Feel the mechanism before naming it: you know the EV maths, approve of the future, and will still buy one more petrol scooter. Why? Because the petrol habit is endowed (it feels owned, and losses loom larger than gains), the new thing demands behaviour change (charging routines, range planning), and its benefits are probabilistic while its costs are certain and immediate. Gourville's '9x problem' formalises it: consumers overweight the incumbent's benefits about 3x, developers overweight their innovation's benefits about 3x, so the innovation must feel roughly nine times better, or ask for dramatically less change, before the switch feels fair.
Rogers' five attributes, run in reverse, become the diagnostic checklist. Relative advantage: is the gain big, or merely real? Compatibility: how much existing habit, infrastructure, and identity must the buyer abandon? Complexity: can the value be understood in one sentence, or does understanding itself cost effort? Trialability: can they test without commitment? Observability: can they see others benefiting? Every stalled innovation is failing a specific subset of these, and 'the market isn't ready' is almost always a refusal to name which.
Each barrier maps to a lever, which is what makes this design rather than fate. Advantage gap → sharpen the gain or shrink the ask (target the niche where the advantage is 9x, not the mass where it's 1.2x). Incompatibility → reduce the behaviour change, not the technology (fit the old habit's shape, as UPI did by digitising the exact gesture of handing money over). Complexity → simplify the story before the product. No trial → sachets, test rides, free tiers, return windows. Invisibility → engineer visible usage: distinctive design, public rituals, referral moments.
The chasm is where barrier work concentrates, because the customer type flips. Early adopters buy your vision and forgive gaps. The early majority buys a whole product: device plus charging plus service plus resale value plus 'people like me already did this'. Moore's crossing strategy is barrier removal by concentration: pick one beachhead segment, build the complete whole product for it alone, harvest its references, then widen. Spreading thin across segments rebuilds every barrier in every segment simultaneously, the default way good innovations die.