What 7 days of programmatic SEO actually looks like (numbers inside)

Open log of Week 1 of a free-marketing growth system for a small
AI video aggregator. Public so the next person doing this has real
numbers to compare against, not VC blog post fairy tales.

Goal: 10 paid customers in 90 days, $0 in paid promotion. Day 7
of 90.

What landed in Week 1

• 4 long-form blog posts (model tutorials + prompt structure)
• 8 programmatic /compare/<a-vs-b> pages
• 6 programmatic /model/<slug> spec sheets covering Veo 3.1, Sora 2,
  Kling 2.5 Turbo, Seedance 2.0, Nano Banana Pro, FLUX 2 Pro
• /compare hub + /model hub with ItemList JSON-LD
• UTM attribution captured into the user table on signup
• Each spec page carries SoftwareApplication + FAQPage JSON-LD —
  the schemas LLM search engines lift from most readily

Total: ~16 indexable surface-area pages live, each cross-checked
against vendor primary docs (no scraped or made-up specs).

What the numbers actually look like

• PostHog 7d: 10 unique visitors / 30 pageviews (up from 5 / 12 a
  week ago — first WoW jump on no outbound spend)
• Google Search Console 7d (last pull two days ago): 1 click,
  33 impressions, avg position 10.1 (up from 0 / 29 / 11.1)
• 0 paid customers (Day 0 baseline still — 90-day target is 10)

What I'd flag for anyone copying this

1. Programmatic SEO compounds earlier than the conservative
   curve suggested — uniques doubled WoW with zero new outbound
   effort. The surface area was the lever, not social-post volume.
2. SoftwareApplication + FAQPage JSON-LD on every model fact
   sheet costs ~5 minutes per page and shows up in LLM result
   formats with no extra work. Cheapest LLM-SEO move there is.

Open to anyone running a similar plan trading notes — drop the
URL of what you're building.