It is tempting to think of automation as the same job search, just faster. The risk is not speed. It is the judgment that used to happen before each send, quietly removed.
A safe tool is defined by what it will not do.
Lines worth refusing
- Writing again after a reply. A hard stop on any real response — yes, no, or "not now" — is not a preference. A tool that keeps going after a no is damaging the company for you, without you in the room.
- Contacting a company you blocked. Current employer, a competitor, a place you cannot risk. The block has to hold at the moment of send, not only as a filter on the job list.
- Inventing a contact when there isn't one. A small team with nobody sensible to write should come back as "no one to reach," not as a message to the closest name. Always finding someone is a way of looking complete.
- Spending your account to look busy. No weekly pace, no variation, executives on day one: that trades your LinkedIn standing for a dashboard that looks active.
- Knowing more than a stranger should. A referral note does not need a dossier. It needs the role, a real reason, and a small ask.
Automation should make the careful version easier. It should not make the careless version faster.
The test
Before a send, a person would ask: did they already answer, is this my own company, is this the right level, am I sure of the address, can I still use this account next month?
If the software cannot answer those, it is not saving you time. It is spending things you cannot get back.