Hey everyone, Neil here. You're reading High-Signal Hiring. Hiring systems from 20+ years of global recruitment experience and 500+ technical hires. Zero noise and instantly actionable.
Last issue, we looked at the most underused sourcing channel in startup hiring. Rehiring former colleagues. 35% of tech hires in early 2025 were boomerangs, and the ones who left well are often your fastest path to a senior seat.
This week, I want to push back on something a lot of founders are being sold right now. AI-first interviewing.
You've seen the tools. HireVue, Humanly, Apriora, a dozen others. They promise to screen hundreds of candidates for you so you can focus on the best ones. The pitch is efficiency. The problem is that the candidates you want most are the ones hanging up.
You'll learn why AI interviewers filter out your top of funnel, where AI in hiring is still worth using, and the four-stage loop (across three interviews) a five-person team can actually run without one.
Not a subscriber yet? Sign up here
| What the candidates are telling you
A Reddit thread on AI interviews hit 17,600 upvotes in a matter of days. 64% of the comments linked AI-only interviews directly to poor company culture. Fortune ran a piece where candidates said they'd rather stay unemployed than take another AI-led call. Slate covered a candidate who showed up in a suit, only to be told she'd be talking to a bot. Her response, paraphrased. I dressed up for this?
This isn't a vocal minority. It's the dominant sentiment in the engineering job market right now. And it matters because of who's saying it.
The engineers we've been profiling across Issues 13 through 18, the AI-capable senior crowd with multiple approaches a week, have the leverage to be picky. They're not going to burn 30 minutes talking to a chatbot when they have three live offers on the table. They'll ghost the process and move on.
The candidates who sit through it anyway are the ones without options.
You're not screening. You're filtering in reverse.
| The culture signal is the product
When a candidate's first interaction with your company is a bot, you've told them exactly what you think of their time. Before you've explained the role. Before you've pitched the mission. Before they've met a single human. That's the signal.
Hiring is a sale in both directions. You're closing the candidate as much as they're closing you. The founders who win at hiring in 2026 treat every touchpoint as a recruiting moment. The founders who lose hand the first one to a bot and wonder why offer acceptance rates keep dropping.
Google and McKinsey have both started reintroducing mandatory in-person rounds. These are companies with unlimited AI budgets choosing to move the other direction. They've seen what happens when the top of funnel goes cold, and they're correcting course.
| Where AI in hiring is still useful
I'm not anti-AI in hiring. I use AI every day in my own work. The question is where you deploy it.
AI works well for the back-office parts of hiring. Sourcing candidates at scale. Enrichment and research. Scheduling. Note-taking after a live call. Rubric scoring against structured criteria. Pulling themes out of interview transcripts so you can run sharper debriefs (Issue 6).
All of that is AI doing work you used to do manually. None of it is AI replacing the conversation.
AI should make your interview process faster and more consistent. It should not be the interview.
| Four stages, three interviews, zero bots
For a 5-person team, you don't need seven rounds. You need four stages across three interview sessions, plus a final async stage to close.
Interview 1: Founder screen, 20 minutes
You, live, on video. Just you and the candidate. No formal script. You're looking for motivation, communication, and whether there's a genuine fit with the role. 20 minutes is plenty. If there isn't a signal in 20 minutes, there isn't one.
Interview 2: Mission and Depth, 60 minutes
Run with whoever on the team can assess the actual work. Share your 90-day mission in three lines and see how the candidate reacts. Then pick one real problem from that mission and go deep. You're testing how they think, not what they know. Orchestration, verification, system trade-offs. Not LeetCode puzzles. I broke this framework down in detail in my recent guest post with Gregor on Engineering Leadership.
Interview 3: Alignment, 45 minutes
A conversation with one or two people they'd work with day to day. No new technical bar. Working style, decision-making, ambiguity tolerance, expectations of founder involvement. This surfaces mismatches before you hire, not after.
Stage 4: Reference and close
Two backchannel references. A final conversation with you to handle comp, start date, and any remaining concerns. Close them the way we covered in Issue 4.
Three interview sessions. Four stages. Total candidate time under three hours. Total calendar time one to two weeks if you run it tight. No bots anywhere in the loop.
| What to do this week
Pull up your current hiring process and list every candidate touchpoint. For each one, ask two questions:
Is this a moment where the candidate is being sold on us?
Would I be happy if my best friend went through this step?
If the answer to either is no, that step is filtering backwards. Start there.
The best candidates don't need you. You need them. Every time you replace a human with a bot in the early stages, you're telling them you haven't worked that out yet.
Next week: the tactical playbook on AI interview fraud. Cluely, stealth earpieces, proxy candidates, and what small founders can actually do about it without an enterprise security budget.
Cheers
Neil