In Arthur Conan Doyle's "Silver Blaze," Sherlock Holmes solves the case by noticing something that didn't happen: the family dog never barked the night of the crime. A silent dog means the intruder wasn't a stranger — it means the dog knew whoever came. The clue was never a physical object at the scene. It was an absence nobody else thought to register as evidence.1
Ordinary attention fixates on what's present — what can be directly seen, heard, and pointed at — and passes straight over what should have happened but didn't. Greene's name for this class of clue: negative cues, information carried by the missing thing rather than the visible one.2 The bias makes sense mechanically: presence is a stimulus your senses register automatically. Absence requires actively knowing what should be there in order to notice it isn't — a comparison against an expectation, not a direct perception, which is a fundamentally harder cognitive operation.
For centuries, disease was understood almost entirely as something attacking the body from outside — a contagious germ, a cold draft, bad vapors in the air. Treatment meant finding something to counteract whatever external agent was doing the damage.3
In the early twentieth century, biochemist Frederick Gowland Hopkins, studying scurvy, reversed the frame entirely. What if the problem wasn't something attacking from outside, but something missing from inside the body? That reversal — looking for absence instead of presence — led directly to the identification of vitamin C, and reshaped the entire concept of what disease could even be.4 Centuries of medical attention had been fixed on presence. The breakthrough required looking the other way.
Greene extends the same move into commerce: the ordinary approach to a market looks at what already exists and asks how to make it better or cheaper — a presence-focused question. The harder, rarer move is to ask what need currently isn't being met at all — an absence-focused question, and correspondingly harder to answer because there's no existing product pointing at it.5 His practical suggestion: look at newly available technology and imagine an application nobody's built yet, aimed at a need people sense but haven't articulated. If the need is obvious enough to already be visible, someone's likely already working on it — genuine absence-spotting means looking where the obviousness hasn't kicked in yet.
Worth being precise about the mechanism in the fictional case, since it's the clearest illustration even though it isn't historical. Holmes didn't notice the silence directly — silence, by definition, produces no stimulus to notice. What he noticed was the mismatch between a model (dogs bark at strangers) and the evidence (the dog was reported not to have barked). The negative cue only exists at the intersection of a held expectation and a checked reality. Without the prior model, the absence of barking isn't even data — it's just nothing happening, indistinguishable from every other silent moment in a normal night.
The strategy has a failure mode worth naming even though Greene doesn't: not every absence is meaningful. Most silences are simply nothing occurring, not evidence of anything in particular. The strategy only works when the expectation being violated is itself solid — Holmes could trust "dogs bark at strangers" because it's a reliable generalization; Hopkins could trust "healthy bodies need specific nutrients" because prior nutritional science, even incomplete, supported the model. An absence checked against a weak or wrong expectation produces false signal, not insight.
Of the four shorthands named on the parent page, this is the one requiring the most prior structure to even attempt. Noticing an anomaly requires knowing the paradigm well enough to spot what breaks it. Noticing an absence requires something harder still: holding a complete mental model of what should be present, detailed enough that a gap in it registers as a gap rather than simply going unnoticed. Holmes needed to already know dogs bark at intruders. Hopkins needed a working model of what a healthy body requires. Absence-spotting is expertise-dependent in a way the other three shorthands aren't quite as strictly.
You're evaluating a situation, a market, or a body of evidence. Before cataloguing what's there, ask explicitly what should be there, given everything else you know, and check whether it actually is. The gap between expectation and reality is where this strategy's information lives.
You're looking at an existing product or explanation and your instinct is to improve what's already visible. Pause and ask instead what need isn't being addressed by anything currently on the table — not a better version of the existing thing, but the thing nobody's building because nobody's named the gap yet.
The Holmes case is fictional, used illustratively rather than as historical evidence — worth flagging plainly, since Greene cites it with the same weight as his factual cases. The Gowland Hopkins case is accurate history, and vitamin C's discovery is well documented.
Open question: Greene's business extension assumes an unmet need, once correctly identified, is straightforwardly actionable. But an absence can also mean a need doesn't actually exist, or exists but isn't solvable at acceptable cost — the strategy doesn't distinguish a genuine gap worth filling from an absence that's absent because nobody wants what would fill it.
This page pairs most directly with its sibling Anomalies, Not Paradigms — both describe noticing what a dominant frame fails to account for, but an anomaly is something present that doesn't fit, where an absence is something missing that should be there. Greene treats them as parallel but doesn't note that absence-detection is strictly harder, since it requires a complete model to notice a gap in, where anomaly-detection only requires noticing a misfit in something already visible.
The Evolutionary Hijack — Paul Graham's Viaweb success, covered later in this chapter, came partly from building toward a gap nobody else had named: a web-based application server when the entire industry's attention was fixed on desktop software. The insight the pairing produces: Greene's business advice here — look for the unmet need rather than the existing market — is exactly what Graham's case demonstrates in practice, again without Greene drawing the explicit connection between his own earlier strategic advice and his own later case study.
The Shadow — from Laws of Human Nature, this page argues everything repressed and denied doesn't vanish, it operates underground and leaks on its own schedule — a psychological claim about absence carrying signal, structurally close to Holmes's silent dog. The insight the pairing produces: in both cases, what's missing from the visible record — a bark that should have happened, a trait a person insists they don't have — is more diagnostically reliable than what's present, precisely because presence can be performed or explained away while a genuine absence is much harder to fake convincingly.
Sharpest implication. Competitive advantage built on this strategy is unusually durable, because most competitors are structurally worse at seeing it — the presence bias isn't a matter of effort or intelligence, it's built into how attention naturally allocates itself, so an absence spotted correctly tends to stay unnoticed by rivals for longer than a visible improvement would.
Generative questions.