Press velocity, decoded: reading the 0-to-40-articles-in-72-hours curve
A product going from zero to forty articles in three days looks nothing like steady background press. How rate-of-change on coverage volume separates a real breakout from a one-day spike.
Zero to forty is the headline. A product launches, a press release goes out, and within seventy-two hours forty outlets have covered it. That number gets screenshotted into a pitch deck and shared in a Slack channel within the hour. On its own, it's close to useless. The same seventy-two-hour spike shows up in front of a product that's flatlined by week three and a product that's still compounding six months later. The volume looks identical. What comes after it doesn't.
Launch Sentinel doesn't score coverage by how many articles land in the first three days. It scores the shape of what happens after, day over day, measured against the product's own baseline rather than an arbitrary threshold. Rate of change instead of raw count is the entire argument of this piece. Almost everything in the corpus collapses into three repeating curves, and only one of them is worth acting on.
Volume tells you it happened. Velocity tells you what it means.
Attention online decays fast, and it decays on a predictable curve. Fang Wu and Bernardo Huberman's 2007 study of a million Digg users found that collective interest in a novel story follows a stretched-exponential decline with a characteristic timescale close to one hour. A 2014 study by Carlos Castillo and colleagues, presented at the ACM's CSCW conference, went further: the eventual shape of an online story's traffic, whether it fades, holds, or rebounds, is largely set within the first ten to twenty minutes of social reaction. By the time a comms team is celebrating day-three coverage numbers, the shape that actually matters was decided days earlier.
That's the uncomfortable part for anyone reading press coverage as a scoreboard. Forty articles in three days can mean a product just cleared the noise floor for good. It can also mean the noise floor is the only thing that ever moved. The count alone can't tell the two apart. The curve can.
Three shapes, one week-five test
Pull enough of these curves out of the corpus and they collapse into three repeating shapes. We call them Spike, Sustain, and Silence, and the divergence between them shows up well before a full news cycle plays out.
Illustrative composite: three archetypal coverage-velocity shapes drawn from patterns Launch Sentinel's press-tracking sees repeat across the corpus, not a single product's real time series.
For the first three days, Spike and Sustain are indistinguishable. Both climb from nothing to roughly forty articles a day. The divergence shows up in the two weeks that follow, when Spike falls back toward zero and Sustain settles into a steady background rate and stays there. Silence never gets the initial climb at all, no matter how hard the launch was pushed.
What each shape predicts
Match the three shapes against outcomes we've already written about and the correlation holds. The twelve breakout products Launch Sentinel confirmed this quarter carry the Sustain shape in their press-velocity history: an early climb that never fully retreats. Spike alone shows up constantly and rarely means anything by itself. It's the shape behind a funding announcement or an app-store feature, sometimes just a single thread that traveled further than the product did. Silence is one of the quieter precursors behind the products we've watched flip from ALIVE to dead: a launch that never caught, followed by months of flat signal score and a changelog that eventually stops updating.
Forty articles in three days is loud. What predicts anything is whether someone's still writing article forty-one three weeks later.
Reading it before the plateau, or the crash, is obvious
None of this requires waiting until day thirty to find out which shape is in front of you. The slope over the first nine to eleven days already separates Sustain from Spike with reasonable confidence: Sustain's daily count stops falling and flattens, while Spike keeps dropping past the point where a plateau would have shown up. Silence, the easiest of the three to call early, simply never produces a slope worth measuring in the first place.
None of the three shapes is rare, and Silence is by far the most common response to a launch push. What's left splits between a handful of real Sustain curves and a much larger set of Spikes that read as one-day stories in hindsight. The corpus is unforgiving about the base rate: twelve confirmed breakouts against 79,573 tracked products this quarter, and press velocity that holds its shape past week two is one of the earliest tells that a product might be headed toward that list rather than toward the 353 that quietly went the other way.
Watch your own market's press curve
Launch Sentinel tracks day-by-day coverage velocity across 79,573 products in 18 categories, and flags the Sustain shape before the rest of the market calls it a breakout.