Skip to main content

Reading a Signal

A signal's detail view is built to be read top to bottom in the order you'd naturally ask questions: what broke, how bad is it, which way is it going, and what do we do.

Written by Schae Lilley

Opening the detail view

Click any signal — on the Scan board or in the Overview list — and the detail view opens. The header shows the severity, the signal's tier, and its title, along with who owns the signal and an optional due date. Anyone on the team can set or clear the date, it shows on the signal's board card, and it turns red once overdue. When someone sets or changes the date on a signal you own, you're notified in Omega and by Slack DM. From here you can also jump to a Full analysis page for a deeper dive (covered in its own article).

The structure: problem → stakes → trend → action

The problem. The view opens with a plain-language statement of what's wrong, color-edged by severity so the visual cue matches what you clicked.

The stakes. Right under the problem sit the numbers that tell you how much to care: the estimated dollars at risk per month, how confident the diagnosis is, and the headline metric move. A caution appears here if the underlying data looks suspect.

Why it matters. A short narrative connecting the problem to the client's business — the context you'd use to explain it in a client meeting.

The trend. Small before-and-after charts for the key metrics involved, so you can see the movement rather than take the prose's word for it. The trend is labeled with the exact timeframe it covers — say, the last 7 days against the prior 7 — so there's no guessing what window you're looking at.

Cost of waiting. Collapsed by default — expand it to see what doing nothing is likely to cost. Useful when you need to make the case for acting now.

What to do. The recommendation itself — the specific action Omega suggests, open by default because it's the reason you opened the signal.

Recommended agent. The specialist agent (or multi-step plan) suggested to work the fix. If there are options, you can pick which one runs.

Below the fold: check the work

Everything above the divider helps you decide. Everything below it helps you verify:

  • Evidence — the verbatim source behind the diagnosis, such as a quote from a client call.

  • Classification — how the signal is filed: the business pillar and problem category it rolls up to.

  • How we got here — the diagnostic trail, expandable when you want to see the reasoning path that led to this conclusion.

Acting from the detail view

The footer holds your options: dig deeper in a chat with Iris, dismiss the signal, queue it, or run the recommended agent now. There's also a discussion section for comments and @-mentions. Each of these has its own article.

Tip: For a quick client-ready summary, you usually only need three pieces: the problem statement, the dollars at risk, and "What to do." The rest of the view exists so you can defend those three when asked.

Note: Dollar figures here are ranked estimates of monthly exposure, not booked losses — hover any dollars-at-risk figure for the definition.

Did this answer your question?