How to build an AR dashboard CFOs will actually use

A good AR dashboard leads with DSO and its trend, the aging mix, CEI, total and overdue receivables, and the bad debt ratio, then lets a CFO drill into the accounts driving each number. The best ones are live and action-linked, updating as cash applies, not a static BI export refreshed once a month.

A dashboard is only useful if it changes what someone does. Most AR dashboards fail that test: they look thorough, get admired at month-end, and drive no action. Here is how to build one that does.

What belongs on an AR dashboard

Start from the questions a finance leader asks, then put exactly those answers on screen. Three questions cover most of it. How fast is cash arriving. How much of the book is at risk. Which accounts are causing both. Anything that does not help answer one of those is clutter.

Resist the urge to show every field your ERP exports. A dashboard with 40 tiles is a report nobody reads. Pick the dozen metrics that move a decision and make them legible. For the metric definitions, see our accounts receivable metrics and KPIs guide.

It also helps to know who the dashboard is for, because a CFO and a collections lead want different things from the same data. The CFO wants the cash and risk story at a glance: is DSO trending the right way, how much is at risk, what is the expected cash. The collections lead wants the worklist: which accounts, how late, what to do next. A dashboard that tries to serve both as one undifferentiated grid serves neither. Build the layers so each person finds their view fast.

Core metrics and views

Build it in three layers.

Headline cash and risk (top of the dashboard).

• DSO with a rolling trend line, not just today's value.

DSO = (Accounts Receivable / Total Credit Sales) x Number of Days

• Best possible DSO alongside it, so the gap is obvious.
• Percent of receivables current.
• Expected cash from collections over the next 30 and 60 days.
• Bad debt ratio.

Operational mix (middle).

• The aging mix as a stacked bar: current, 1 to 30, 31 to 60, 61 to 90, 90-plus, shown as percentages of the book.
• CEI for the period.

CEI = ((Beginning AR + Credit Sales - Ending Total AR) / (Beginning AR + Credit Sales - Ending Current AR)) x 100

• Disputes outstanding and average days to resolve them.

Accounts (bottom, drill-down).

• A ranked list of top overdue accounts by balance and days late.
• A worklist of invoices that just crossed an aging threshold.

The top layer is for the CFO. The bottom layer is for whoever does the chasing. The middle ties them together.

A worked read of the top layer shows why this ordering helps. Say DSO reads 41 and trending up, best possible DSO reads 28, percent current reads 71, and expected 30-day cash reads 1.2 million. In one glance the story is clear: there is a 13-day gap to the floor, the gap is widening, and nearly a third of the book is already past due. The CFO does not need to ask a single follow-up question to know collections need attention, and the next click into the aging mix and top-overdue list shows exactly where. That is what a dashboard is for: turning a ledger into a decision in seconds.

Design principles for clarity

A dashboard a CFO actually uses follows a few rules.

1. One screen, no scrolling for the headline numbers. If it does not fit, it is not prioritized.
2. Trends over snapshots. A single DSO figure means little; the line over six months means everything.
3. Show the gap, not just the value. DSO next to best possible DSO. Actual CEI next to target.
4. Use color for exception, not decoration. Red should mean act, and it should be rare.
5. Round to the precision that matters. A CFO does not need DSO to two decimals.
6. Label every metric with its definition on hover, since people read AR metrics differently.
7. Date the data. A small as-of timestamp stops every meeting from opening with the question of how fresh the numbers are.

A common failure is precision theater: dashboards crowded with decimals, sparklines, and gauges that look sophisticated and say nothing. Sophistication in an AR dashboard is the opposite. It is the discipline to show five numbers a CFO can act on instead of fifty they cannot. If a tile has never once changed a decision, remove it and the dashboard gets better.

Static reports vs real-time dashboards

A BI export refreshed monthly is a photograph of a ledger that has already changed. By the time the deck circulates, the 30-day bucket it shows has aged into 60. The numbers are real but late, and late numbers drive late action.

A real-time dashboard reads DSO, aging, and CEI as payments apply and invoices age. The 60-plus bucket lights up the day an invoice crosses into it, while there is still time to do something. That timing difference is the whole value of the dashboard.

There is a second cost to the static version that teams underrate: the build itself. Someone exports from the ERP, cleans it in a spreadsheet, refreshes the BI model, and circulates a deck. That is hours of skilled finance time every cycle spent producing a snapshot that is stale before it lands. A live dashboard wired to the source removes the manual build entirely, which is both cheaper and more accurate, since there is no hand-cleaned export to introduce errors.

Connecting the dashboard to action

A number on a screen recovers no cash. The dashboard has to link to the next step. When the 90-plus bucket rises, a CFO should drill straight to those invoices and the accounts behind them, then hand them to collections or trigger follow-up directly. A dashboard that only displays a red figure, with no path from the number to the action, is decoration. The point is to shorten the distance between seeing a problem and working it.

The best test of a dashboard is to ask, in front of any tile, what would I do if this number turned red. If there is no clear answer, the tile is reporting for its own sake and should go. If the answer is clear but reaching the underlying accounts takes another three tools and a spreadsheet, the dashboard is leaking the value it just surfaced. Every important number should be one click from the accounts driving it and one more from the action that resolves them.

From dashboard to autonomous AR follow-up

Rex is an autonomous AR agent, and it closes the last gap a dashboard leaves open. Rex keeps every metric on the dashboard live, DSO, aging mix, CEI, expected cash, as payments apply and invoices age. Then it acts on what the dashboard shows. When an invoice crosses an aging threshold, Rex follows up. When a payment lands, Rex applies it and the numbers update without a human touching them. When a dispute opens, Rex works it toward resolution.

This is the part a traditional dashboard cannot do on its own. A BI tool can show you the 90-plus bucket is rising, but a person still has to notice, open the accounts, and write the emails. Rex collapses that chain. The number and the action are the same event: the bucket rises because an invoice aged, and Rex is already following up on it. The dashboard becomes a window onto work in progress rather than a backlog of work to start.

So the dashboard stops being a report you read and react to. It becomes the live view of work Rex is already doing across the whole ledger, surfacing only the accounts that need a human decision.

See how Rex turns a live AR dashboard into collected cash, end to end.