The common indie trap isn’t “zero revenue.” It’s this:

Several projects each make a little money—and your attention gets stretched thinnest where it matters most.

On the P&L it looks safe: A does a few thousand a month, B does one thousand, C closes the occasional deal.
On the calendar it’s the opposite: the best earner often gets leftover scraps of time.

Pat Walls has described that shape in public posts: around 2020, Starter Story already carried most of the revenue while he still juggled another business and many side threads—energy scattered, with the main product not getting firepower matched to its revenue share. After a full week offline to think (a Think Week), he shut the distractions and went all-in; he later wrote that revenue stepped up quickly.

This guide is not his biography. It extracts one decision:

With multiple live projects, where does attention go—and what do you kill?

1. Start with one table: revenue ÷ hours

Don’t ask “which has more upside” first. Ask:

Project	Trailing 90-day revenue (or steady monthly)	Real weekly hours	Revenue ÷ hour
A (your name)	e.g. $8,000	e.g. 8	≈ $1,000
B	e.g. $2,000	e.g. 25	≈ $80
C	…	…	…

“Real hours” include support, bugfixes, content, pricing debates, and midnight anxiety. Skip those and you’ll overrate anything that “almost runs itself.”

Sorting rules:

1. Highest revenue ÷ hours is the default bet.
2. High revenue share + low time share usually means you’re starving a winner—not that it’s “fine on autopilot.”
3. Low revenue + high time goes on watch unless a key hypothesis is still being tested.

Revenue ÷ hours is only the first filter, not the final answer.
Also check stability, growth slope, market room, and whether more attention would actually amplify the line.
Otherwise a $500/mo project that takes one hour a week can look like a “main bet” on hourly math—when it’s often mediocre success, or already at its ceiling.

Pat’s later diagnosis was basically this table: most revenue, least focused firepower. The gap wasn’t ideas—it was under-executing the already-validated answer.

2. Three kinds of “paying”—not one business

Type	Signs	Default move
Validating	First paid users; growth unstable; core bets untested	Fixed window (e.g. 4–8 weeks) + written pass/fail; then decide
Mediocre success	Covers some bills; plateau; you don’t want to double down	Exit line on revenue/growth/energy; hit it → kill or sell
Main bet	Highest revenue share or best slope, and skill fit	Default attention here; everything else maintenance-only

The easy misread is the middle cell: mediocre success.

It pays, so shutting it feels rude.
It won’t climb, so you keep opening “the next product” for safety.
Result: the main bet never gets a full breath; side bets never get a clean death.

Three checks:

• Would 2× time clearly grow revenue—or only grow exhaustion?
• Does this match what you’re good at, or only “I already built it”?
• Would you start this again from zero?

Two “no”s out of three: stop using “but it makes money” as a shield.

3. Kill order (more important than “ship another”)

Default move isn’t another launch. It’s:

Unload first, then press.

Suggested order:

1. Zero growth + high maintenance → kill or hard maintenance (fatal bugs only).
2. Pays but skill-mismatch → sell / hand off / shut; it steals decision bandwidth.
3. Validation window expired without clear signal → stop; “wait a bit more” is usually the costliest line.
4. One main bet + at most one validation slot → normal ceiling for a tiny team.

“Validation slot” means only one project may be in active build/growth mode. Everything else maintains or dies.

For a solo maker, the real ceiling isn’t how many products you own—it’s how many are allowed to burn your active growth energy. Default: one.

A ten-product portfolio can work—if maintenance is systematized and you still actively grow only one. Ten products each demanding “what’s next?” isn’t a portfolio; it’s attention debt.

4. When not to go all-in

All-in is conditional, not moral.

Don’t all-in yet if:

• The “main” product still lacks stable pay or a clear growth signal;
• You’re escaping: new projects dodge pricing/distribution on the old one;
• The second project is only insurance theater—killing it wouldn’t grow the first.

All-in is reasonable if:

• The revenue÷hours filter, plus stability / slope / amplify-ability, still points at the same project;
• What’s missing is execution density (content, conversion, delivery, sales)—not another idea;
• You can write: what dies, and which 2–3 moves the main bet gets for 30–90 days.

Pat’s Think Week helped not because road trips are romantic, but because he translated “I’m doing a lot” into “I’m starving the one thing worth feeding.”
You don’t need three thousand miles—you need one ranking session that daily firefighting can’t interrupt.

5. 30-day card: from parallel to pressure

Week	Do this
1	Fill revenue÷hours; label main / mediocre / validating
2	Write exit or maintenance rules for every non-main (date + metric)
3	Kill / sell / freeze; delete their “build time” from the calendar
4	Main bet keeps only 2–3 growth moves; weekly slope review

Pass criteria aren’t “I feel focused.” They’re:

• Weekly hours on the main bet clearly up;
• Non-mains no longer eat “what feature next?” brainspace;
• In 30–90 days, revenue or a key conversion metric moves in an explainable way.

If pressure still yields a flatline, the issue may be product/channel—then change direction. Don’t open a new pit to hide the old one.

When this applies

Best for solos / tiny teams with ≥2 products or services that already earn.
Weaker for pure pre-revenue exploration, or pure calendar-sold consulting (time is inventory; portfolio math differs).
Not financial advice; case figures are founder self-reports used for mechanism, not audited truth.

30-second: which one?

① Trailing 90 days—who brings the most revenue?
② Do your real weekly hours mostly go there?
③ If not: is the gap execution—or mediocre success?

• Revenue king + starved hours → press first, ideas later
• Revenue king + more hours still flat → set an exit—or change the main bet
• Pennies everywhere, growth nowhere → cut to one active growth slot

Not every paying project deserves to be kept alive. The one worth going all-in on already proves people pay—and still has room to grow when you add attention.

Related cases

• Starter Story: how a forced-revenue case library became a media business