Your territory model is a spreadsheet. That's the problem.

Territory design in most revenue orgs is an annual exercise. The VP of Sales and a RevOps analyst spend two to four weeks in a spreadsheet with dozens of tabs, publish a plan, and the plan is stale by February. Four structural reasons that pattern fails:

1. The inputs are static and thin. Account count, geography, sometimes revenue. Almost never scored potential, expansion whitespace, renewal risk, buying-capacity health, or procurement complexity. The plan is balanced on variables that do not correlate with outcomes.
2. Territory design is disconnected from quota setting. A territory is defined, a quota is assigned, and nobody validates whether the territory can structurally yield the quota. When the territory cannot, the rep fails and leadership blames execution instead of architecture.
3. The model does not rebalance. A rep leaves, a whale churns, a new market opens, a competitor exits. The territory map stays frozen until next year's offsite. Meanwhile, reality has moved on.
4. The plan optimizes for equity of inputs, not equity of opportunity. Everyone gets 50 accounts. But if my 50 accounts are 3x the potential of your 50 accounts, we do not have balanced territories. We have a balanced spreadsheet.

A working territory model is a scored, continuously computed, rebalanceable system, not a slide. Phase 1 work moves it from the spreadsheet to the operating substrate.

Account potential is not ARR. ARR is what the account is worth today. Potential is what the account is worth over the planning horizon, weighted by how likely you are to realize it. Five inputs combine into a potential score.

One formula that validates whether a territory can structurally yield its quota. If this number is below 3.0, the rep cannot hit quota except through exceptional execution or luck.

The useful version of this formula is computed per rep and per motion. A new-business rep needs higher coverage (closer to 5x) because the motion is harder and slower. A renewal-plus-expansion rep can work at lower coverage (closer to 3x) because the motion is warmer and more predictable. A dual-motion rep needs coverage computed against the blended target.

Before a quota number gets published, it passes three validation checks. Without them, you are setting up some reps to miss before the year even starts.

A quota that fails any of the three checks should not be published. Either the quota comes down, the territory gets supplemented, or the motion assumption gets rebuilt.

Annual rebalancing is the default pattern and the wrong pattern. The business moves faster than the planning cycle, and every territory drifts away from balance during the year. Five triggers that should force a rebalance regardless of calendar:

1. A rep leaves. Their book needs to be distributed, not parked. Redistribution is an opportunity to rebalance the entire team, not just to reassign the orphaned accounts.
2. A whale churns. A single-account loss of 10%+ of territory potential destabilizes the affected rep's book. Rebalance or supplement.
3. A new market opens. New ICP segment, new geography, new product line. Territory boundaries should shift to capture it, not wait until next year.
4. Quarterly drift exceeds threshold. If coverage variance across reps exceeds 30%, the team is no longer on the same plan. Mid-year rebalance is necessary.
5. ICP weights change. When Phase 1 ICP work adjusts composite weights, territory potential scores shift. Any material change in ICP definition is a rebalance trigger.

Rebalancing carries political cost. Reps get attached to accounts. Managers advocate for status quo. The way to reduce that cost is to make rebalancing routine and transparent, with the rebalancing logic visible to the team. When the trigger is explicit and the math is on the table, rebalancing is operational hygiene, not a threat.

Orgs running two motions on shared infrastructure face a design question that does not exist for single-motion orgs: do reps carry one motion or both? Three patterns:

1. Motion-specialized reps. Rep A works Motion 1 only. Rep B works Motion 2 only. Pros: deep motion expertise, clean quota math per motion. Cons: doubles the rep count for overlapping accounts, introduces cross-motion handoff at the account level.
2. Account-specialized reps (dual-motion). Rep A owns all motions for Account 1. Rep B owns all motions for Account 2. Pros: unified customer experience, single relationship to manage. Cons: requires reps to master both motions, blended quota math harder to calibrate.
3. Hybrid: pod model. A team of two or three specialists co-own a book. Named lead rep per account, but specialist reps engage when their motion is the active one. Pros: best of both. Cons: most operational complexity, requires clear ownership rules.

The decision is less about which pattern is "right" and more about which pattern your org can operationalize. If you have clear cross-motion handoff governance (see the six-handoff governance framework), pattern 1 is viable. If you do not, pattern 2 or 3 is safer.

Twelve yes/no questions to pressure-test the model. The no's rank your rebuild priority.

Ship the potential score. Validate against current territories. Surface the imbalance numerically. Then rebalance with the math on the table.

Anti-pattern to avoid: designing the model but not rebalancing the team. A potential score that reveals imbalance and produces no action is political theater. The Phase 1 work is done when the rebalance has shipped, the new quotas are validated, and the review cadence is calendared.