Blog · Territory, capacity and quota planning
A method for territory planning where every scenario adds up: the four inputs, the fairness measures, the identity that must hold before two plans can be compared, and a worked comparison of moving two reps between regions.
Territory planning season is one analyst, one workbook and three weeks, ending in a plan nobody can fully explain. The cause is rarely the plan. It is that the scenarios along the way did not reconcile, so nobody could compare them, so the last one standing won. This guide sets out how to build scenarios that add up and compare them on measures that mean something.
Every scenario is built from the same four files. Change the files and you have a different exercise, not a different scenario.
Before two plans can be compared, each has to satisfy one identity:
Σ revenue(accounts) = Σ revenue(reps) = Σ revenue(regions) = company total
and one constraint:
every account is assigned to exactly one rep
A scenario that assigns an account twice inflates two territories. One that drops an account deflates the company. Neither can be compared with a plan that does not have that error, because the difference between them is the error, not the design. Assert both on every scenario before looking at anything else.
Once two plans reconcile, compare them on the measures that describe what the plan is for.
| Measure | What it says | Computed as |
|---|---|---|
| Coverage | Can the reps actually reach the accounts | Accounts within capacity ÷ accounts assigned |
| Revenue at plan | What the plan is worth if quotas are hit | Σ quota, from potential |
| Accounts per rep | Workload spread | Min, median, max across reps |
| Potential per rep | Opportunity spread | Min, median, max across reps |
| Disruption | How many relationships move | Accounts whose rep changed |
Fairness is the spread measures. A plan with a tight spread of potential per rep is fair; one with a wide spread has winners and losers before the year starts.
A team with six reps in the Northeast and six in the Midwest. The Midwest has grown; the Northeast has not. Plan A is the current assignment. Plan B moves two reps from Northeast to Midwest and rebalances quota from potential.
| Measure | Plan A | Plan B |
|---|---|---|
| Reps in Midwest | 6 | 8 |
| Accounts per Midwest rep | 88 | 66 |
| Midwest coverage at capacity 70 | 71% | 89% |
| Quota per Midwest rep | $1.2m | $1.0m |
| Revenue at plan, company | $7.2m | $8.0m |
| Potential per rep, spread | $0.6m to $1.9m | $0.9m to $1.4m |
| Accounts that change rep | 0 | 176 |
Both reconcile to the same 1,056 accounts and the same company potential. Plan B covers more, spreads potential more evenly and is worth more at plan. It moves 176 relationships, which is the cost. That is a decision the planning meeting can make, because both numbers are on the table and both plans add up.
Different account lists per scenario. An analyst removes a few dormant accounts in scenario three and the totals no longer match scenario one. Freeze the list at the start and mark dormant accounts as a flag, not a deletion.
Potential that moves with the rep. If potential is estimated per rep rather than per account, moving the rep moves the potential and every scenario flatters itself. Estimate potential per account from the account's own attributes.
Quota set by hand after the fact. A plan whose quotas were typed in cannot be compared with one whose quotas were computed. Use one rule for quota from potential across every scenario, and adjust by hand only at the end, visibly.
No disruption measure. The best plan on every other measure can be the worst to implement. Count the moved relationships and put it on the table.
When a scenario is built by asking, "move two reps from the Northeast to the Midwest and rebalance quota", and comes back reconciled with the comparison table, the planning meeting stops being a review of one workbook and becomes a session that tries ideas until the measures stop improving. Covirage builds scenarios that way, asserts the identity on each, and writes the chosen plan back to the CRM with the scenarios that lost kept for next year. The territory planning page describes it.
Fair is measurable: the spread of accounts per rep, revenue per rep and quota attainment at plan across the team. A plan where the best and worst territory differ by ten percent in revenue potential is fair enough for most teams. One where they differ by a factor of two is a plan that will lose the reps at the bottom.
As many as the questions you have, which is usually five to fifteen. The constraint used to be that each one took an analyst a day. When a scenario takes minutes, the planning meeting can try ideas live and stop when the fairness measures stop improving.
The territory. Quota is a property of the accounts and their potential. A rep who moves takes their skill with them, not their number. Set quota per territory from potential, then assign reps.