The uncomfortable truth is that a stealth search isn’t a lighter version of hiring — it’s a harder one. You’re running a full executive search while pretending you aren’t, and most of the corners people cut to maintain the illusion of secrecy are exactly the corners that make searches fail.
Why founders reach for it anyway
The instinct is understandable. A VP is underperforming, or has plateaued, but firing them outright feels risky — you might lose institutional knowledge, spook the team, or simply not have a replacement ready. So the founder tells themselves they’re "quietly exploring options" while keeping the incumbent in place. It feels like taking action without taking the leap.
The trouble, as one widely read founder-community essay on executive hiring points out, is that most people who say they’re running a stealth search aren’t really running one — they’re running a half-committed open search with extra secrecy layered on top, which is the worst of both worlds. They talk to a handful of candidates their board already knows, do a handful of interviews, and call it coverage. Meanwhile the existing VP often finds out anyway, because the very act of casually reaching out to people creates the kind of small signals that are impossible to fully suppress.
The Starbucks case — and why it isn’t a template
The example that gets cited as proof stealth searches can work is Starbucks’ 2024 hire of Brian Niccol, then CEO of Chipotle, to replace Laxman Narasimhan. The board moved fast: it identified its top candidate, approached him directly and personally, without a broad recruiter-run process, and closed the deal in what looks from the outside like weeks rather than months. The market reacted immediately — Starbucks shares jumped 24% in a single session on the news, while Chipotle’s dropped as much as 13%. Board chair Mellody Hobson later said the decision had been under discussion for months and framed it as the board "owning the outcome," not reacting to outside pressure.
That story is often held up as evidence that stealth works. It’s more accurate to say it’s evidence that targeted works. Starbucks wasn’t sifting a pipeline — it had already decided who it wanted before the search began. That is a fundamentally different exercise from "we should quietly see who’s out there." A single-target approach and an open-ended stealth search share almost nothing except the fact that both happen out of public view.
Three paths, not one default
Framing this as "stealth or nothing" misses the real decision. There are at least three distinct strategies once a founder concludes a VP role needs a change, and they carry very different risk profiles:
| Path | Target clarity needed | Candidate coverage | Confidentiality risk | Typical failure mode |
|---|---|---|---|---|
| Stealth search | High — works best with 1–2 named candidates already in mind | Narrow, usually board/team network only | High — incumbent often finds out mid-process | Under-resourced search mistaken for real coverage |
| Open search | Low — can start broad and narrow over time | Wide, full market | Low — everyone knows a search is on | Incumbent friction and interim disruption, but process integrity stays intact |
| Top the role | Low — you’re hiring a boss, not a replacement | Wide, full market | Moderate — incumbent may sense the shift | Incumbent leaves anyway, or the new hire clashes with the person they now oversee |
The table makes the trade-off explicit: stealth only makes sense when target clarity is already high. If you don’t have named candidates in mind, you are choosing the option with the worst combination of coverage and confidentiality risk — a narrow pool searched with maximum friction.
A simple decision flow before you commit
Before starting anything quiet, it helps to walk through the sequence in order rather than jumping straight to "let’s keep this discreet."
flowchart TD
A[Assess: is the VP truly underperforming?] --> B{Do you already have 1-2 named candidates?}
B -->|Yes| C[Run targeted stealth approach]
B -->|No| D{Is the VP still performing, just capped?}
D -->|Yes| E[Top the role: hire their boss]
D -->|No| F[Move on now, run open search]
The checkpoint that matters most is the second one. If you can’t name real candidates, a stealth search isn’t a strategy — it’s a delay tactic dressed up as one.
Spotting the search that feels active but isn’t
The clearest sign a stealth search has quietly failed is that it feels like progress while producing none of the substance of one. A few tells:
- The "candidate pool" is entirely people your board or team already know socially.
- Interviews are happening, but at a fraction of the volume and rigor of a normal search.
- Nobody has written down what "done" looks like — there’s no defined target profile, just a vague sense that someone better exists.
- The current VP has started asking pointed questions, and the team is quietly speculating.
None of these on their own is damning. Together, they describe a search that consumes months of calendar time and goodwill while never coming close to matching what an open search would have produced in the same period.
Writing it down: a short decision memo
Because the biggest risk in a stealth search is self-deception, it helps to force the decision onto paper before committing real time to it. A short memo — even one page — should cover four things: the target (a named role profile, and ideally named candidates, not a vague sense of "someone stronger"); the constraints (what confidentiality level is actually required, and from whom); the risks (what happens if the incumbent finds out mid-search, and what the fallback is); and the exit criteria (a date by which you either have a candidate in hand or you convert to an open search). If you can’t fill in the target and exit-criteria fields with something concrete, that’s the memo telling you the stealth path isn’t ready yet.
If confidentiality really is non-negotiable
Sometimes secrecy isn’t optional — a pending fundraise, a restructuring, or a board-level succession may require it regardless of how the search is run. In that case, the more honest move is usually to hand confidentiality to a partner built for it rather than improvise it internally. What separates a serious confidential-search partner from a generic recruiter is largely about documented process, not charisma:
| Criterion | What to ask for | Why it matters |
|---|---|---|
| Disclosure protocol | A written document describing how candidates are approached without naming the role or company | Vague answers signal improvisation on every mandate |
| Off-limits map | A specific list of which companies and contacts are excluded from outreach | A broad client book can quietly narrow the real candidate pool |
| Named-partner accountability | The specific person running the search, and their track record on confidential work | Confidential work handed to associates tends to break protocol |
| Reference-handling rules | Written rules on who can be contacted, and when in the process | Careless references are the most common way confidential searches leak |
Retained-search relationships tend to suit this kind of work better than contingent or ad hoc recruiting, precisely because the incentives favor careful, long-term handling of both the mandate and the candidate rather than a fast placement. None of this guarantees confidentiality holds — no process can promise that — but it shifts the odds toward discipline instead of hope.
The bottom line
A stealth search isn’t inherently wrong; it’s simply expensive in a way that’s easy to underestimate. It demands the same target clarity, coverage, and time commitment as an open search, plus the added tax of secrecy. If you don’t have a named candidate or two already, and you can’t commit real recruiting bandwidth, you’re usually better off moving on openly or topping the role instead. And if confidentiality is genuinely required, the smarter move is rarely to improvise it yourself — it’s to find a partner whose disclosure discipline is something you can actually verify before you sign.
