Some of the most useful questions in a business fall between departments, which is why nobody owns them. Which manager has fourteen direct reports? Which sales rep is carrying half the revenue while another carries almost none? Is that capital project stalled, or just quiet? Which of our subsidiaries is nobody actually using? HR has part of the answer, finance has part, and the system administrator has the rest.
The Sonar edition of the prompt library went live this month, and among the new prompts are nine that go after those in-between questions. This post covers them in five groups: the org chart as data, who is carrying the number, capital spend, entities and currencies, and the shape of the close.
If you're new to this, Sonar AI is an AI agent that runs inside NetSuite. Every prompt in the library is a playbook that I engineered and tested against live NetSuite data, and you run it inside your own account, against your own records.
The Org Chart as Data
Employee Tenure and Org Topology reads the employee records and produces the picture that usually lives in someone's head. A tenure curve, with headcount by years of service. Span of control per supervisor, flagging anyone with more than twelve direct reports or fewer than three. Headcount and average tenure by department. A retention risk map showing where short tenure concentrates. And the anomalies: employees with no department, no supervisor, or both, which tend to be the ones that fall through every process. The prompt is explicit that restructuring proposals and span-of-control recommendations go to a person, and that a small team's statistics don't support conclusions about individuals.
Billable vs Non-Billable Hours Monitor is for services businesses, and it has a rule I like: never apply one utilization target to every role. A partner at 38% billable may be exactly where they should be. The prompt uses role-specific benchmarks, trends the billable ratio over time, and categorizes the non-billable hours so that you can see whether they're going to sales, internal projects, or nothing in particular. It also checks first whether the account has time entries at all, and stops with a clear message if it doesn't, rather than producing an empty analysis.
Who Is Carrying the Number
Sales Rep Conversion Efficiency follows each rep's pipeline through its stages, estimate to sales order to invoice, and compares each rep's conversion at each stage against the team median. That comparison is the point. A rep who converts estimates at half the team's rate has a different problem from one who converts orders to invoices at half the rate, and the prompt keeps the stages separate so you can see which it is. Individual performance conclusions are flagged for human review.
Sales Rep Workload Distribution answers the coverage question. It counts opportunities, transactions, revenue, and assigned customers per rep, and it identifies who is overloaded and who is underused. It also reports something I've found most companies have never measured: the attribution rate, meaning the share of revenue transactions that have a rep assigned at all, and the dollar value of the ones that don't. The unassigned bucket is usually bigger than anyone expects.
Capital Spend
Capital Expenditure vs Budget Analyzer pulls capital spending from the fixed-asset accounts in the general ledger, compares it to budget, and adds the part that a budget report leaves out: open commitments. A capex budget that's 60% spent with purchase orders outstanding for another 50% is over budget, and the prompt says so. It produces the executive summary, spend by asset account, the commitments schedule, the monthly trend, and a year-end projection from the run rate.
Capital Project Spend Tracker takes the same data at the project level and applies four health states with stated criteria. On track means spend percent is within five points of completion percent. At risk means it's ahead of completion. Over budget means spend plus commitments exceed the budget. Stalled means no spend in sixty days with budget remaining. Each state has an action, and "stalled" is the one I'd watch, because a stalled project is either done and never closed or stuck and nobody has said so.
Entities and Currencies
Currency Exposure Heatmap maps open receivables, open payables, and bank balances by currency and by subsidiary, to show unhedged exposure. It has a hard rule that sounds obvious and gets broken constantly: never sum open amounts across currencies, because the unpaid amount on a transaction is denominated in that transaction's currency. Any single-currency exposure above 10% of equity is flagged for review. If the account turns out to be single-currency, the prompt says so and reports that as the finding.
Subsidiary Activity Asymmetry is for OneWorld accounts, and it answers a question that comes up in every acquisition and every cleanup: which of these entities is real? It computes each subsidiary's share of activity across six dimensions, customers, vendors, revenue, spend, GL activity, and master-data freshness, and it turns the shares into an asymmetry index. A subsidiary with 2% of revenue but 30% of the customer records is worth a look. So is one that exists only because it was the default on a form. The prompt is careful about the limits of what a query can see, since the subsidiary on a transaction header isn't exposed to SuiteQL, and it says which lens each number came from.
The Shape of the Close
Period Close Crunch Profile is a prompt I don't think I've seen anywhere else. It plots transaction posting by day of the month across the trailing twelve months, for all types together and for journals, vendor bills, and invoices separately. Every controller knows the close is hectic. This shows where the load actually sits: an end-of-month spike from accruals and adjustments, or a day-one spike from reversals, and how intense the spike is relative to a routine day. The close-day load multiple is the number to watch, because intensity is what correlates with overtime and with errors. The recommendations are tied to specific days: which tasks could move earlier, which accruals could be scheduled, what could be automated. The prompt also includes its own caveats, such as the fact that a transaction dated the 30th may have been entered on the 5th.
Wrapping Up
What these nine have in common is that the data was always there. Supervisors are on employee records. Reps are on transactions. Commitments are on purchase orders. Posting dates are on everything. The questions went unasked because answering them meant joining three departments' worth of records, and nobody's job description included that. A prompt doesn't need a job description.
All nine are in the paid tier of the Sonar AI Prompt Library, under HR and Workforce Analytics, Budgeting, Forecasting and Planning, Multi-Entity and Consolidation, and Month-End Close and Reconciliation.