Classes 67 through 79 handed you a shelf of engines: audience and community, paid advertising, direct search, social, and email routes, plus this class's partnerships, marketplaces, referrals, and product loops. A shelf is not a strategy. This closing lesson turns the collection into a portfolio — three chosen routes, one measurement cadence, and ninety days of testable work.

The engines, together at last

Recall the map from Lesson 67.2, now with every shelf stocked:

• Audience and community routes (68–71): niche, positioning, recurring value, trust, gathering, contributions, flywheel. Slow to start, compounding when they catch.
• Paid routes (72–75): ads, creative production, testing, affiliates. Fast to test, metered by budget, honest only with message match and math.
• Direct routes (76–79): search pages, social assets, email relationships, partner bridges, marketplace aisles, referral handoffs, product first-value loops. Durable when each points to owned follow-up.

No serious project rides one engine. Search brings strangers with intent; social earns attention and language; email keeps the relationship; community multiplies contributions; paid buys timed reach; partners borrow trust; marketplaces borrow aisles; referrals and product loops turn wins into arrivals. The art is choosing which three deserve your limited weeks — and which dozen you deliberately defer.

Deferral is a decision, not neglect. Name the routes you will not touch for ninety days and why: no capacity, no audience fit, no servable destination yet. Lesson 67.1's discipline returns — one audience, one first route, everything else queued.

The portfolio: one compounder, one test, one direct line

A workable ninety-day portfolio holds exactly three routes:

1. One compounding route. The asset that stacks if you tend it — weekly search pages, the Tuesday newsletter, the template gallery, the community archive. Judge it on a slope, not a spike: are returns, saves, and qualified arrivals growing month over month?

2. One controlled test route. The bounded experiment with a budget, question, and stop date — a small paid-creative test, one marketplace listing, one partner co-send, one referral nudge. Lesson 74.1's rule governs: one question, one variable, one decision rule.

3. One direct relationship route. The line you own regardless of platforms — email list, owned docs, customer follow-up sequence, direct booking channel. When a rented surface shifts, this line keeps Tuesday alive.

Plus a small measurement cadence: weekly check (fifteen minutes, Appendix 79.A scorecard), monthly review (one hour, keep/change/stop per route). Without the cadence, three routes become three hobbies.

Running examples choose differently — as they should

There is no universal stack. Watch the same logic produce different portfolios:

• ByeBuy Classroom compounds with search-ready lessons plus the weekly note (direct), tests one partner newsletter co-send per month, and keeps the course map plus email archive as the owned line. Paid stays deferred until a lesson's activation math is proven.
• Sonariq / Research Desk compounds with the template gallery and sample briefs, tests one workflow-tool marketplace listing, and keeps the saved-brief library plus follow-up email as the owned line. Social is a listening post (77.4), not the compounder.
• Local business launch compounds with map listing plus review habit and the Friday email, tests a small geo-targeted paid offer with a real appointment, and keeps the booking channel plus care follow-up as the owned line. Virality is explicitly deferred.
• Creator-led niche publication compounds with the weekly research memo and archive, tests one integration partnership per quarter, and keeps the subscriber list plus contribution inbox as the owned line. Paid amplification waits until one issue's forward rate earns it.

Same framework, four honest answers. Copying another project's mix is how a tailor ends up buying national video views and a classroom ends up discounting hems.

Exercise: commit to ninety days on one page

Create DISTRIBUTION-PLAN.md. If it needs a second page, it holds two plans — cut one.

# DISTRIBUTION-PLAN.md — [Project], 90 days from [date]

## Audience + promise
- Audience (specific people + moment): ___
- Promise (what arrives, how often, why open): ___

## Three routes (1 compounding + 1 test + 1 direct)
1. Compounding: ___ / weekly action: ___ / slope metric: ___
2. Test: ___ / question: ___ / budget+time cap: ___ / decision date: ___
3. Direct: ___ / weekly action: ___ / health metric: ___

## Weekly (15 min) + monthly (1 hr)
- Weekly scorecard file: ___ (see 79.A) / owner: ___
- Monthly review date: ___ / keep/change/stop per route: ___

## Owner + budget
- Owner per route: ___ / total time per week: ___ / total spend cap: ___

## Evidence to collect
- Exposure / visit / first value / return / lead / cost (5 rows weekly): ___

## Stop rules (write before starting)
- Stop test if: ___ / stop compounding tweak if: ___ / never sacrifice: ___

## Deferred (named, not forgotten)
- Routes deferred + why + recheck date: ___

Worked mini-example — Research Desk, next ninety days: audience "solo analysts citing filings weekly"; promise "one runnable check each Tuesday"; compounding is template gallery plus Tuesday note; test is one marketplace listing with a $300 support-time cap and a six-week decision; direct is saved-brief follow-up email; weekly scorecard owned by the founder, fifteen minutes Fridays; monthly review first Monday; stop the listing if activation stays under 10% after forty arrivals; defer paid social and affiliates with an October recheck.

Finish line: a DISTRIBUTION-PLAN.md with audience, promise, three named routes, weekly and monthly cadence, owner, budget and time caps, evidence fields, stop rules, and an explicit deferred list.

Verify quickly: read the plan aloud in under two minutes. If a teammate cannot repeat back the three routes, the owner, and the stop rule, shorten until they can. A plan nobody can recite is a plan nobody follows.

Common failure mode: the "we should market it" fog — five half-started channels, no owner, no stop rule, vanity metrics celebrated monthly. The portfolio cures fog with scarcity: three routes, one owner each, numbers every Friday, a pause button with its conditions written in advance.

Check your understanding

1. Why does the portfolio require one compounding, one test, and one direct route rather than three tests? 2. How can four projects use the same framework and choose different mixes? 3. What do stop rules protect that enthusiasm alone cannot?

Next

Part XIV closes here. You can now choose a first route, build audience and community around recurring value, run a controlled paid test, publish search, social, and email work that points somewhere, borrow trust through partners, marketplaces, referrals, and product loops — and measure whether discovery became value. Part XV asks the next question: when attention becomes exchange, what is the fair, sustainable price? Appendix 79.A gives you the scorecard to carry across the bridge.