A sales leader inherits a region, splits it down the middle, hands one half to each rep, and sets both of them the same number. It feels even-handed. By March one rep is 140% to plan and coasting; the other is at 55%, demoralised, and updating their CV. Nothing was wrong with the people. The map was wrong, and the quota inherited the map's mistake.

The quick version

  • Segment first, then draw lines, then set numbers, in that order. Quotas are the last decision, not the first.
  • Balance territories on opportunity and workload, not on map area or account count. A "fair" split of geography is usually an unfair split of money.
  • A good quota is one most reps can reach with effort. Aim for roughly 60% of the team at or above plan; if almost everyone misses, the problem is the quota, not the team.
  • Re-cut once a year, gently. Territory churn destroys relationships and pipeline; redesign on a schedule, not on a whim.

The idea in depth

Most teams treat territory, segment, and quota as three separate admin chores handled at three different times of year. They are one chain of decisions, and it only holds if you work it in order. Segmentation decides who is worth selling to. Territory design decides who sells to them. Quota decides how much we expect back. Skip to the quota, the usual mistake, and you're setting a target before you know what the territory can actually produce.

Segment on behaviour and value, not just firmographics

Segmentation is the habit of grouping customers so that the customers inside a group want roughly the same thing and are worth roughly the same effort. The lazy version sorts accounts by size band alone: enterprise, mid-market, SMB. That's a start, but size is a proxy. What you're really after is two things at once, how much an account is worth (potential spend, lifetime value) and how it buys (long committee-driven cycle versus a quick self-serve sign-up). Two companies with identical headcount can sit in completely different segments because one needs a six-month enterprise motion and the other will swipe a card after a demo.

So the move is to score your account list on two axes, value and effort-to-win, and let the quadrants define your segments and the selling motion each one gets. High-value, high-effort accounts earn a named rep and a real plan. Low-value, low-effort accounts get a lighter touch or a self-serve path. This is also where territory design connects to your wider go-to-market motion: a product-led segment and a sales-led segment should not be carved on the same map.

One caveat worth saying out loud: segmentation is a model, and models go stale. The account that was a sleepy SMB last year just raised a round and now buys like an enterprise. If you only re-segment when you redesign territories, you'll spend months selling the wrong motion to a moved target. Treat the segmentation as a living view, refreshed far more often than the territory map.

Balance territories on opportunity and workload, not on the map

Here's where the research is unusually clear and unusually ignored. Andris Zoltners and Sally Lorimer, in Sales Territory Alignment: An Overlooked Productivity Tool (Journal of Personal Selling & Sales Management, 2000), drew on work with hundreds of sales forces to argue that territory misalignment is one of the most common and most expensive sales-productivity problems, and one leaders rarely look at. The mechanism is simple. If a territory holds more opportunity than one person can work, that opportunity goes uncovered: too few leads chased, too little time with the accounts that matter. If a territory holds too little, you're paying full freight for a rep who runs out of road by summer. Either way you lose money you can't see on any report.

The fix is to design territories so each one carries a comparable load of workable opportunity, where opportunity blends two things: the potential in the accounts (their likely spend) and the workload to serve them (calls, visits, travel, account complexity). A territory with twenty whales is not equivalent to a territory with twenty corner shops, even if the account count matches. Splitting a region by postcode or by headcount feels fair and almost never is.

A fair split of the map is usually an unfair split of the money.

So stop balancing on the things that are easy to count, area, number of logos, and start balancing on a rough opportunity score per account, summed per territory. You don't need a data-science team. A spreadsheet with two columns, estimated potential and a workload weight, gets you most of the way. When the per-territory totals land within shouting distance of each other, the quotas that follow will feel fair, because they'll be fair.

The catch is that perfectly balanced territories on paper can still be unfair to a person, because the maths ignores relationships and ramp. A rep who has spent three years building trust in a patch shouldn't lose half of it to a rebalancing exercise that looks neat in a model. The pull toward team trust and the pull toward optimal coverage point in different directions here, and the redesign that's cleanest on the spreadsheet can be the most corrosive to the team. Balance the numbers, then sanity-check against disruption before you ship the new map.

flowchart TB
    A(["Total market / TAM"]) --> B(["Segment: value x effort-to-win"])
    B --> C(["Draw territories: balance OPPORTUNITY + workload"])
    C --> D(["Set quotas off territory capacity"])
    D --> E(["Pay plan links to quota"])
    E -.->|"re-segment often, re-cut once a year"| B
The chain runs one way: segment, then carve, then quota. Quota is the last link, not the first. Leaders Loop

Set quotas reps can actually reach

A quota does two jobs at once: it tells a rep where to aim, and it usually decides what they get paid. That double duty is exactly why bad quotas are so damaging, get it wrong and you've simultaneously misdirected effort and broken trust in the pay plan. Zoltners, Lorimer and Sinha lay out the tension plainly in 7 Ways Sales Teams Can Set Better Goals (Harvard Business Review, 2019): set goals too high and motivation collapses; set them too low and you pay out for results you'd have got anyway. The target has to sit in the band where stretch still feels achievable.

What does "achievable" mean in numbers? Industry data is sobering. Salesforce's State of Sales research, widely reported across recent editions, has repeatedly put average quota attainment in the rough range of only about half the reps hitting their number in a given period. That is the symptom of quotas set off wishful top-down targets rather than territory capacity. By contrast, the Alexander Group, a go-to-market consultancy, argues a healthy distribution looks like a bell curve where roughly 60% of reps land at or above plan and about 40% below. If almost nobody clears the bar, you haven't found a weak team; you've found a broken quota.

So build quotas up from territory capacity rather than down from the board's growth wish. Start with what each balanced territory can realistically produce, last year's base, plus the workable opportunity you scored in the design step, then apply growth. Where the bottom-up total falls short of the company target, that gap is a real business problem (hire more reps, raise prices, enter a segment), not something to paper over by inflating every rep's number until the spreadsheet adds up. Once quotas are set, check the resulting distribution against that 60/40 shape. If it's wildly off, you've learned something before the year starts rather than after it's lost.

Hold that 60/40 figure loosely, though. It's a practitioner benchmark, not a law of nature, it shifts by industry, deal size, and pay-plan design, and a team selling huge, lumpy enterprise deals will have a noisier distribution than a high-volume transactional one. Use it as a smell test, not a target to engineer toward.

A worked example

Take a fictional company, call it Northwind, a B2B software firm with two reps covering one country. (All figures here are illustrative, to show the method, not real benchmarks.) The outgoing plan split the country east/west and gave each rep a $1.2m quota. East is hitting 130%; West is at 60% and the rep is quietly interviewing elsewhere.

The leader runs the chain properly. Segment: she scores every account on potential spend and effort-to-win, and discovers the east holds three large, slow-cycle accounts worth most of its revenue, while the west is a long tail of smaller, faster deals. Same map area, completely different work. Territory: instead of east/west, she re-cuts so each rep carries a comparable opportunity score, the eastern whales are split, and a block of mid-tier western accounts moves to balance the load. The new territories come out at roughly $1.0m and $1.1m of workable opportunity. Quota: she sets numbers off those capacities, about $0.95m and $1.0m, modest growth on a realistic base, not a flat $1.2m pulled from the annual plan.

The outcome isn't magic. One rep's quota actually drops, and the leader has to sit with the board and explain that the old $2.4m combined target was never grounded in territory capacity, that's the real conversation, surfaced in January instead of October. But both reps now believe their number, the pay plan feels honest, and the West rep stops reading job ads. The point of the exercise was never a tidier map. It was a quota two people would actually run at.

flowchart LR
    subgraph OLD["Before: split the map"]
      W1(["West rep · $1.2m quota · 60%"])
      E1(["East rep · $1.2m quota · 130%"])
    end
    subgraph NEW["After: balance opportunity"]
      W2(["West rep · $0.95m · believable"])
      E2(["East rep · $1.0m · believable"])
    end
    OLD ==>|"re-segment + re-cut"| NEW
Illustrative only. The map looked fair; the money didn't. Balancing opportunity fixed both the attainment gap and the trust gap. Leaders Loop

Frequently asked questions

Should I balance territories by geography or by accounts?

Neither, on its own. Balance on workable opportunity, the potential spend in the accounts weighted by the effort to serve them. Geography and account count are easy to measure and usually misleading; a compact patch of large accounts can hold far more opportunity than a sprawling region of small ones.

How often should we redraw territories?

Re-segment your accounts frequently, quarterly is common, because accounts move between segments fast. But redraw the actual territory boundaries sparingly, ideally once a year on a known schedule. Frequent re-cutting destroys the relationships and pipeline a rep has built, which costs more than the imbalance you're trying to fix.

What if the bottom-up quotas don't add up to the company target?

Then you've found a real gap, and that's useful. Don't close it by inflating everyone's number until the maths works, that just guarantees misses and resentment. Close it with business levers: more reps, higher prices, a new segment, or a revised company target. Surfacing the gap early is the whole value of building quotas from capacity.

Isn't a quota everyone hits just a quota that's too low?

If everyone sails past it, yes. But the opposite failure is more common and more expensive: a quota almost nobody hits demotivates the whole team and tells you nothing about who's actually good. A healthy spread has most of the team at or near plan and a tail above and below it. Aim for stretch that's reachable, not stretch that's punitive.

How does this connect to commission and pay?

Tightly, which is why fairness matters so much. When the quota is tied to pay, an unbalanced territory isn't just a coverage problem, it's a pay-equity problem, and people leave over it. Design the territory and quota fairly first; the comp plan then sits on a foundation people trust. See commercial unit metrics for the economics underneath the targets.

Related in the Toolkit

Where to go next