Reference
Metrics
Compass computes 22 business metrics on every scan — deterministically, from your Xero data, adapted to service-business language. Every number is unit-tested; nothing is invented by the AI. Grouped by decision surface below.
Revenue / growth
How the top line is behaving — recurring, growing, retained, expanding.
Monthly recurring retainer (MRR)
What it measures. Predictable revenue that repeats each month, from clients on a regular billing cadence. For services, this is your retainer income; for SaaS, subscription income.
How it is computed. Sum, across clients who invoiced 3 or more times in the selected period, of their average monthly billings. Non-recurring one-offs excluded.
What healthy looks like. No universal threshold — meaningful in comparison to your total revenue. A book that is 60%+ MRR is structurally more stable than a 40% MRR book of the same size.
Annualised retainer revenue (ARR)
What it measures. Projected annual value of your current retainer book — a forward-look version of MRR.
How it is computed. 12 × MRR.
What healthy looks like. Same framing as MRR — no absolute threshold. Growth-over-time is the useful measure, not the absolute number.
Revenue growth (month over month)
What it measures. How this month's sales invoices compare to last month's. Positive means growing; negative means contracting.
How it is computed. (revenue this month − revenue last month) / revenue last month × 100. Uses invoice.total (gross of receipt) for both months.
What healthy looks like. For a services business: >= 0% is a "not shrinking" pass; 0-2% is flat; 2%+ is compounding growth. For SaaS the bar is materially higher.
Average revenue per client (ARPC)
What it measures. How much revenue you extract per active client on average. Anchors your capacity assumptions.
How it is computed. Total revenue in the selected period ÷ number of clients with at least one invoice in-period.
What healthy looks like. No universal threshold. Meaningful in trend and comparison — if ARPC rises, you are either raising rates or growing existing accounts; if it falls, you are either discounting or growing the long tail.
Net revenue retention (NRR)
What it measures. Revenue this period from clients who were also present in the prior comparable period, divided by their prior-period revenue. Above 100% means expansion is outpacing contraction across your existing book.
How it is computed. For every client with revenue in the prior window: sum current-window revenue, sum prior-window revenue, then current ÷ prior × 100.
What healthy looks like. >= 110% is best-in-class expansion. 100-110% is holding steady. Below 100% means you are losing more from existing clients than you are gaining.
Gross revenue retention (GRR)
What it measures. Same as NRR but capped at 100% per client — expansion cannot offset losses. The "no-fluff" retention number that investors watch.
How it is computed. Per client, min(current-window revenue, prior-window revenue) ÷ prior-window revenue × 100.
What healthy looks like. >= 95% is best-in-class. 90-95% means you are bleeding some existing revenue. Below 90% means the base is genuinely shrinking before expansion is counted.
Compound monthly growth rate (CMGR)
What it measures. The average monthly growth rate that would explain revenue moving from the first month of the window to the last. A cleaner trend measure than raw month-over-month because it averages out one-off spikes.
How it is computed. (revenue in last month ÷ revenue in first month) ^ (1 ÷ n) − 1, where n is the number of complete months between them.
What healthy looks like. For a services business, 2% CMGR is comfortable growth. For SaaS the bar is 5%+. Negative CMGR means revenue is compounding downward — a slow but persistent contraction.
Bookings
What it measures. Total value of sales invoices raised in the selected period, gross of receipt. Deferred revenue growth is bookings-minus-billings.
How it is computed. Sum of invoice.total for ACCREC invoices with a date in-window.
What healthy looks like. No universal threshold. Bookings > billings signals a growing backlog; bookings < billings signals a shrinking one. Use the book-to-bill ratio (below) for the shape.
Billings (cash collected)
What it measures. Payments actually received in the selected period. Divergence from bookings is your true deferred-revenue position.
How it is computed. Sum of payment.amount for payments dated in-window.
What healthy looks like. No universal threshold. Match to bookings for a healthy through-put; sustained gap in either direction is worth understanding.
Growth vs churn ratio (quick ratio)
What it measures. New-client revenue this period divided by silent-churn revenue. Above 4 means you are winning growth faster than you are losing existing revenue.
How it is computed. Sum of first-invoice-in-window client revenue ÷ sum of currently-silent client historic revenue.
What healthy looks like. >= 4 is healthy expansion. 1-4 is barely replacing churn. Below 1 means you are losing revenue faster than winning it.
Customer
The shape of the client book — active, new, retained, quietly churning.
Active clients
What it measures. Number of clients with at least one sales invoice in the selected period.
How it is computed. Count of distinct contactIds with invoice.type=ACCREC and invoice.date in-window.
What healthy looks like. No universal threshold — meaningful in trend. Rising is good; falling is a leading indicator that silent-churn or slowing revenue-quality signals will surface next.
New clients this window
What it measures. Clients whose first-ever sales invoice fell inside the analysis window.
How it is computed. Per client, find the earliest invoice.date across all their ACCREC invoices; count them if that date sits in-window.
What healthy looks like. No universal threshold. A book that is not adding new clients at all is over-reliant on retention; a book adding lots of small new clients that never grow is over-reliant on volume acquisition.
Client silent-churn rate
What it measures. Share of tracked clients who have gone silent — a gap since their last invoice longer than the maximum of 90 days or 2× their historic median gap.
How it is computed. For clients with 2+ invoices historically: mark quiet if days-since-last-invoice > max(90, 2× median gap). Divide quiet by total tracked.
What healthy looks like. For services/agencies: < 10% is healthy, 10-25% is elevated, > 25% is a retention emergency. For SaaS: < 5% is healthy.
Client retention rate
What it measures. 100% minus churn rate. The share of tracked clients still active by silent-churn thresholds.
How it is computed. 1 − (churn rate as above).
What healthy looks like. >= 95% is best-in-class. 90-95% is acceptable but leaky. Below 90% means you are bleeding customers.
Client lifetime value (CLV)
What it measures. Rough projection of the total revenue you expect from a client over their entire relationship — a proxy, no cost side.
How it is computed. Per client, tenure (first invoice → most recent invoice, in months) × average monthly billing. Averaged across clients with revenue in-window.
What healthy looks like. No universal threshold — meaningful relative to your acquisition cost. A 3× CLV to CAC ratio is a common SaaS benchmark; services businesses usually don't track CAC precisely enough for the ratio to be meaningful.
Cash / working capital
Working capital — how long money is tied up, and how many months of runway.
Cash conversion cycle (CCC)
What it measures. The gap in days between money leaving (paying suppliers) and money coming back (clients paying). Positive means you are your clients' bank.
How it is computed. DSO − DPO.
What healthy looks like. Under 30 days is healthy for a service business. 30-60 days means cash is tied up longer than ideal. Over 60 days means you are effectively your clients' bank.
Days sales outstanding (DSO)
What it measures. Average days clients take to pay after you invoice them. Lower is healthier.
How it is computed. For every ACCREC invoice with a payment: days from invoice.date to payment.date. Weighted average across the window.
What healthy looks like. Under 30 days: clients are paying inside standard terms. 30-60 days: payment cycle slipping. Over 60 days: chronic late-payer pattern; treat as a collections risk.
Days payable outstanding (DPO)
What it measures. Average days you take to pay suppliers. Higher preserves your cash (within agreed terms).
How it is computed. For every ACCPAY bill with a payment: days from bill.date to payment.date. Weighted average across the window.
What healthy looks like. Above 30 days: holding cash long enough within normal terms. 15-30 days: paying a touch fast. Below 15 days: paying suppliers very fast — your cash could work harder for you.
Runway (months)
What it measures. How many months you can keep going with no new sales, based on cash on hand + outstanding AR against average monthly burn.
How it is computed. (bank cash + outstanding AR) ÷ average monthly burn. Burn = (in-window bills + posted payroll) ÷ months in window. Proxy — non-Xero costs not counted.
What healthy looks like. 12+ months is comfortable. 6-12 months is worth watching. Under 6 months is a priority.
Efficiency
The unit-economics view — concentration, book-to-bill, Rule-of-40 proxy.
Client concentration risk
What it measures. Share of turnover held by your single biggest client. A single-point-of-failure warning above 30%.
How it is computed. Revenue(top client, in-window) ÷ total revenue(in-window) × 100.
What healthy looks like. Under 20% is a resilient book. 20-40% is meaningful concentration — losing the top client would hurt. Over 40% is a single point of failure.
Book-to-bill ratio
What it measures. Bookings divided by billings. Above 1.0 means booking faster than collecting (growing pipeline); below 1.0 means backlog is shrinking.
How it is computed. Sum of invoice.total in-window ÷ sum of payment.amount in-window.
What healthy looks like. 0.9-1.15 means bookings and billings are tracking together. 0.75-1.35 is a meaningful gap worth understanding. Outside that range means bookings and billings are badly out of sync — check timing.
Rule of 40 (proxy)
What it measures. Revenue growth % plus an operating-margin proxy. Investors treat >= 40 as the health cut-off for venture-scale SaaS businesses.
How it is computed. Growth % (current window vs prior window) + margin proxy ((revenue − bills) ÷ revenue × 100). Margin here excludes payroll and non-Xero costs — a proxy.
What healthy looks like. >= 40 is venture-scale healthy. 25-40 is okay but not great. Under 25 means either growth or margins need work.
Where these live in the app
All 22 metrics are available on the dashboard\'s Metrics tab. The Overview tab surfaces the most-common four (Total revenue, Active clients, Sales invoices, CCC) as headline stat cards; the rest are one tab away.
Related
- Widgets — the visual dashboards that consume these metrics.
- Signals catalogue — the 24 decision-shaped analyses on top of the metrics.