Jonathan Ingles: Everett Rogers synthesized six thousand studies and published it in 1962. Six thousand. And the conclusion was essentially — technology adoption follows a curve. Every time. Slow, then explosive, then nothing. And that finding has survived sixty years of scrutiny.
Ben Okonkwo: Descriptively, yes. The shape shows up everywhere.
Jonathan Ingles: Descriptively. That's the word doing the work, isn't it.
Ben Okonkwo: Hm. So — okay — think about it like this. You're watching a wave from the beach. You can see the shape perfectly after it's broken. What you can't tell is how tall the next one will be, or when it'll hit, or where it'll reach on the sand. The S-curve is the shape of the broken wave. We know it in hindsight. Landline phones, nearly a century to saturate. Smartphones, roughly a decade. The pattern is real. But knowing the pattern doesn't mean you know where you're standing in it right now.
Jonathan Ingles: And that's the sleight of hand. Rogers documented this. Accurately. And then — frankly — an entire industry of forecasters took that descriptive framework and started presenting it as predictive. Which it isn't.
Ben Okonkwo: Right. The S-curve tells you what happened. Not what's happening.
Ben Okonkwo: But even that framing — 'the pattern is real' — I think most people stop there and miss what's actually producing the shape. Because there are three separate engines running. Information diffusion first — word-of-mouth, observational learning, people watching their neighbor use something and deciding to try it. That's what Rogers is really documenting in those six thousand studies. Early phase is almost entirely social copying.
Jonathan Ingles: Right, but that can't be the whole mechanism. Otherwise every technology would accelerate at the same rate.
Ben Okonkwo: Exactly — so that's where the second engine kicks in. Network effects. A telephone is worthless if nobody else has one. The value compounds as adoption grows, which actively lowers the barrier for the next cohort. That's what drives the steep acceleration phase — the inflection point becomes self-sustaining because utility itself is climbing.
Jonathan Ingles: Critical mass. Once you hit it, the curve basically runs itself.
Ben Okonkwo: Right. But then — and this is the part I think actually explains the compression data in that 1860-to-2019 household dataset — the third engine is complementary infrastructure. Electricity didn't just need wiring. It needed appliances. The network and the ecosystem had to co-develop, which... I mean, that takes time. Physical time.
Jonathan Ingles: The last-mile problem. Landlines took nearly a century precisely because you needed physical copper to every dispersed household. Smartphones had a cellular network that already existed. Two completely different constraint profiles.
Ben Okonkwo: Which means Rogers' five segments — innovators at two-and-a-half percent, early adopters at thirteen-and-a-half, and so on — they're sitting on top of all three of these engines simultaneously. But the taxonomy treats each person as making an individual decision. And once network effects are doing the heavy lifting, that assumption starts to feel... actually, no, it doesn't just feel odd — it might be structurally wrong.
Ben Okonkwo: So that's exactly where Christensen lands, actually. The 1992 paper — disk drives — managers inside the industry, with real-time sales data, still couldn't tell where they sat on the curve. Not because they were incompetent. Because the curve itself doesn't... I mean, it doesn't announce its own inflection point while you're living through it.
Jonathan Ingles: Wait. Insiders couldn't read it in real time?
Ben Okonkwo: That's his finding. And his distinction is important — component S-curves versus architectural S-curves. A component technology hitting its ceiling gets masked when someone reconfigures the whole architecture. Two separate curves, moving independently. Managers were tracking the wrong one.
Jonathan Ingles: Okay, so if insiders can't locate themselves on the curve with actual data in hand — what does it mean when a policy brief says we're 'early on the S-curve' for some energy transition? That's not analysis. That's a political claim wearing a framework.
Ben Okonkwo: Right, and the Brendon Tankwa INET Oxford paper names this directly — six recurring errors, and near the top is exactly that: projecting current trajectory to a confident end-state the data doesn't actually support. The ceiling isn't in the curve.
Jonathan Ingles: That's the laundering move. 'Inevitable' because S-curve — but the curve never said inevitable. Someone added that.
Ben Okonkwo: And the compression data makes it worse, because — ChatGPT, three months to a hundred million users. Rural electrification, decades. Bank of America cites these side by side as proof of acceleration. But those aren't commensurable phenomena. One has zero marginal cost, no last-mile problem. We might be calling two completely different processes by the same name and then forecasting the second using data from the first.
Jonathan Ingles: And Geoffrey Moore names the exact moment where that breaks down. The chasm — the gap between early adopters and the early majority — that's where most technologies just... die. Quietly. And the S-curve, the one everyone's citing, it's built from the survivors. The ones that crossed. So the framework is, by construction, a portrait of success. It doesn't contain the failures.
Ben Okonkwo: Hm. Survivor bias baked into the empirical foundation.
Jonathan Ingles: Which means — look — it's validated across railroads, electricity, telephones, the internet, smartphones, yes. Robust descriptively. But the policy brief that says 'trust the curve' is drawing on a dataset that, by design, excluded every technology that stalled in the chasm. I don't know what to do with that, actually. I keep turning it over. The best tool we have for escaping linear thinking about adoption is... honest about its own limits in a way that nobody using it seems to be.
