The owner's dashboard: which numbers to see every day so you stop managing blind

According to a survey by Aktion Finansy, a Russian publisher of professional resources for finance teams, 27% of Russian companies hit a cash gap in 2026 for the first time, or for the first time in a long while. Most of them find out after the fact — when there is already nothing to pay with. An owner's dashboard closes exactly that hole: the key numbers of the business are gathered from working systems on their own and show the state of the company today, not in last month's report.

Below is a working set of metrics with viewing frequency, an explanation of why dashboards lie and where the data should come from, and a step-by-step rollout plan.

What is an owner's dashboard?

An owner's dashboard is a single screen with the key metrics of the business, collected from working systems automatically and updated without human involvement. It answers three questions: how much money the company has, how sales are going, and where trouble is brewing — at the moment you open the screen. No "please send me the report" requests, no two-day wait, and no taking numbers on faith.

The classic definition was formulated by Stephen Few, author of the most-cited book on dashboard design:

"A dashboard is a visual display of the most important information needed to achieve one or more objectives, consolidated and arranged on a single screen so the information can be monitored at a glance" — Stephen Few, Perceptual Edge.

The contrast with reality is striking: according to a study by Moe Delo, a Russian online accounting service (a survey of 1,400 entrepreneurs, autumn 2025), 12% of small business owners keep their books "in their head", another 2% keep none at all, and 28% admit they struggle to control cash flow. A dashboard is the opposite pole of that scale.

Which numbers should you see every day?

The daily minimum is five blocks: cash with a payment calendar, revenue against plan, the state of the sales funnel, overdue obligations, and one or two operational metrics specific to your industry. Everything else — margin by line of business, customer churn, advertising performance — is looked at weekly or monthly. The frequency is set by a simple rule: how fast the metric can go bad and how much a day of delay costs.

The working set looks like this:

The list is a frame: a manufacturer will add shop-floor utilization to the daily block, a retail chain will add a location-by-location comparison, a service company will add specialist utilization. The principle is the same everywhere: what you look at daily is what dies fast.

How many metrics should be on the first screen?

Five to nine metrics is the limit at which a screen can be read at a glance; anything beyond that turns the dashboard into a report you have to study all over again. Detail hides on a second level: you notice a dip in revenue, click, and break it down by location, sales rep or product. The first screen answers "is everything all right", the drill-down answers "why".

Few's "single screen" requirement is a matter of principle: a metric you have to scroll to reach stops being looked at within a couple of weeks. If a twentieth number is asking to be on the first screen, that is a signal to revisit priorities — an owner never has twenty equally important metrics.

How is a dashboard different from an Excel report?

An Excel report describes the past and costs working hours: the finance manager assembles it by hand from exports, and by the time it is ready the numbers are stale. A dashboard updates without a person and shows the current day. The scale of that manual labour has been measured: according to an hh.ru survey, one in four employees in Russia spends more than six hours a week on reports nobody has automated.

If your accounting still lives in spreadsheets, it is better to wait on a dashboard: first check against seven signs whether it is time to leave Excel — a dashboard on top of scattered spreadsheets inherits every one of their ailments.

Building that loop — from data sources to screen — is something we deliver as a custom development project: we work out where your numbers are born, connect the systems, and take contractual responsibility for their accuracy.

Why do dashboards "lie"?

A dashboard lies when the data underneath it is entered by hand, entered late, or entered by different rules in different departments. The picture still looks convincing — and that is the main danger: decisions get made confidently on wrong numbers. The price has been measured: Gartner estimates that poor data quality costs companies an average of $12.9 million a year.

The usual causes repeat from project to project:

• Data is entered by hand and entered late — the screen shows yesterday's reality as if it were today's.
• The same metric is calculated in different ways: "revenue" means with VAT to sales, without VAT to finance, and payments received to marketing.
• Some transactions live outside the systems — in chats, notebooks and personal spreadsheets of sales reps — and never reach the analytics.
• Reference lists diverge: one customer is entered three times under different names, and their history is scattered.

Hence the main conclusion: the dashboard is the tip. Underneath it you need a loop where transactions are recorded the moment they occur, reference lists are unified, and metric definitions are written down. Putting that order in place takes up most of the project — and delivers most of the value.

Where does a dashboard get its data?

The data comes straight from the systems where it is born: balances and payments from the bank and from 1C, the accounting and ERP platform most Russian companies run on; deals and the funnel from the CRM; revenue by location from the online cash registers; stock from the inventory system. The exchange runs over APIs — the term for how systems pass data to each other without a person involved. Manual entry is absent from the design by construction: nobody re-keys anything, so nobody makes mistakes or embellishes.

In practice this means a dashboard project is two-thirds an integration project: connect the sources, align the reference lists, agree on definitions. How such a link-up is built technically — events, queues, failure handling — we covered in the article "A zoo of services: how to connect 1C, a CRM, the bank and Telegram into one loop".

How does a dashboard warn you about a cash gap?

Through the payment calendar: the dashboard combines the schedule of upcoming receipts and payments one or two months ahead and highlights the day the balance goes negative. The warning arrives weeks before the problem — there is time to renegotiate a deferral, speed up collections or postpone a non-essential payment. Without a calendar, the gap is discovered on the day itself, when the options have already run out.

The problem is widespread: according to the same Aktion Finansy survey, for 26% of companies a shortfall of funds for current expenses has become a routine part of the operating model. That habit is expensive: gaps end up being covered with short-term loans or the owner's personal money.

Looking at the numbers every day is also a routine, and it is also worth automating. A properly built loop does not wait for you to notice a problem:

• The cash gap forecast is recalculated with every new invoice and arrives as a warning in Telegram.
• A drop in conversion or abnormal churn is recorded the moment it happens, without a month-long retrospective.
• Overdue receivables escalate on their own: a reminder to the counterparty, a task for the sales rep, a notification to the manager.

What to build a dashboard on: a BI platform, custom development, or a Telegram bot?

The choice of tool is secondary: all three options work on clean data, and all three lie on dirty data. A BI platform is strong on flexible analytics, a custom web dashboard on precise fit to your process and to any source, a Telegram digest on the fact that people actually open it every morning. There is plenty to choose from: the BI market in Russia, as estimated by TAdviser, passed 63 billion rubles for 2024, and there are dozens of domestic platforms.

In practice the options complement each other: a morning digest in Telegram on top, a screen with drill-down for the deeper look. A prototype of that screen, incidentally, can be put together in an evening with vibe coding — a cheap way to test the set of metrics before a full project. Our catalogue of ready-made solutions includes a financial calendar for this job — a narrow starting point without a large project.

How to roll out an owner's dashboard: five steps

The working sequence runs from definitions to automation: first fix the metrics and the rules for calculating them, then connect the sources, and only then build the screen. Projects that begin with buying a BI tool usually end with a pretty picture drawn on manually entered data. The first working screen on connected sources appears within weeks.

1. Pick 5–9 metrics and write down the definition of each: what counts as "revenue", whether it is recognized on payment or on shipment, with or without VAT.
2. For every metric, find the source system. Anything that lives in chats and notebooks has to move into the systems — otherwise the metric stays manual and will be the first to lie.
3. Connect the sources and align the reference lists: one customer, one product, one location — one identifier across all systems.
4. Build the first screen and spend a month or six weeks reconciling it against your familiar reports. Trust comes from the matches, and the mismatches expose accounting errors — which is more useful than it sounds.
5. Add alerts and the payment calendar, then switch off the manual reports the dashboard duplicates.

A dashboard is rarely the first automation project — it usually crowns the work of putting processes in order. If that order is still a long way off, start with the roadmap in "Business automation: where to start".

Which mistakes turn a dashboard into a picture?

Four mistakes come up more often than the rest: metrics without fixed definitions, manual entry "temporarily, until we connect it", an overloaded first screen, and no owner responsible for accuracy. Each on its own looks like a small thing; together they destroy trust in the numbers within a couple of months, and the owner goes back to calling the finance manager.

• No definitions — sales and finance keep arguing about "revenue", now while looking at the same screen.
• Manual entry "for now" — the temporary field lives for years, and people forget to fill it in.
• Thirty charts on the first screen — the dashboard turns into a report and stops being opened.
• No owner — a broken data exchange gets noticed a month later, through a mismatch with the bank.

All four are closed off at the design stage, which is why the cheapest time to fix them is before the start, using the list above.

In our fintech case study, end-to-end analytics on top of a single loop gave the owner the main thing: managing on real numbers in the moment instead of a feeling that "things seem fine". If you recognized your own situation in this article — tell us about your task: we will work out where your numbers are born, connect the sources, and build a screen you can trust.

Sources

• Vedomosti: businesses are hitting cash gaps more often — Aktion Finansy survey, May 2026
• Vedomosti: "The economics of financial chaos" — studies by Moe Delo and Luchi on financial record-keeping in small business, May 2026
• Gartner: How to Improve Your Data Quality — the cost of poor data quality, 2021
• Stephen Few, Perceptual Edge: Dashboard Confusion Revisited — the definition of a dashboard, 2007
• TAdviser: key trends in the Russian BI market — market size estimate for 2024
• hh.ru / CNews: one in four employees spends more than six hours a week on non-automated reports, 2025