Every Monday, in thousands of small companies, someone spends 2 to 4 hours exporting data from the ERP, merging it in Excel and sending a report that is already out of date on arrival. A custom dashboard connects those sources once and shows your KPIs in real time. It costs between 1,500 and 6,000 EUR and pays for itself sooner than you would expect, in recovered hours and in decisions no longer made blind.
Why are you still closing the month with Excel exports?
The pattern repeats across almost every company with 5 to 50 employees: the data exists, but it lives scattered between the ERP, the CRM, online banking and half a dozen spreadsheets. To know how the business is doing, someone has to export, copy, paste and reconcile. The result is a report that consumes 8 to 16 hours a month and describes last week, not this one. With that lag, decisions arrive late by definition.
A real example from our work: a consultancy of around 20 people closed its month in 5 working days because invoicing, billable hours and expenses lived in three separate systems. After connecting those sources to a dashboard with automatic reconciliation, the close dropped to 2 days. They did not replace their ERP or their expense tool, they simply stopped moving data by hand. The build cost around 4,000 EUR and paid for itself in under a year on administrative hours alone.
What is a custom dashboard and how is it different from Excel?
A custom dashboard is a web application that connects directly to your systems, calculates KPIs using your business rules and keeps them updated without anyone lifting a finger. Nobody exports anything: the data refreshes on its own, hourly or every few minutes depending on the source. The whole team looks at the same number, calculated the same way, which ends the recurring argument about which version of the spreadsheet is the right one.
The difference with Excel is structural, not cosmetic. A spreadsheet depends on someone feeding it and breaks when that person goes on holiday. The difference with a generic Power BI setup lies in the rules: a properly calculated margin in distribution includes freight, rebates and returns, and that requires modelling your business, not dragging columns. If your entire operation still runs on spreadsheets, the first step is often to move from Excel to custom software and build the dashboard on that foundation.
Which data sources should you connect first?
The four sources with the best return are almost always the same. The ERP provides sales, purchasing and stock. The CRM provides pipeline and sales conversion rates. The bank, via PSD2 aggregation or standard statement files, provides your real cash position and forecast. And if you sell online, the ecommerce platform provides orders and margin per product. With just two of these four connected, most of the manual reporting work in a typical SME disappears.
Technically, each source connects through whatever channel it offers: an official API, a database replica or, at worst, automated file ingestion. Everything consolidates into a central database, usually PostgreSQL, acting as the single source of truth. The choice of engine matters less than people think at this scale; in our guide to SQL versus NoSQL for business we explain why a well designed relational database covers 90 percent of SME use cases.
How much does a dashboard cost and what stack does it use?
From our experience with Spanish SMEs, a dashboard with 2 or 3 data sources and 8 to 12 KPIs sits between 1,500 and 3,000 EUR, delivered in 3 to 4 weeks. With more sources, automatic reconciliation, email alerts or role-based access, the range moves to between 3,000 and 6,000 EUR and 4 to 6 weeks. In both cases we work on a fixed price agreed before starting: the quote is prepared by the same senior engineer who will build the system.
The typical stack is deliberately boring: PostgreSQL as the store, scheduled connectors for each source and a lightweight web frontend that works on the manager's phone as well as on the office screen. Infrastructure runs at 20 to 50 EUR per month on a European server. The code and the data belong to the client from day one, with no per-user licences and no lock-in: if you want to continue with another provider tomorrow, you can.
Which mistakes leave a dashboard gathering dust?
The most expensive mistake is not technical: it is building a dashboard nobody looks at. It usually happens when the design is copied from a template instead of starting from the decisions someone actually makes each week. The practical rule we apply: every indicator needs a named owner and an agreed action if it goes out of range. A panel with 30 metrics and no owners is decoration; one with 10 watched metrics changes behaviour.
The second mistake is vanity metrics: web visits, followers, downloads. They go up, look good in a meeting and support no decision. For an industrial or distribution SME, margin per order, stock rotation, average collection period and 30-day cash deviation are worth far more. And not everything needs real time: cash can refresh hourly, but forcing real time where a daily refresh is enough only makes the project more expensive for no benefit.
How do you start without risking the budget?
The lowest-risk path is to start with a 190 EUR technical audit: within 72 hours you get an initial diagnosis of which data sources you have, what state they are in and what can be connected, and within 5 working days a full report with a fixed quote. If the project goes ahead, the 190 EUR is deducted from the total, so the real cost of deciding with proper information is zero.
From there, we recommend starting with a pilot covering one or two sources and the 5 KPIs that cause the most argument in your management meetings. Within 3 or 4 weeks the team is working with live data and you can see whether the tool changes the Monday routine. Adding more sources or alerts later is incremental rather than a rebuild, because the central database is already in place and documented.
Frequently asked questions
How much does a custom KPI dashboard cost for an SME?
Between 1,500 and 6,000 EUR depending on sources and complexity. A panel with 2 or 3 sources and 8 to 12 KPIs runs 1,500 to 3,000 EUR; with automatic reconciliation, alerts and role-based access it moves to 3,000 to 6,000 EUR. Always at a fixed price agreed before work starts.
How long does it take to get the dashboard live?
Between 3 and 6 weeks from budget approval. A pilot with one or two sources can be in production in 3 weeks; projects with 4 or more sources and reconciliation logic take up to 6. The initial diagnosis arrives within 72 hours through the 190 EUR audit.
Which systems can be connected?
Practically anything with an API, an accessible database or an automatable export: ERPs such as Sage, A3 or Odoo, CRMs such as HubSpot, banks via PSD2 aggregators or statement files, and platforms such as Shopify, PrestaShop or WooCommerce. In companies of 5 to 50 employees, 2 to 4 sources cover 80 percent of the need.
Do I have to replace my ERP or stop using Excel?
No. The dashboard reads from your current systems without touching them; the ERP remains the system of record. Excel can stay for ad hoc analysis. What disappears is the 8 to 16 hours a month spent exporting and reconciling data by hand for recurring reports.
What maintenance does it need after delivery?
Little, but not zero: APIs change and connectors need watching. Infrastructure costs 20 to 50 EUR per month and we offer a technical retainer from 500 EUR per month for support and improvements. No lock-in: the code is yours and any team can maintain it.
