Owner’s dashboard: cash, sales and risks on one screen

Why an owner needs one screen and why numbers differ

When the owner doesn’t have a single clear screen, meetings quickly turn into arguments about whose numbers are “right”. Sales shows one thing, finance another, operations brings a third report. Time is spent not on decisions but on figuring out who counted what and how.

Numbers diverge for simple reasons: different sources (CRM, bank, 1С, Excel), different dates (by payment or by shipment), different rules (including VAT or not) and different levels of detail (some look at the company level, others by branch or project). Even the word “revenue” can mean different things in different reports.

A management dashboard differs from accounting reports because it’s not for compliance but for decision-making. Accounting is responsible for correctness under accounting standards, while the owner cares about something else: are we making money, where are we slipping, and what could hit cash in the coming weeks.

Most arguments revolve around three topics:

• revenue: by invoices, by shipment, or by actual cash received
• margin: what was included in cost and which discounts were applied
• liabilities: what counts as overdue and who’s responsible for stuck payments

The “one screen” format is not an attempt to fit every metric in the world. It’s a short set of numbers you check regularly (daily or weekly) and that make it clear what to do next. A good owner dashboard doesn’t explain the past across 20 pages; it quickly highlights deviations and the questions to resolve today.

This is especially important in companies with mixed models (sales, projects, manufacturing, integration), where a single deal moves through many systems and stages. Without a unified screen, it’s easy to lose the big picture.

Decisions first, metrics second

A dashboard exists to speed up decisions, not to look pretty. So start not with metrics, but with a list of decisions you actually make. Otherwise you get a “kitchen sink” screen: numbers with no actions.

Write down 8–12 regular decisions: some weekly, some monthly. For example: where to hire or cut people, which sales channels to push, which payments to postpone, which projects to speed up or stop, which clients need changed terms (limits, prepayment).

Turn each decision into a short question the dashboard answers in three minutes. “Will we have enough cash until month end?”, “Are we growing without margin erosion?”, “Are there clients dragging us into a cash gap?”. If you can’t state the question simply, the metric will probably be contentious.

A good filter: a metric is unnecessary unless there’s a clear action attached to it. “Number of calls” is useless if you don’t change scripts, reassign leads, or manage the funnel based on it.

Also think about horizons. For cash, week and month matter more (to spot gaps in advance); for sales, a week often suffices to catch trends without noise. Leave daily numbers for decisions where you can actually act today.

Cash metrics: the minimum that actually helps

Cash on the dashboard isn’t for pretty charts. You need to quickly understand two things: will there be enough funds in the coming weeks, and what is driving the change.

Start with a simple cash flow view: receipts, payments and account balances. Next to that it’s useful to show “net flow for the week/month” and “ending balance”. Paper profit can grow while cash declines.

To avoid confusion between margin and profit, keep just two levels side by side: gross margin (after COGS) and operating profit (after main expenses). Add more profit types and meetings will quickly become about which number is the “real” result.

Receivables and payables should also be on one screen, but in a form that prompts action. Typically enough are: overdue amount and share of top-3 clients (risk concentration), average days to pay (DSO) and a 4–8 week trend, amount due to suppliers in the next 7–14 days, share of overdue supplier payments and dynamics.

Plan-vs-actual is not needed everywhere. Plan what you truly control (expenses, payroll (FOT), investments). For revenue, sometimes trend and period-over-period comparison suffice if sales are volatile.

Early signals of a cash gap are simple: balance falls for 2–3 periods in a row, receivable overdue rises, and upcoming payments in the next 14 days already consume more than half of the current balance. For example, if receipts are stable but DSO rose from 28 to 45 days, the gap usually arrives not suddenly but in a couple of weeks. You can see this in advance.

Sales metrics: how to see growth without noise

For sales to support decisions rather than debates about “why your numbers differ”, show not everything but only what explains growth or decline. The owner needs dynamics and causes more than every manager-level detail.

Start with revenue in clear breakdowns: product, channel, region, key clients. This shows where growth is real (a specific product increased) and where it is accidental (one large deal skewed the month). Also show share of top-5 clients: it immediately reveals concentration risk.

A funnel without extra layers

The funnel shows where the flow is constricted: leads → meetings → proposals → deals. Look at volumes and conversion rates between steps. If there are many leads but few meetings, the issue is often lead quality or response speed, not sales performance alone.

To read the funnel quickly, keep 4–5 numbers on the screen: leads (and their cost if paid channels exist), meetings, proposals, deals and conversion from meeting to deal.

Average ticket, frequency and speed

Revenue growth can come from higher average tickets, more frequent purchases, or new customers. When average ticket falls, check the share of discounts: “revenue growth” is often bought by margin.

Sales speed directly affects cash. The deal cycle length (from first contact to payment) shows when revenue turns into cash. Example: if the cycle grew from 20 to 35 days, revenue targets may be met but cash gaps will become more frequent.

For sales quality keep nearby the share of discounted sales, returns and cancellations. If returns rise, apparent growth may be noise that turns into negative results tomorrow.

Risk metrics: what should alert you early

Risk metrics are not scare tactics but ways to spot a problem before it becomes a cash gap, a broken contract or a lost client. A good rule: a risk metric must answer “what do I do if it’s red?”.

Usually 4–6 signals updated frequently with clear thresholds are enough. Examples: revenue concentration (share of top-1 and top-3 clients or suppliers), overdue on key obligations (share of late orders and average days overdue), team load (overload in narrow roles and idleness in others), critical incidents (IT, production, logistics downtime), quality (complaints, defect rate, repeat issues).

Example: if the share of top-1 client rises and at the same time delays occur on their orders, these are usually not two separate problems. Often one bottleneck (people, supply, testing) failed under load.

To reduce arguments at meetings, set thresholds in advance: green/yellow/red ranges, metric owner (who acts when red), update frequency (day/week), and a short “reason/plan” field next to the number.

Data sources: where to get numbers and whom to trust

A dashboard rests on clear sources, not charts. If the company has many systems, numbers will differ until you agree which data is “official” and who owns it.

Financial dashboards usually use data from 1С or ERP (revenue, COGS, receivables), CRM (funnel and deals), bank (payments and balances), warehouse (stock and turnover), HR systems (headcount, payroll (FOT), turnover). Even within one group there may be different versions of 1С, and that’s already a risk.

Assign an owner for each source — a business role, not IT. Sales head owns CRM, chief accountant or CFO owns 1С, logistics owns warehouse, treasury owns bank data, HR owns personnel data. The owner confirms entry rules and resolves data quality questions.

To avoid arguments, record in one place: where each metric comes from (system and report), the source owner and a backup, how often it’s updated, and allowed delay (for example, bank — same day, 1С — after day close).

If some data lives only in Excel, admit it and make one master file, ban copies and enforce simple version control. Plan migration to a system, but don’t wait for perfection.

Minimum checks that catch most errors: reconcile revenue and payments with bank statements and primary documents, monitor anomalies (margin spikes, negative balances, “missing” deals), spot-check 5–10 transactions weekly, compare with last week and the same period last year.

Unified definitions: stop arguing about formulas

If meetings turn into arguments about whose number is right, the problem is almost always definitions. A dashboard works only when metrics have one meaning for everyone and that meaning is locked in.

Mini dictionary of metrics

Start with a dictionary of 10–15 key indicators. Don’t overcomplicate: agreement is more important than perfect math.

Typically you fix first:

• revenue: by payment or by shipment (and why this option was chosen)
• margin: gross or operating, and which costs are included
• active clients: what counts as active and over which period
• returns and discounts: whether they reduce revenue and how they are allocated by period
• VAT and currency effects: show with or without VAT and where FX is accounted for

Also agree on dates. “Sale date”, “payment date” and “shipment date” give three different pictures. For example, at month end you may ship a lot and receive payment next month. If not fixed, sales will claim growth while finance claims a cash shortfall.

How to change formulas without chaos

You need a “single version of data”: an approved set of rules and sources used for reporting. Usually the owner or CFO approves it, and a responsible person (analyst or BI lead) stores and updates it.

Treat formula changes like product changes: effective date, reason, calculation example and whether history is recalculated or left as-is. This improves metrics without breaking trust in past periods.

Step by step: how to build the first dashboard

The first owner dashboard doesn’t have to be perfect. It must answer key questions quickly and be computed the same way for everyone so meetings focus on actions, not numbers.

A practical five-step cycle:

• Gather 10–15 owner questions and turn them into 8–12 KPIs with clear calculation logic.
• For each KPI pick 1–2 sources and write rules immediately: who owns the metric, how we calculate it, when it’s updated, what to do on discrepancies.
• Sketch a one-screen draft: top row for cash, below sales, right side 2–3 risks. Agree on the layout before automating.
• Set up updates and access: who views, who edits, how changes to formulas and sources are recorded.
• Run 2–3 meetings strictly using the dashboard, collect questions and only change what obstructs decisions.

Small example: in a manufacturing or systems integrator company owners often argue whether “sales” are by shipment or by payment. In the first dashboard you can show both numbers but make one the primary and display its definition next to the KPI. In companies at the scale of GSE.kz such agreements are especially important: the more supply chains, projects and accounting systems, the easier it is to lose a consistent meaning of the numbers.

A simple success criterion: after the meeting you have 2–3 concrete decisions and a clear answer which numbers supported them.

Common mistakes that break a dashboard

Most dashboards fail for one reason: they are built as a showcase of all numbers at once. They stop helping decisions and become a source of arguments about whose table is right.

Usually the problem isn’t the BI tool but data discipline and rules. Frequent mistakes:

• too many indicators without priorities
• mixing cash and accruals (for example, showing payment-based revenue next to accrual-based expenses)
• no metric owner and no source owner
• metrics without context (no period, comparison or short reason for deviation)
• manual edits over data that erode trust

Example: the owner sees margin at 28% while the CFO says it’s 19%. One report treated discounts as a revenue reduction, another as a marketing expense. Both are “correct” under different rules. Without a single definition the argument repeats every week.

A good test: if you can’t explain in 30 seconds what a metric means and where it comes from, it doesn’t belong on the main screen.

Pre-meeting checklist: quick checks in 2 minutes

Meetings turn into disputes when people look at different numbers or data of different freshness. A short pre-meeting check is enough to focus the discussion on decisions, not tables.

First check data hygiene. The screen should show last update time and source. If updates are delayed (sales already happened but cash and shipments aren’t pulled), say so and don’t draw conclusions from incomplete data.

Then quickly assess deviations. Plan-vs-actual should highlight automatically, and next to each major deviation there should be a one-line reason, not a 10-minute explanation. For example: “margin fell due to discount on Project X” or “revenue increase due to one-off large shipment”.

A short checklist that fits in 2 minutes:

• data updated on schedule and dashboard shows “updated at…”
• plan-vs-actual highlighted and each big deviation has a short reason
• top-3 drivers of revenue and margin are clear immediately (channel, product, key client)
• overdue receivables and limits visible without switching tabs
• 1–2 main risks for the next 1–2 weeks already have an owner, an action and a deadline

If any item fails, start the meeting with one question: “What needs fixing in data or definitions so the same argument won’t happen tomorrow?” Then the financial dashboard becomes a decision support tool, not an excuse.

Example scenario: one week with the owner’s dashboard

Monday. The morning call shows something odd: last week revenue is down 8% while number of deals is up 12%. Previously this triggered debates about who’s counting what. The owner’s dashboard shows three hints together: average ticket, share of discounts and share of returns.

Tuesday. You drill down and see average ticket fell 15%. It’s not a manager issue: discounted deals rose (from 9% to 18%) and the product mix shifted toward cheaper, lower-margin items. Returns also rose slightly and hit revenue harder than the count suggests.

Wednesday. The meeting focuses on what to change, not why sales fell. Decisions split between sales and finance. Sales limits discounts and refocuses the product mix. Finance checks adjustments and speeds up return handling.

Thursday. Agreements are recorded as tasks to avoid repeat disputes:

• Limit discounts above 10% without approval: owner — head of sales, deadline — Friday.
• Rebuild product plan (mix): owner — commercial director, deadline — Wednesday.
• Analyze top-10 returns and causes: owner — head of service/quality, deadline — Tuesday.
• Control metrics: average ticket, share of discounts, share of returns, gross margin — daily.

Friday. You watch the same four metrics. If discounts fall but average ticket doesn’t rise, the issue is deeper (price, packaging or lead quality). If returns climb, it becomes a priority even if revenue temporarily holds.

Next steps: lock the result and avoid backsliding

After the first launch the main goal is to build trust in the numbers and the habit of using them. Otherwise in a month the dashboard becomes a pretty image opened only before the meeting.

Agreements that prevent 80% of future disputes:

• Fix the final KPI set for the one screen and assign an owner for each indicator.
• Create a source map: where each number comes from, how often it’s updated and who verifies quality.
• Define update rules: what is pulled automatically and what may be adjusted manually (and how this is marked).
• Plan reliability: where data is stored, who has access and what to do if systems fail.
• Set a rhythm: 10 minutes weekly to check data quality and a monthly review to ensure no unnecessary metrics appear.

Expand the dashboard not when you want another chart, but when a new type of decision appears that can’t be made without new metrics.