← All guides

Best practices

How to Write Achievements Using the XYZ Formula

Most resume bullets fail the same way. They tell the reader what you were near, not what you changed. The XYZ formula is the cheapest fix in resume writing: accomplished X, as measured by Y, by doing Z. A result, a number, a method. Get all three into one sentence and the bullet starts doing actual work.

The three parts, in order of difficulty

X, the result, is usually easy. You remember what happened. Y, the measure, is where most people stall, because nobody wrote the number down at the time. Z, the method, is the part people skip entirely, and it is the part that proves the achievement was yours. Anyone can be in the room when a metric moves. The method is your evidence that you moved it.

Take "improved website performance." It sounds fine and says nothing.

The reader now knows what changed, by how much, and what you personally did to cause it. That last part matters more than people think, because the method is where your skills live.

The same transformation, three more times

Finding numbers you never wrote down

Most people assume the number is gone. Usually it is just somewhere other than memory, and a half hour of digging turns three vague bullets into three defensible ones. Worth checking, in rough order of yield:

• Performance reviews and promotion documents, written back when the details were fresh and someone else was motivated to record them
• Old messages to your manager, where you almost certainly reported the result the week it happened
• Ticket and project trackers, for counts you can defend: bugs closed, stories shipped, incidents handled, tickets deflected
• Dashboards and reports you used at the time, including screenshots buried in old decks
• Your own calendar, for how long a project actually ran and who was in the room

When none of it survives, arithmetic will do. If you automated a task that took thirty minutes a day and ten people did it, that is roughly 1,250 hours a year. Say roughly. An estimate you can explain out loud beats a precise number you cannot.

When the work resists numbers

Some roles genuinely do not produce clean metrics. Research, design, compliance, policy, and early-stage anything all suffer from this, and the usual advice to "just quantify it" is useless in the face of it. The measure does not have to be a percentage. Scope counts: how many teams, markets, or systems. Frequency counts: weekly, for two years. Adoption counts: how many people used the thing after you shipped it. Avoided cost counts, if you can name what would otherwise have happened.

So does influence, as long as you name the decision it changed. "Research that killed a planned $400K integration after 12 customer interviews showed nobody wanted it" is a fully measured achievement without a single percentage in it.

Two habits that keep the formula honest

First, if you cannot quantify precisely, estimate and say so; "roughly 20%" beats silence, and you can defend an estimate in an interview. Second, use the strongest verb that is still true. Led beats helped with, but only if you actually led. Interviewers pull on these threads, and a bullet you cannot back up costs more than a modest one.

One achievement per line

Resist the urge to pack two accomplishments into one bullet. Each achievement gets its own sentence, its own number, its own method. Cramming dilutes both, and a reader skimming for evidence will find the seam and stop.

A useful length test: if the bullet runs past two lines on the page, it is usually two achievements sharing a sentence. Split it and give each one its own Y and Z. Two strong bullets always beat one crowded one, and they give a hiring manager two separate things to ask you about.