P1–P5: A 5-priority system for indie SaaS roadmaps that actually works
Why 3 priority levels aren't enough, 10 are too many, and the practical framework solo founders can use to stop shipping by feeling.
P1–P5: A 5-priority system for indie SaaS roadmaps that actually works
Every prioritization framework an indie founder reads is wrong for them.
RICE assumes a product team with a researcher and an analyst. ICE collapses to one number you'll game without realizing it. MoSCoW gives you four buckets so amorphous they're functionally one bucket. Eisenhower is for inboxes, not roadmaps.
What a solo founder actually needs is a system that:
- Forces a decision within 30 seconds per item
- Scales from 5 items to 50 without collapsing
- Plays nice with the way a one-person team actually works — week to week, with a single brain
- Stops you from re-ranking everything every time a customer complains
The thing I've used for the last 18 months, after trying everything else, is a 5-level priority system. P1 through P5. Below is how it works, why each part exists, and the small rules that keep it from degenerating into a 1-tier system (everything is "high priority").
The bug with 3-tier systems
"High / medium / low" sounds clean. In practice, everything urgent becomes "high," which after a week means "high" has 12 items, "medium" has 4, and "low" has the things you'll never build.
The bug: you can't act differently on item #1 in the high pile vs. item #12. They're labelled the same.
A 3-tier system gives you the illusion of prioritization without the discrimination.
The bug with 10-tier systems
"P1 through P10" or RICE-style numeric scores look rigorous, but no one can mentally distinguish P4 from P5. The bottom half gets treated as one bucket: "stuff I'm not doing right now."
Worse, fine-grained systems invite re-ranking every time a customer breathes. P3 becomes P2 because of a single Intercom message, then becomes P3 again three days later when a different customer pings. You spend more time re-ranking than building.
The bug: the precision is fake. Your real resolution is 5 levels max.
The 5-priority system
| Priority | Time horizon | Selection rule |
|---|---|---|
| P1 | Ship this week | Revenue or retention at risk. Something is on fire. |
| P2 | Ship this sprint (≤2 weeks) | Friction at the moment of value. Trial-to-paid blockers. |
| P3 | This month | Theme with strong mention count, no immediate dollar attached. |
| P4 | This quarter | Strategic bet. Compounds over time. |
| P5 | Watchlist / backlog | Weak signal worth tracking. Don't delete; don't build. |
Five buckets. Each with a time horizon, not just a vibe. Each with a selection rule — a sentence you can apply in under 30 seconds.
The rules that keep it from collapsing
Rule 1: No more than 3 P1s at any time
P1 means "ship this week." A solo founder cannot ship more than 3 meaningful things in a week.
If your P1 pile has 4 items, one of them is lying. Demote the weakest to P2.
This is the rule that does the most work. The instinct is to put everything urgent at P1 — but if you can't actually ship 5 P1s in a week, then 2 of them are functionally P2s pretending. Be honest.
Rule 2: P5 isn't "no." It's "not yet, watch it."
Most ideas die in P5 — that's the point. P5 is the morgue for ideas that aren't dead-dead, just deferred.
Don't delete P5 items. Re-read the P5 list at quarterly review. If a P5 item still feels alive after a quarter, it might be a P4. If it feels dead, archive it. If it's getting louder in customer feedback, it might be jumping to P3.
The watchlist is where you avoid two failure modes at once: building things nobody asked for (you'd have stopped tracking) and deleting things you'll later wish you remembered (you'd have lost the signal).
Rule 3: Every item has a verb and an effort estimate
"Mobile app" is not a roadmap item. "Build a PWA installable on iOS, Android, and desktop, with offline draft-saving for the editor (L effort)" is a roadmap item.
The verb forces you to know what you're building before you prioritize it. The effort estimate forces you to know what it'll cost.
Use a 5-tier effort scale to match the 5-tier priority:
- XS: < 1 day
- S: 1-3 days
- M: 4-8 days
- L: 2-3 weeks
- XL: 1 month+
If you can't put a verb + effort on an item, it doesn't belong in the roadmap yet. It belongs in a "needs scoping" doc.
Rule 4: Re-rank weekly, not whenever a customer complains
The roadmap is re-evaluated on a fixed cadence — Monday morning, 30 minutes. Not "whenever someone DMs you with a feature request."
A single hot email cannot move a P3 to a P1. A pattern across multiple customers can. The 30-minute weekly review is where you assess whether new signal has actually changed the picture — not where individual feelings re-shuffle the deck.
Rule 5: Confidence is separate from priority
You can be very sure something is P1, or you can be guessing it's P1. Those should be treated differently.
Pair each priority with a confidence rating: high / medium / low. P1-high-confidence gets shipped this week, full stop. P1-low-confidence gets a 1-day investigation first to validate before committing engineering time.
This is the underrated piece of the framework. It tells you when to de-risk before building, vs. when to just go.
A worked example
Here's what my roadmap looked like last quarter after running 60 pieces of feedback through this system:
P1 (ship this week):
- Fix the CSV import bug that drops rows on files >5MB. (S effort, high confidence — 8 mentions, 1 churn risk attached.)
- Add SSO via WorkOS. (M effort, high confidence — gating two enterprise deals worth $4K MRR combined.)
P2 (this sprint):
- Build the Linear push integration. (M effort, medium confidence — 5 mentions, mostly Team-tier prospects.)
P3 (this month):
- Onboarding wizard with sample data. (M effort, high confidence — 11 mentions, mostly day-1 churn signal.)
- Slack delivery for weekly digest. (XS effort, high confidence — easy unlock for Team tier.)
P4 (this quarter):
- Notion two-way sync. (L effort, low confidence — strategic bet; need to validate demand more before committing.)
P5 (watch):
- Mobile app / PWA. (L effort — keeps showing up, but in the language of "would be nice," not "I'm churning over it.")
- Bulk delete contacts. (XS — workaround exists; come back if it spikes.)
- Public API. (M — only one mention, but from a high-value account; watch.)
That whole list lives on one page. I look at it every Monday for 5 minutes. New items get triaged into the right bucket using the selection rules. Old items get re-evaluated against any new evidence.
It is the only roadmap system I've used that I actually re-read more than once.
Where Omentoo fits in
The P1-P5 system requires some synthesis input to feed it — themes, mention counts, severity ratings, real quotes. That's the part Omentoo automates.
Every Omentoo recommendation comes back with:
- A priority score (P1 through P5) using exactly the rules above
- An effort rating (XS / S / M / L / XL)
- A confidence rating (high / medium / low)
- Quoted text from the source feedback
- A rationale for why this beats other options at the same priority
So the loop is: paste your customer feedback into Omentoo → get back a P1-P5 sorted list with evidence → spend 5 Monday-morning minutes confirming or adjusting → ship.
The whole system collapses without the synthesis step. P1-P5 only works if you actually know which themes show up 8 times vs. 2. Without honest counts, you're just shipping by feeling with extra steps.
Try Omentoo → — paste your feedback, get back a P1-P5 sorted roadmap with effort and confidence ratings on every item.