--- title: How to switch on and read the daily brief url: https://tideswell.xyz/docs/analyst/daily-brief description: A written briefing each morning on what changed overnight, with severity, evidence, and a task you can create in one click. --- > Documentation index: fetch https://tideswell.xyz/llms.txt to discover every page before exploring further. # How to switch on and read the daily brief The **Brief** is a written briefing produced each morning: a headline, then a list of insights about what changed in your store overnight and what to do about it. It is the analyst working without being asked. Find it as **Brief** in the sidebar. ## Before you start [#before-you-start] Three things have to be true, and the page tells you which is missing. The **Brief** module has to be switched on, per . You need store view access. And the **Daily analyst brief** automation has to be switched on, per . Until it has run at least once, the page says there is no brief yet and that it runs each morning once enabled. ## What is in a brief? [#what-is-in-a-brief] A headline, then insights. Each insight carries: * **A severity**, shown as a coloured chip and a left rail: **Act** in red, **Watch** in amber, or **Info**. Act means it needs a decision today. * **The detail**, in plain English. * **The evidence** behind it, as bullets. * **A recommendation**, where there is one. When the brief covers both, insights are split under **Store** and **Marketing** headings. At the foot sits a confidence note, the date the brief is for, and which model wrote it. ## How do you turn an insight into work? [#how-do-you-turn-an-insight-into-work] Press **Create task** on an **Act** or **Watch** insight. The normal task form opens with a suggested title and description already written. You still pick the department and anything else, exactly as with any task, so nothing is created behind your back. See . **Info** insights carry no suggested task, because they are context rather than something to do. This is the feature that makes the brief worth reading rather than skimming. An observation you act on in one click is worth more than a longer report you file. ## How should you read the severity? [#how-should-you-read-the-severity] **Act** is the only one with a deadline attached. If you read nothing else on a busy morning, read the Act insights and create tasks from them. **Watch** is worth a glance and often worth a task with a date a few days out. **Info** is background, and skipping it on a busy day costs nothing. ## What is the confidence note for? [#what-is-the-confidence-note-for] Declaring what the brief could not see. A brief is written from whatever data sources were reachable that morning. When one was unavailable, the brief says so rather than quietly writing a thinner briefing that reads just as confident. Read it before acting on a surprising insight, the same habit as . A dramatic claim written while a data source was down deserves a second look. ## How do you read an older brief? [#how-do-you-read-an-older-brief] **Earlier briefs** holds the previous 30, and expanding one shows it in full. That archive is more useful than it looks. Reading back a fortnight tells you whether a problem is new or whether the brief has been flagging it for two weeks while everybody skimmed past. ## Where do you switch it on and choose the model? [#where-do-you-switch-it-on-and-choose-the-model] On the **Automations** screen, where the **Daily analyst brief** card carries its own switch, a **Run now** button and a **Brief model** setting. The model label shown on a brief is the model that wrote **that particular** brief, so changing it does not rewrite history. See . Generating one on demand lives with the automation rather than on this page. ## Who else finds out when it is ready? [#who-else-finds-out-when-it-is-ready] Anyone set to be notified. A note that the brief is ready lands in the Inbox, and it can email as well. Choosing recipients is .