A lot of security products focus on what happens after something is already risky.
A file was detected.
A link was blocked.
A threat was found.
A scan completed.
That is useful, but we keep coming back to a different question while building Avao:
What should the user know before they take the action?
Before they install an extension.
Before they enter their email into a form.
Before they download a file.
Before they log in to a website.
Before they approve a browser permission.
Most risky digital decisions happen quickly.
The user is not sitting there thinking about threat models.
They are trying to get something done.
That’s why we’re thinking of Avao less as a traditional security tool and more as a real-time helper for everyday internet use.
The goal is not to scare people with warnings.
The goal is to give them enough context to make a better decision.
For example:
“This extension is asking for broad access.”
“This login page looks different from the real one.”
“This site may be trying to collect sensitive information.”
“This download source looks unusual.”
Small, clear feedback at the right moment can be more useful than a big report later.
The challenge is balance.
Too little feedback and the user is still guessing.
Too much feedback and the product becomes annoying.
That is the product problem we’re working through now:
How do you make security feel helpful in the moment, without slowing people down?
Curious how other founders think about this.
When your product gives real-time guidance, how do you decide what deserves an interruption?
Great question — I'm dealing with the exact same balance in my study planner app (Clutch). We send task reminders, focus timer alerts, and evening nudges, and the "interruption vs. helpful" line is razor thin.What's worked for us so far:1. Severity-based interruption. Only high-priority items (task is due NOW) trigger a sound + notification. Low-priority stuff (evening reminder if you haven't studied) is silent — just a badge or in-app banner.2. User-controlled granularity. We let users toggle three separate notification types (reminders, sound, push) instead of one master on/off. People who hate sound can keep visual alerts.3. Time-window awareness. We don't send anything after 10 PM or before 7 AM local time. Learned that the hard way — a 11 PM "time to study!" push is a great way to get uninstalled.4. Quiet mode after repeated ignores. If a user dismisses 3 reminders in a row, we stop pushing for that task and just show it in the dashboard. No point nagging someone who clearly isn't going to do it.The hardest part is that "annoying" is personal — what feels helpful to one user feels intrusive to another. Defaults matter a lot.Curious: are you thinking of letting users set their own "interruption threshold," or is that too complex for a security tool?