MobCorp — Meeting Brief

Mobley & Helms · Thursday, November 19, 2026 · Internal, not for distribution

0paying customers
133ventures (125 with a live product, 1 validated)
66%asks closed without Johnny (goal 80%)
47open items needing a human

1. What has been holding us up

Each holdup with its diagnosis: root cause, evidence, and the fix in flight.

HoldupRoot causeEvidenceFix in flight
No revenue yetDemand problem, not a product problem: almost no real prospects reach checkout. Sales and distribution lag far behind building.0 paying customers. 54 businesses emailed, 18 opened, 0 replied. Most site traffic is bots. The only "purchases" were our own checker bot.Outreach batch to 23 Tacoma businesses ($15/mo site rebuild) ready for approval; focus on two money paths; tag bot checkouts so the numbers stay honest.
Release tool stops for a personSafety guards treat false alarms the same as real failures and hold for Johnny.Releases pause at the hold step waiting on a human, even when the issue is a false alarm.Making releases self-healing: clear false-alarm holds automatically, stop only for real failures.
App Store pipelineThe working TestFlight setup was built for the Mobley app only and never generalized.Mobley app reaches testers; GameGob titles, Heatline and Barnlight do not.Extending the Mobley TestFlight path to the other apps, one app first.
AI capacityHeavy reliance on outside coding agents with hard usage quotas.Codex, Gemini and Antigravity are out of quota; work concentrates on fewer agents.Shift work to available agents; plan quota and spread across providers.
Machine healthEverything runs on one Mac, and a process leaks memory over long runs.Long-running work slows or stops on the main Mac.Find and fix the leaking process; move long runs off the single machine.
One-person sign-offsSign-ins, permissions and money movement can only be done by Johnny.47 items open for Johnny, but only about 6 truly need him (Google sign-in, Claude and Mac permissions, outreach batch, moats decision). Mobley already closes 59% of asks alone."Needs Johnny" ledger with an owner per ask; batch the remaining steps into one sitting; target 80% closed without Johnny.
Breadth over depthWe build faster than we finish and release.133 ventures, only 1 validated. About 30 finished pieces of work (code branches, Worker edits) were never released or merged.Release or delete the ~30 unreleased items before building anything new; pause all but two ventures for 30 days.

2. What we are doing about it

3. Strengths and weaknesses as a conglomerate

Strengths

  • Very high build speed: AI agents ship sites, apps, games and tools in hours.
  • Owned "civil stack": our own deploy, release, records and agent tooling on Cloudflare, low fixed cost.
  • Broad portfolio across games, AI tools, government services, media and anime.
  • Payments, sign-in and checkout already built and reusable across ventures.
  • Institutional memory: decisions, ledgers and reports written down, so work survives handoffs.

Weaknesses

  • No customers, no revenue, no proof of demand yet.
  • Too many ventures for the attention available; focus is diluted. About 30 finished pieces of work are sitting unreleased.
  • Founder bottleneck on sign-ins, approvals and money movement.
  • Distribution and sales are underdeveloped next to engineering.
  • Dependence on outside AI quotas and one machine.

4. Prospectus

What MobCorp is: an AI-operated holding company that builds and runs many small digital ventures on one shared, low-cost platform. Each venture reuses the same sign-in, payments, hosting and agent workforce, so a new business costs days, not months.

Where value comes from:

The ask of the next 30 days: prove that one venture can repeatedly take money from real strangers.

5. What will serve us best next

  1. Pick the two closest-to-revenue ventures and pause the rest. Recommended: Abstergo site rebuilds (local outreach ready now) and the Argo / WeylandAI government track (LOI and bid in hand).
  2. Approve the outreach batch today and measure replies, not traffic.
  3. Do the one-sitting tasks (Google sign-in, permissions) so the agents stop waiting.
  4. Get the Argo letter of intent signed and finish the Madera bid.
  5. Ship one app beyond Mobley through TestFlight to unblock the App Store path.
  6. Fix the Mac memory leak and diversify AI capacity so work does not stop overnight.

Decisions for this meeting