Territory build: split accounts by fit → Sheets

The problem this solves

Territory design is one of those decisions that shapes a year of revenue and gets made in a hurry. The classic approach - split alphabetically, geographically, or by whoever shouts loudest - balances the number of accounts while ignoring what the accounts are worth. Two reps can hold lists of identical length where one list is dense with high-fit, in-market accounts and the other is a graveyard, and the difference shows up later as a performance gap that gets attributed to the people instead of the map.

Balancing territories properly means knowing each account's fit and current intent, which means the same data problem that undermines scoring undermines carving. The inputs live in the CRM at whatever freshness the last cleanup left them, and rebuilding them account by account for a territory exercise is weeks of analyst time nobody has between planning cycles. So the map gets drawn on stale data, defended in one heated meeting, and frozen until next year.

The fair version is mechanical once the inputs exist: score every account on live data, then split so each territory carries an equivalent weight of fit and intent rather than an equivalent row count. What makes it stick politically is transparency, meaning the assignment logic is written down, the balance is checkable, and RevOps signs off on a document instead of a vibe.

How the mission runs

1. Pull and score the account base. The agent pulls every account from HubSpot and scores it for ICP fit, layering current buying signals from Intent Data on top. Fit and intent are kept as separate dimensions, because a territory balanced on fit alone can still concentrate all the in-market-now accounts on one rep's list.
2. Define the balance criteria. You set what balanced means: equal aggregate fit weight, equal counts of high-intent accounts, geographic coherence if travel matters, and any hard constraints like named accounts that must stay with a specific owner. The agent treats these as the rules of the carve rather than suggestions.
3. Propose the three-way split. Accounts are assigned to three territories so each carries an equivalent weight of fit and intent under your constraints. The output includes the balance math per territory - aggregate score, high-fit count, high-intent count - so the claim of fairness is a checkable table instead of an assurance.
4. Write out the assignment logic. Alongside the mapping, the agent documents the logic itself: how accounts were scored, what the balance targets were, how ties and edge cases were resolved. This is the page that survives the territory meeting, because disputes get argued against a stated rule instead of a suspicion.
5. Deliver the mapping in Sheets. Everything lands in Google Sheets: one row per account with its territory, score components, and intent level, plus a summary tab with per-territory balance figures and the written logic. Built as the sign-off document, RevOps can filter, spot-check, and approve it - or move specific accounts and see the balance recalculate.

The prompt

This is the exact objective the agent receives. Swap the obvious placeholders for your own domain, segment or channel and run it as-is from the console, Slack, or the API.

What comes back

A Google Sheets workbook holding the complete territory proposal: every account mapped to one of three territories with its fit score, intent level, and score components on the row, a summary tab proving the balance per territory, and the assignment logic written out in full. Nothing is changed in HubSpot; the workbook is the sign-off artifact, and ownership updates happen after RevOps approves.

Make it yours

• Change the territory count or add a strategic-accounts pool that is carved first, with the remaining base balanced across the field team afterward.
• Weight the balance toward intent for a team in land-grab mode, or toward fit for a team working long enterprise cycles, and state the weighting in the logic page.
• Re-run it quarterly as a drift check: keep assignments stable but report how balance has shifted as accounts moved segments, so rebalancing is a data-backed conversation.

Frequently asked questions

Go deeper

• ICP definition with live data: from slideware to instrument
• Lead routing and scoring that reflect reality
• Governing autonomous marketing agents