An alert is Omnyra telling you something without being asked.
Everything else in the platform answers a question you went looking for. Alerts are the opposite: they exist because the problems that hurt a small business are usually the ones nobody thought to check on.
Severity — how quickly it needs you
Fix now — immediate action. Ignoring it could seriously hurt the business.
Important — a significant issue. Worth handling within the week.
Heads up — not urgent, but worth keeping an eye on.
For your information — no action needed. Something worth knowing.
The four exist so the top one stays meaningful. A system where everything is urgent gets ignored entirely, which is worse than no alerts at all, so Fix now is deliberately rare.
Category — what it is about
Cash — your balance, how fast you are spending, and whether you will run out.
Money owed to you — overdue invoices, slow payers, collection risk.
Bills you owe — upcoming due dates, late payments, vendor terms.
Spending — unusual or increasing expenses.
Revenue — declining income, lost customers.
Profitability — whether you are making enough on the work.
Customer risk — relying too heavily on one customer, or losing an important one.
Categories matter because they route. A cash alert and a profitability alert are usually acted on by different people and on different timescales, even in a business where that is the same person on different days.
Where alerts go
Alerts feed the action plan, which is where they are grouped by urgency and given somewhere to go. That is the screen to work from — the alerts themselves are the raw signal.
Why an alert is not an opinion
Alerts fire on thresholds crossed in your own data, not on judgement. That is a strength and a limit at once.
The strength is that an alert is checkable. Every one traces back to the figures that triggered it, and if you disagree, the numbers are there to look at.
The limit is that a threshold cannot know your context. A cash alert during a month you deliberately prepaid a year of insurance is technically correct and practically noise. Read alerts as "this is what the data says", and expect to overrule some of them.