Giving your teams the context to navigate uncertainty and the permission to pivot on the route without losing the destination.
Just as the global positioning system (short: GPS) is the technology that powers navigation software, the GPS Framework (Goal, Positioning, Strategy) enables you to navigate your company to success.
Three questions. Three time horizons. One structure.
Most teams already think in cycles: sprints, quarters, trimesters. The cycle GPS gives structure to what you're already doing.
| Time Horizon How much time do we invest? |
Goal Where are we going? |
Positioning Are we there yet? |
Strategy What are our route preferences? |
|---|---|---|---|
|
Cycle GPS
1 week – 1 year
Play to win this round. Higher uncertainty → shorter cycle.
|
What you're achieving this cycle. Concrete end state. Derived from the current bottleneck in your long-term strategy. (Think: an OKR objective.)
format: plain text
"Ship public beta with business licensing by end of cycle."
|
Current and target values on metrics that show the gap. (Think: OKR key results.)
format: metrics from → to
Live processes: 6 → 18. Organizations: 2 → 5. Billing MVP: no → yes.
|
What to prioritize when you hit a fork. Not a to-do list. (Think: strategic bets.)
format: visual map
Map axes: customer segment (small teams → mid-size → enterprise) × adoption depth (exploring → committed → dependent). Focus on the bottom-left: small teams still exploring. Move them upward to committed. Don't move right into enterprise until billing is proven.
|
The cycle GPS doesn't exist in isolation. The same three questions apply at two more time horizons, forming a fractal pattern:
Infinite GPS gives long-term direction that persists across cycles.
Cycle GPS produces concrete outcomes within a bounded time frame.
Immediate GPS guides autonomous execution of individual tasks.
Each level nests into the one above. An immediate GPS serves a cycle strategy. A cycle GPS addresses a bottleneck in the infinite strategy. The metaphor stays the same. The scope changes.
| Time Horizon How much time do we invest? |
Goal Where are we going? |
Positioning Are we there yet? |
Strategy What are our route preferences? |
|---|---|---|---|
|
Infinite GPS
near-permanent
Play to keep playing. Long-term direction. Persists across cycles.
|
How you imagine the future. Currently out of reach but not unimaginable. Never "completed." (Think: Vision statement.)
format: plain text
"Running any process between people is effortless. The process just works and the history is always there."
|
The single number that tells you if you're succeeding at your goal. Must be independently observable, not self-reported. (Think: One Metric That Matters from Lean Startup.)
format: single metric
Total sessions completed. A session = a process that actually happened between real people.
|
The reinforcing loop that compounds your progress over time. (Think: Flywheel from Jim Collins' Good to Great.)
format: visual loop
Easy process creation → More sessions → Richer data → Smarter automation → Less admin burden → More processes worth running → …
|
|
Cycle GPS
1 week – 1 year
Play to win this round. Higher uncertainty → shorter cycle.
|
What you're achieving this cycle. Concrete end state. Derived from the current bottleneck in your Infinite Strategy. (Think: an OKR objective.)
format: plain text
"Ship public beta with business licensing by end of cycle."
|
Current and target values on metrics that show the gap. (Think: OKR key results.)
format: metrics from → to
Live processes: 6 → 18. Organizations: 2 → 5. Billing MVP: no → yes.
|
What to prioritize when you hit a fork. Not a to-do list. (Think: strategic bets.)
format: visual map
Map axes: customer segment (small teams → mid-size → enterprise) × adoption depth (exploring → committed → dependent). Focus on the bottom-left: small teams still exploring. Move them upward to committed. Don't move right into enterprise until billing is proven.
|
|
Immediate GPS
hours – days
Execute with autonomy. Written by the person accountable. Writing it is the first step of execution.
|
What outcome this task produces and why it matters. Links to the cycle strategy it executes. (Think: a user story.)
format: plain text
"Marketing site explains what we do to a first-time visitor. Unblocks sales conversations that currently require a live demo."
|
Current state: what concretely exists right now. Done state: what's verifiably different when complete. (Think: acceptance criteria.)
format: current → done
Current: No public site. Sales require live demo. Done: Site live. Visitor understands what we do and requests access without talking to anyone.
|
Guardrails, constraints, and reference points that tell the executor what to prioritize, what to avoid, and where to look. Not steps. (Think: an approach brief.)
format: guardrails + constraints + references
Use strategy docs as content source. Follow brand direction. Lead with "process between people" framing, not technical features. Don't over-promise unshipped capabilities.
|
GPS works at every scale. The same three questions apply whether you're a founding team running weekly experiments or a 500-person org with quarterly cycles. The time horizon changes. The scope changes. The fundamental questions stay the same.
The operational guide. How to plan, check in, review, and adapt using GPS.
This guide covers the cycle level, where most of the operational work happens.
Every cycle follows the same rhythm:
How often you check in depends on cycle length. A rule of thumb: quarterly cycles use bi-weekly check-ins, yearly cycles use monthly check-ins, and weekly cycles use daily stand-ups.
At the start of each cycle, re-iterate the GPS from the scope above, then build this cycle's GPS in five steps. The order matters: you start from where you are, not from where you want to be.
Two sheets for running the steps above. Both are Google Docs: opening one makes your own copy, which you can fill in on screen or print.
No Google account? Download the GPS sheet or the cover sheet as Word files.
Both are licensed under CC BY-SA 4.0, like the rest of this page. Adapt them, and pass the adaptation on under the same terms.
The purpose of check-ins isn't tracking. It's creating repeated opportunities for the people accountable at each scope to demonstrate that these goals matter. When someone reports being behind, the reaction from the scope above (curious and problem-solving, not punitive but clearly invested) signals whether the GPS is real or paperwork.
A check-in follows four steps. The order matches the adaptation rule: cheapest change first.
Check-ins serve two purposes: internal course correction and external accountability. You use them to adjust strategy. The scope above uses them to assess whether the investment is on track. If you are significantly off-course, the investment board may offer support, adjust expectations, or reduce scope. Transparency about setbacks builds trust. Hiding problems until the review breaks it.
Four things can end a goal cycle. Only the first one is intuitive. The other three prevent death marches.
Using only #1 is the hallmark of a death march: everyone can see it's pointless, but reconsidering the goal isn't part of the process. Adding #2 through #4 is the hallmark of agility. You get #2 from consistent goal cycles. You get #3 and #4 from questioning your goals during check-ins. At the end of a cycle, you may decide to continue with a goal, but that's a deliberate investment decision, not inertia.
Once you know which ending you hit, assess what happened and extract learnings for the next round. Same four steps as the check-in, different questions:
The review sets the stage for the next cycle's pitch. Those who delivered on their promise earn credibility for bolder asks. Those who missed with good reasons and documented learnings may continue at similar investment levels. A pattern of misses without learning needs a structural conversation, not another pitch.
Goal alignment is bidirectional. The GPS from the scope above cascades down as a frame: here's where we're going, here's where we are, here's our approach. Cycle goals then bubble up as investment pitches: here's what we propose to deliver within that frame, and here's what we need to do it.
The cover sheet template holds both parts, with a goal sheet attached per goal.
Pitches flow upward through the organizational structure. Team leads pitch to unit leads. Unit leads pitch to company leads. Each scope acts as investment board for the scope below. The frame cascades down. Proposals flow up.
Pitches don't happen in a vacuum. A cycle GPS must fit the GPS of the scope above (the super-GPS). Every pitch depends on all three parts of the super-GPS: the goal (pitches must contribute to it), the positioning (pitches must move the metrics), and the strategy (pitches must be coherent with the approach, or explicitly challenge it).
When a pitch doesn't fit, two possibilities: the pitching scope misunderstood the context (clarify and re-pitch), or the pitch reveals a flaw in the super-GPS (surface it and decide whether to adjust). Good pitches don't just fit. They stress-test.
The dependency also flows upward. When the scope above looks at all pitches together, it must ask: "If all these succeed, will we achieve our super-GPS?" If the answer is no:
The pitch model makes gaps visible and negotiable, instead of the scope above dictating sub-goals that "must" add up and being disappointed when reality intervenes.
Four approaches to aligning direction. Same challenge, different trade-offs.
|
GPS
New Work by Design
|
OKR
Intel / Google
|
Shape Up
Basecamp
|
SAFe
Scaled Agile, Inc.
|
|
|---|---|---|---|---|
|
Core components
What you write down
|
Goal + Positioning + Strategy at each time horizon | Objective + Key Results (+ Initiatives, added later by practitioners) | Pitch (problem + appetite + solution boundaries) | Strategic themes → PI objectives (valued by business owners) → features ranked by WSJF on the ART planning board |
|
Time horizons
How far out you look
|
Three: infinite, cycle (1 week–1 year), immediate (hours–days) | One: quarterly (sometimes annual) | One: fixed 6-week cycles + 2-week cooldown | Three: portfolio (themes, roadmap), Planning Interval (8–12 weeks), iteration (usually 2 weeks) |
|
Who sets direction
Where goals come from
|
Scales with maturity. Founder-led early on, progressing to team-pitched as domain ownership grows. At maturity: each scope proposes cycle goals as investment pitches, the scope above acts as investment board. | Varies. Cascaded top-down, bottom-up, or bidirectional depending on implementation. | Leadership selects pitches at the betting table. Anyone can pitch, but leadership decides. | Product management ranks features, business owners assign value. Teams decide how much they commit to during PI Planning. |
|
How progress is tracked
What tells you where you are
|
Metrics on a spectrum (current → target). One metric at infinite level, multiple at cycle level. Observable states at immediate level. | Key Results: measurable outcomes, typically 3–5 per objective. Scored 0–1.0. | Hill charts: uphill (figuring out) → downhill (executing). No metrics. | ART predictability: business value achieved vs. planned on committed objectives. 80–100% counts as healthy. |
|
Role of strategy
How you decide what to do
|
Explicit, separate artifact. Decision guide for forks, not a task list. Described as a visual map. | Not native to the original model. 'Initiatives' are often added, but usually degrade into rigid task lists rather than true strategic guidelines. | Implicit in the shaped pitch. Boundaries ("what's in, what's out") guide the team. | Expressed as a ranked feature backlog. That's a route, not a guide for forks in the road. |
|
Positioning vs. Strategy
Are they separated?
|
Explicitly separated. Positioning stays stable as reference point. Strategy adapts freely. | Blurred. Key Results mix "where we are" with "how we get there." No separation. | Partially. Appetite sets the constraint. Hill chart tracks confidence, not metrics. Approach is free. | Blurred. The planned features are both the route and the yardstick: predictability measures delivery against the plan. |
|
When things stall
What happens when progress stops
|
Explicit sequence: change strategy first → adjust positioning → rethink goal. Cheapest change first. | No prescribed adaptation sequence. Check-ins happen but there's no explicit protocol for what to change first. | Cut scope, don't extend time. If it doesn't ship, it returns to the pitch pile for re-betting. | Uncommitted objectives absorb small surprises. Bigger changes mean renegotiating dependencies across teams, so they tend to wait for Inspect & Adapt and the next PI. |
|
Cycle length
How it adapts
|
Variable: matched to maturity stage. 1 week (ideate) → 1 year (mature). | Fixed: quarterly. Some orgs add annual layer. | Fixed: 6 weeks. No exceptions. | Fixed cadence: the whole train shares one PI calendar. |
|
Autonomy model
Who decides how
|
High. Executor writes their own GPS. Scope above sets the constraint, everything inside is yours. | Depends on implementation. Ranges from heavy top-down cascade to full team ownership. | High within cycle. Team has full autonomy over execution. Zero between cycles: betting table decides what's next. | High over how, low over what. Teams plan and estimate their own work. Features come from the backlog above. |
|
Best suited for
Path certainty assumed
|
The destination is stable but the path will change. You need to navigate, not just track. | The destination is clear, measurable, and the path is trusted enough to commit to outcomes for a fixed period. | The problem is understood well enough to bound the investment, but the solution path is open within that boundary. | Many teams with hard dependencies (hardware, regulation, legacy) that can't be cut yet. Coordination cost is the main problem. |
Not by design. Often in practice. A Planning Interval is a goal cycle with a fixed length, and PI Planning is the strategizing session for a whole release train. Uncommitted objectives give buffer. The trouble starts when the event's output turns from a set of goals into a plan.
Three things push it there. First, the planning board maps features to iterations, so the feature list becomes both the route and the measure of success. Changing the route then feels like failing the PI: the same trap that separating positioning from strategy exists to avoid. Second, the predictability measure only rewards one of the four endings of a journey: arriving as planned. Learning mid-PI that a goal can't be reached, or finding something more valuable, lowers the score. Third, changing course means renegotiating dependencies across teams. That costs enough that people avoid it, even when what they learned says they need to.
The root cause is structural. PI Planning takes dependencies as a given and coordinates around them. The heavier the planning event, the more coupled the organization. Reducing those dependencies does more for agility than any change to the event.
| SAFe | GPS |
|---|---|
| Strategic themes, portfolio vision | Infinite GPS. Add what's missing: one metric that matters and a flywheel. |
| Planning Interval | Cycle GPS. The PI length is the battery for the journey. |
| PI objectives | Goal and positioning, written as current → target, not as a list of features. |
| Feature backlog, WSJF | Options the strategy chooses between. Add a strategy statement per objective: what wins at a fork. |
| Iteration planning, standups | Immediate GPS. |
| ART sync, system demo | Check-ins. Ask in GPS order: strategy first, then targets, then the goal. |
| Inspect & Adapt | Review. Name which of the four endings you hit, and report what you learned alongside predictability. |
Treat the planning board as a map of dependencies, not a promise of scope. That keeps what PI Planning does well (getting a hundred people to see the same picture) without turning the next ten weeks into a fixed route.
Based on Core SAFe 6.0. Scaled Agile's AI-Native SAFe (June 2026) moves in the same direction: a one-day PI Outcome Planning anchored to outcomes instead of feature commitments, and a short Sense & Respond event every iteration in place of the end-of-PI Inspect & Adapt.