• An audit: measure the school's waste, energy or food waste for a week and present the numbers. Cheap, fast and it arms every future project.
• A planting: trees or habitat on the grounds, with a plan for watering after the excitement fades.
• A repair event: a lunchtime stall where things get fixed instead of binned, with a tally of items saved.
• A campaign with one specific ask: not save the planet, but put a recycling bin in every classroom by term 3.

And that closes the topic. The first lesson gave you the honest hierarchy, the second the playbook for your own column, the third the multipliers, and this one the method for moving something bigger than yourself. The case study that follows shows all four lessons at work in a single project, with its numbers intact, so read it as a worked example of everything above.