Quarterly Planning From Data

Use 90 days of Foundry data to plan the next quarter, not vibes, not aspiration.

The reason most quarterly plans fail isn't laziness, it's planning from vibes. You sit down with a blank page in week 13, write what you wish were true, and call it strategy. Three months later, the gap between plan and reality is wider than you remember and you can't quite pinpoint where it diverged.

Foundry's advantage is that you have 90 days of structured data sitting in the system already. Decisions made (and how they aged), projects shipped (or didn't), energy mix observed, commitments kept (or not), themes that kept reappearing in your tags. The quarterly plan isn't a fresh fantasy, it's an update to a running model of how you operate.

The output is a plan that pre-acknowledges what's true about you, not the plan you'd write for an idealized version of you.

Before planning the new quarter, close the old one. This is the strategic portfolio review (see that module for the full sweep). At minimum:

• Close every overdue decision. Verdict each one. VALIDATED, PROVED_WRONG, SUPERSEDED, or STILL_ACTIVE with extended revisit.


• Resolve every High-priority recommendation. Park, archive, or address the underlying condition.


• Mark completed projects COMPLETED. Run the closing reflection. Don't carry done-but-not-closed projects into the new quarter.


• Bulk-sweep parked ideas 60+ days old. Restart, archive, or accept the park.

Open four surfaces in sequence and take notes:

/overview, current snapshot. Note any number that surprises you.

/focus, energy mix history. Was the quarter mostly COMPOUNDING, MAINTENANCE, DRAINING? Did you end up running more projects than you intended? Did the Focus Lane stay focused?

/decisions filtered ACTIVE, what bets are still in play and what's the falsification condition for each? Any that should be SUPERSEDED?

/reviews, read the last 12 weekly snapshots. What intentions kept appearing? What wins concentrated where? What learnings repeated?

If you have Pro, run AI Pattern Detection here. It surfaces patterns across 60+ days that no human reads through manually, recurring decision themes, energy trends, capture/triage velocity gaps.

You're looking for three things: what worked structurally (do more), what kept showing up despite not being planned (rename and own it), and what you assumed would matter but didn't (drop it).

Three explicit settings, recorded as a logged Decision:

Capacity max. Adjust based on Phase 2 evidence. Did you ship at the rate the previous max assumed? If not, lower it. If you genuinely had room, raise it cautiously. Document the reason, not just the number.

Focus Lane. Declare the single project that owns Layer 2 next quarter. If you can't name one, the planning isn't done, keep going until you can.

Energy floor. What's the minimum number of COMPOUNDING projects you'll maintain? If you let it drop to zero in the previous quarter, build in a forcing function (e.g., "if I have 0 compounding projects by week 4, I start one even if capacity is tight").

Not weekly intentions. Bigger. Three sentences that define the success of the next 90 days. Use the same outcome-shaped framing as weekly intentions, but at quarterly scope.

Examples:

• "Get the new pricing live and held by at least 20 paying customers for 60+ days."


• "Build the partner pipeline to a state where 3 partners are running revenue tests independently."


• "Compound the founder library to 10 reusable assets that survive the quarter."

Create a Decision titled Q[N] [Year] plan with the following structure:

`
Capacity: [max projects]
Focus Lane: [project name]
Energy floor: [minimum compounding count]

Three intentions:
1. ...
2. ...
3. ...

Why these three:

• ...

• ...

Set the revisit date to the last week of the quarter. The "explicitly not doing" list is the half most plans skip and most need most, it's the saying-no that operationalizes the saying-yes.

End by sharing the summary with one trusted human (advisor, co-founder, accountability partner). The act of sending it externally raises your bar for the writing. If Pro, use a sharing link on the decision; if not, copy/paste the markdown into a message.