3e. Real portfolios are lopsided — and that changes the answer

# 3e. Real portfolios are lopsided — and that changes the answer

Everything above prices a *uniform* constellation. Real ones aren't: a Director runs one or two hot
Domains on a fast beat, several slow ones on `weekly`, and a Private Domain that crawls nothing.
`portfolioWeeklySpend` costs each member at its own cycles, cadence, and breadth.

**A realistic lopsided 6-Domain constellation** (1 hourly/broad, 2 daily, 2 weekly, 1 private):

| Member | $/wk | Share | Articles/wk |
|---|---|---|---|
| Supply — Fabs (hourly, broad, 4 terms) | $8.64 | **74.1%** | 5,600 |
| Demand — Compute (daily) | $0.99 | 8.5% | 400 |
| Silicon Competition (daily) | $0.84 | 7.2% | 300 |
| Policy — Export (weekly) | $0.42 | 3.6% | 30 |
| Power buildout (weekly) | $0.42 | 3.6% | 30 |
| My Strategy (private, no crawl) | $0.34 | 2.9% | 0 |
| **Total** | **$11.66** | | 6,360 |

**One member is three-quarters of the bill.** And the same six Domains cost **$5.04/wk uniform vs
$11.66/wk lopsided** — a 2.3× spread from shape alone. So:

> **Domain count is a poor cost predictor.** The distribution of cadence × breadth across members is
> the whole story. Any per-seat pricing or in-product estimate that reasons from "how many domains"
> will misprice by >2× in both directions.

The product consequence is that the Operations surface should lead with **concentration** — *"one loop
is 74% of your spend"* — rather than a portfolio average. That single sentence is more actionable than
any total, and it points at the one lever worth pulling.

### The Firecrawl plan cliff — and it lands on the default constellation size

Credits are bought in monthly blocks, so cost steps rather than slopes:

| Domains (uniform default) | $/wk | Credits/mo | Plan needed | Plan $/mo | Utilisation |
|---|---|---|---|---|---|
| 1 | $0.85 | 1,560 | Hobby | $16 | 31% |
| **3** | $2.52 | **4,680** | **Hobby** | **$16** | **94%** ⚠️ |
| 6 | $5.04 | 9,360 | Standard | $83 | 9% |
| 10 | $8.39 | 15,600 | Standard | $83 | 16% |
| 20 | $16.78 | 31,200 | Standard | $83 | 31% |

**A default 3-Domain constellation sits at 94% of the Hobby allowance.** The fourth Domain — or one
step from steady to broad — forces Standard: **$16 → $83/mo, a 5× jump for a ~33% workload increase.**
That cliff lands exactly on the shape onboarding produces by default, which makes it a product problem,
not a footnote:

- **Warn before the step,** in the breadth/cadence picker — *"this widens you past the Hobby
  allowance"* — since the marginal Domain looks cheap ($0.84/wk) while actually costing $67/mo in plan
  step.
- **Beyond the step there is enormous headroom.** Standard covers 20+ Domains at 31% utilisation, so
  the honest advice above 4 Domains is *"grow freely, you've already paid for it."*
- Self-hosting Firecrawl (AGPL) removes the ladder entirely and is the right answer for a heavy
  Director.

---

Raw Markdown source