A goal is one sentence and one number
“Ship faster” is not measurable. “Happier customers” is not measurable. “Go live with client A’s three features this quarter” is. Ontra’s goal record holds a number, a date, and linked sprints. The sentence stays in the customer’s language; the percent comes from the card. A goal that lives on a slide is invented in a meeting at month-end. A bound goal moves itself when the sprint closes.
OKR theatre exhausts teams because the tool and the work live apart. Pretty sentences are written at quarter-start, the sprint board lives elsewhere, Friday’s slide is a third truth. Ontra does not want you to open a second OKR tool. The goal is the top of the project plan. Features bind to milestones, tasks to features, sprints to the goal. Dual records scatter again; one tree stands.
Sprint rhythm, not a vague quarter
Split a quarterly goal into six sprints. Mark which feature counts toward the goal in each sprint. “We’ll look once a quarter” kills visibility. There is no surprise in week twelve; the drift is already on the report in sprint four. The sprint lock is a commitment. A card that does not count toward the goal bloats the sprint and does not move the number. That split hurts — and it should.
Close the month on the report screen. The meeting is for risk the report cannot show: the client slipped, a dependency broke, acceptance criteria are missing. Restating visible drift is a waste of time. Goal permissions (`goals.read` / `goals.update`) are part of fine-grained access. Not every member needs every goal. Holdings see totals in a separate view; a member only sees the goal on their project.
- A goal is one sentence, one number, one date
- Do not open a second OKR tool
- The goal updates when the sprint closes
- Split the quarter into sprints; do not wait for week twelve
- A card that does not count bloats the sprint and does not move the number
Software and agency, the same bind
On a software team the goal is “this module live”; feature, task, and PR sit on one tree. In an agency the goal is “this campaign live”; the brief binds to the task, delivery to the feature. In both, a slide cannot be the source. Rival brands can share a workspace; the goal grant keeps the company boundary. A designer only sees the goal of the assigned brand.
Bad goal examples
“Improve quality”, “strengthen communication”, “be more agile” — those are intentions, not goals. Intention can stay on a slide. A goal lands on the card. “Halve reopened work this quarter” is measurable because the repeat report exists. “Pull delay days under ten” is measurable because the delay list exists. Ontra’s report feeds the goal sentence; do not invent a second dashboard.
A measured sentence versus an unmeasured intention
Intention: faster, better quality, more agile. Sentence: this quarter client A’s three features live, reopened work halved, delay days under ten. Intention can stay on a slide. The sentence lands on the card. Ontra does not want a second OKR tool. The goal is the top of the tree. When the sprint closes the number moves. The month is not invented in a meeting.
A card that does not count toward the goal bloats the sprint. That pain teaches priority. “We’ll look once a quarter” produces a surprise in week twelve; the drift is already on the report in sprint four. `goals.read` is fine-grained. Not every member sees every goal. Holdings see totals in a separate view. In software the goal is a module live. In an agency the campaign is live. In both, a slide cannot be the source.
Bringing the goal sentence down onto the card
Intention stays on the slide: faster, better. The sentence lands on the card: three features live, repeats halved, delay under ten. A second OKR tool splits the record. The goal is the top of the tree. When the sprint closes the number moves. The month is not invented in a meeting. A card that does not count bloats the sprint. That pain teaches priority. We’ll look once a quarter explodes in week twelve; drift is on the report in sprint four.
goals.read is fine-grained. Not every member sees every goal. Holdings see totals in a separate view. In software the goal is a module live. In an agency the campaign is live. A slide cannot be the source. The report feeds the sentence: repeat report, delay list, close speed. Do not invent a second dashboard.
Bad goal examples are intentions. A good goal has a counterpart on the report. Improve quality is not measurable; halve reopened work is because the repeat count exists. Be more agile is not measurable; pull close speed under ten days is. Do not lock a sprint before writing the sentence. Otherwise the sprint produces work and the goal stays empty.
At month-end the goal screen replaces the meeting. The meeting is only for risk the report cannot show. This order is boring. What is boring is inventing the goal on a slide at quarter-end.
Sprint rhythm, not a vague quarter
Split a quarterly goal into six sprints. Mark which feature counts toward the goal in each sprint. “We’ll look once a quarter” kills visibility.
More posts
Product
When tasks and docs share one record, work stops scattering
List, board, and roadmap are views of the same record. A decision is made once and every surface shows the same truth. This piece explains how a pile of tools delays delivery and how Ontra binds one record.
Security
Fine-grained access lets the team breathe
Not everyone needs every project. Add project, folder, and list access on top of roles, and privacy becomes the default. In agencies and holdings the leak is rarely malice — it is the oversized shortcut “admins see everything.”
Setup
Finish the workspace in the 7-day demo, not the tool tour
No credit card. When the demo ends the app locks and data stays. Load real work in week one instead of touring screens: a project, roles, the live sprint, the docs tree, and the first report. A showroom tour does not pick a tool; a sitting record does.
