A recruiter once kept me waiting 22 minutes past the scheduled Zoom start time, then opened by saying the team “moves fast and doesn’t have a lot of process.” Both things were true. I left eight months later.
That gap between what you notice during an interview and what you act on is where most bad hires (on both sides) happen. The signals are usually there. The problem is knowing which ones actually matter.
Process signals that tell you how decisions get made
Scheduling chaos is one of the most reliable indicators I know. If it takes four back-and-forth emails to set up a 45-minute screen, that’s not an accident. That’s how the organization functions. You’re seeing it at its best, when someone is presumably motivated to make a good impression.
On the opposite end, extreme urgency is just as suspicious. “We need someone to start Monday” is occasionally a real constraint. More often it means the role has been open for months and the hiring manager is under pressure to close it, which means you’re inheriting someone’s problem.
The number of interview rounds matters too, but not the way most candidates assume. It isn’t that more rounds means more rigorous. Five rounds with no clear rubric and no feedback loop means nobody has authority to make the call. That ambiguity doesn’t go away after you join.
Things interviewers say that you should write down
The word “family” in a company description is, in my opinion, almost always a warning. Real families don’t fire you when revenue misses. What “family culture” tends to mean in practice is that boundaries are expected to be flexible, usually yours.
“Work hard, play hard” is functionally identical. I’ve never heard that phrase from a company where it turned out to mean “we take vacations and leave by 6.”
If an interviewer speaks disparagingly about a former employee or a previous team, pay attention to the specifics of what they’re saying, but pay more attention to the fact that they’re saying it at all. People who lack discretion about former colleagues rarely develop it.
High turnover that nobody can explain clearly is a different kind of red flag. Turnover happens everywhere. But “people tend to leave after a year or two” with no follow-on explanation is worth probing. Ask what the last three people in this role went on to do. If the interviewer doesn’t know, that’s information too.
When you don’t meet your actual manager
This one doesn’t get enough attention. If you get three rounds into a process and you still haven’t spoken with the person you’d report to, ask why. There are legitimate reasons (travel, internal approvals, timing). But sometimes the manager isn’t involved because the team doing the hiring and the team doing the managing are basically separate groups with different standards.
When you do meet the manager, notice how they respond when you ask technical or process questions. You’re not testing whether they’re smarter than you. You’re watching for defensiveness, vagueness, or signs that they’re selling rather than explaining. A manager who’s excited about the work will talk about it differently than one who’s trying to fill a headcount target.
According to LinkedIn’s Economic Graph research, manager quality is one of the top-cited reasons people leave jobs within the first year. That’s not surprising. What’s surprising is how rarely candidates ask direct questions about management style during the interview itself.
Compensation conversations that should make you nervous
Vague equity discussions are a real tell. Not because equity is necessarily worthless (it sometimes isn’t), but because a company that understands what they’re offering can explain it clearly. If you ask about preferred share structure, liquidation preferences, or the current 409A valuation, and the recruiter redirects to “we think this will be very valuable,” that’s not confidence, that’s deflection.
Artificial deadline pressure on an offer is common enough that it’s worth naming. “We need an answer by Friday” two days after you received the offer is a negotiating tactic, not a real constraint. Most companies will extend the deadline if you’re direct about needing more time. If they won’t, that tells you how they’ll treat you later when you need flexibility.
The Bureau of Labor Statistics Occupational Outlook Handbook gives you solid benchmarks for compensation by role and region. Using real external data when negotiating isn’t aggressive, it’s just grounding the conversation in facts. Companies that react defensively to that are worth noticing.
Questions that actually reveal things
Generic questions get generic answers. These tend to work better:
- “What’s the last thing this team changed its mind about?” Teams that never update their views on process, tooling, or priorities are usually stuck. Teams that can answer this immediately are usually healthier.
- “How does a typical engineer know what to work on next week?” This reveals planning maturity, autonomy, and whether there’s a real system or just whoever shouts loudest.
- “Can you describe how performance reviews work here?” Not “do you have them,” but how they actually work. Vague answers suggest the system exists on paper only.
- “What happened to the last two people who were in this role?” One left for a promotion elsewhere is fine. Two in a row left for undisclosed reasons is a pattern.
If you’re using an AI interview tool like Craqly to prep your responses, it’s also worth running your candidate questions through it beforehand. Hearing how your questions land out loud, and whether the follow-ups make sense, is a genuinely useful exercise before a real conversation.
Green flags are real too
Interviewers who push back on your answers respectfully. Honest answers about what isn’t going well. A process that finishes when they said it would. A hiring manager who asks “what would make this role not work for you” rather than only selling. These aren’t the norm everywhere, but they exist. When you see them, notice that too.
The instinct you have at the end of an interview, the one you sometimes override because the comp is good or the brand name is impressive, is usually picking up on something real. I’m not saying follow every gut feeling blindly. But the feeling is data. Treat it like data.