Attack on many fronts
Marketing gets a doctrine: attack from multiple fronts, because one channel isn't enough. The Spanish French-progress shorts go on YouTube; separately, brain-rot reels in English go on Instagram, funded by the free Veo credits from the Gemini Pro plan. The load-bearing insight is that each platform rewards a different format: Instagram photos haven't gone viral since before Meta bought it a decade ago, but reels do, so posting the wrong format on the right platform is wasted effort. The product half is the visual-exercise stock, pre-generated dense images per topic (starting with A1) rather than generated on the fly, both to control cost and to give the user and Severo creative fuel.
// trace: where this idea came from
- ↳ video diary @ 6:54 (las fotos no se hacen virales en Instagram; los reels sí)
- ↳ video diary @ 7:31 (reels brain-rot con los créditos de Veo)
- ↳ Entry 241-2: The bottleneck moved (la falta de exposición que aquí se ataca)
The visibility problem gets a strategy: attack marketing from several fronts at once, because a single channel makes it hard ▸ 6:24. Front one is the Spanish shorts of their French progress, on YouTube. Front two is new: brain-rot reels of Severo, in English, on Instagram, funded by the thousand free Veo credits the Gemini Pro plan includes ▸ 7:31.
The insight underneath is that platforms reward different shapes, and posting the wrong shape wastes the platform. Instagram photos don’t go viral, they haven’t since before Meta bought Instagram roughly a decade ago; it isn’t built for photos to spread ▸ 6:54. What travels there is reels, so their Instagram, which so far held only decent photos that were never going to move, needs video to earn reach. It’s the same lesson that took them off social media, existing for the record not the feed, now inverted into using the feed deliberately: know what each channel amplifies before you feed it.
cada plataforma premia un formato distinto →
The product half of the day is the visual-exercise stock. Rather than generate images on the fly, where cost scales with every user even on the cheap Nano Banana model ▸ 5:10, they’ll pre-generate one or two dense, detail-rich images per topic, starting with A1 ▸ 1:31, a market plaza full of food, a family tree, so the picture doubles as creative fuel: the user invents sentences from it, and Severo reads it as input to generate creative exercises ▸ 2:38. The open problem Julia inherits is aesthetic coherence, color without turning “payaso” ▸ 4:32. Reach and retention, it turns out, are the same discipline pointed at different screens…
// continued in
no entry has continued this idea yet: the arc is still open