Service · Software development

Custom KPI dashboard.

A dashboard that aggregates live data from your own systems, embedded in your application or as a white-label environment for customers. Not an off-the-shelf BI tool, but custom software for where standard connectors, performance requirements or branding reach their limits.

A custom dashboard is not a replacement for Power BI.

Power BI, Tableau and Looker are excellent standard tools for reporting and analysis. We do not replace them; we often work alongside them. We build a custom KPI dashboard when a standard tool reaches its limits: because there is no ready-made connector for your source system, because the dashboard must sit inside your own application, because it must run white-label for your customers, or because the calculations and visualisations are so specific that a drag-and-drop tool cannot cover them.

We build dashboards that bring together real-time data from multiple sources, branded to your look and feel, with exactly the calculations and visualisations your situation calls for. For marketing teams who want Google Ads, GA4, Meta, HubSpot and Mailchimp on one screen. For operations teams who need live MES data from the shop floor. For CFOs who want ERP data and BI in a single view. And for SaaS companies who want to give their customers a dashboard with their own data, in the right brand.

The question is almost never "can we do this in Power BI too?" The real questions are who will use the dashboard, how live the data needs to be, how deep the integration with other systems goes, and whether the dashboard is part of your brand or Microsoft's. We weigh our recommendations on those questions, and our advice follows from them. Sometimes it means a better Power BI setup; often it means a custom dashboard alongside what is already running.

We work across a broad range of sectors: B2B SaaS, e-commerce, manufacturing, financial services, healthcare and media companies. The common thread is that the value lies in how well the source data is made accessible, not in how many charts appear on screen. A good dashboard helps someone make a decision or spot an anomaly within ten seconds, and then offers a deeper layer for anyone who wants to dig further.

What sets a custom dashboard apart.

Live data
Real-time via WebSocket or SSE, not scheduled exports from yesterday.
Multi-source
Aggregates sources that do not share a standard connector.
White-label
Branded per client, embeddable in your application or portal.
Custom logic
Calculations and visualisations that do not fit standard tools.

Six types of dashboards we build.

In practice they often overlap. A marketing dashboard at a SaaS company is also a customer-facing dashboard. We work out together which combination fits, based on a good conversation about who will use it and for what.

Type 01

Marketing dashboards

Google Ads, GA4, Meta Ads, LinkedIn Ads, HubSpot, Mailchimp and your own CRM on one screen, with campaign attribution across channels. For marketing teams who don't want to open five tabs to see whether a campaign is working. We build the data pipeline so that new sources can be added later without touching the visualisation layer, and connect via our API integrations to any source system that has an API.

GA4Google AdsMeta AdsHubSpotMailchimpAttribution
Type 02

Sales and CRM dashboards

HubSpot or Pipedrive alongside your ERP, inventory and order data. A sales manager sees not only pipeline value but also which deals are at risk because stock is tight or margin is slipping below a threshold. Designed for organisations where sales and operations need to see the same reality in order to make decisions. We combine CRM, financial and operational data in a single view, so no sales rep has to reconcile a spreadsheet by hand.

HubSpotPipedriveSalesforceERP integrationPipeline forecastMargin monitoring
Type 03

Operations and production dashboards

Real-time dashboards that bring together data from MES, SCADA, inventory systems and order books. Built for production floor managers and logistics planners who need to know now which line is falling behind, which machine has stopped and where the bottleneck is. Here, sub-second latency is often the difference between intervening and playing catch-up, a use case where standard BI with scheduled refreshes falls structurally short.

MES dataSCADAReal-time alertsOEE trackingLive inventoryWebSocket
Type 04

Financial and executive dashboards

A CEO or CFO view with the top KPIs for your organisation: revenue, margin, cash flow, headcount, customer retention, project status. Often combines data from Exact, AFAS, Twinfield or another accounting system with BI and operational sources. An executive dashboard should fit on one screen, be readable in ten seconds, and show the same figures you quote in a management team meeting. We build dashboards that pass that test.

ExactAFASTwinfieldTop KPIsForecast modelDrill-down
Type 05

Customer-facing and embedded dashboards

A dashboard where your customers or partners see their own data: usage, performance, progress, ROI. Suited to SaaS platforms that want to deliver more value from their product, agencies that want to report transparently to clients, and suppliers who want to give partners a real-time view. Fully branded, with SSO and role-based access so each customer sees only their own data. Embedded via iframe, web component or SDK in your existing application or web environment.

White-labelSSOMulti-tenantEmbedded SDKRole-based accessIframe
Type 06

ERP dashboards with KPIs

For organisations that use their ERP as the operational core but lack a manageable KPI layer on top of it. We build dashboards that draw directly on the ERP database or ERP API and show the KPIs that the ERP's standard reports don't deliver, or deliver only with difficulty: margin per customer, lead time per order flow, stock turnover per product group, revenue per product line. Often combined with a custom ERP or added as a KPI layer on top of SAP, Exact, AFAS or Microsoft Dynamics.

SAPMicrosoft DynamicsExact GlobeAFASKPI layerOperational reporting

What you get at the end.

A production-ready dashboard, plus everything around it so you can manage and extend it yourselves.

The dashboard

Production and staging environments, hosted in your cloud or with us.

Data pipeline

ETL or streaming pipeline into the data warehouse, with monitoring.

Integrations

Integrations with your source systems, with clear documentation for each integration.

Codebase and documentation

Full source code, build instructions and an architecture overview.

Managed service (optional)

Monitoring, security patches, new metrics and ongoing development.

When a custom dashboard is the right choice.

Six patterns we often encounter before a client opts for custom development instead of a standard BI tool.

Multiple sources without a connector

Standard connectors don't cover every source

Power BI or Tableau has no native connector for your order portal, your legacy ERP or that one Python microservice. Workarounds with scheduled exports and CSV files are fragile and often break down. A custom dashboard can connect directly to any API or database.

Real-time requirements

You need data from right now, not from last night

Operational decision-making can't wait for a scheduled refresh. On production floors, in logistics planning and in customer support, everything depends on live data from the source. WebSocket or SSE keeps the screen continuously in sync with reality.

Branded experience

Your customers shouldn't see Microsoft

An embedded Power BI report in your client portal shows Microsoft branding, Microsoft icons and Microsoft UX patterns. A white-label dashboard reflects your brand, with your colours, typography and navigation. For a SaaS or agency product, that difference is not a minor detail.

Embedded in your own application

The dashboard belongs in your product

Customers shouldn't have to leave your application for a separate BI environment. KPIs belong where the work happens. Embedded via a web component or SDK, with the same login and the same permissions as the rest of your application.

Public-facing KPIs

Anyone may view it

A transparency page for customers, a performance overview for regulators, a status page for your market. A public-facing dashboard has very different requirements for caching, performance and uptime than an internal tool, and its content must be explainable to every visitor.

Specific calculations

The formula doesn't fit in DAX

Your KPI calculation relies on a domain model of its own, an industry-specific formula or a step-by-step procedure that is hard to squeeze into a BI tool's formula language. In those cases, a custom calculation engine in TypeScript or Python is faster, easier to test and easier to maintain.

How an engagement works.

01Introduction→ 02KPIs and data mapping→ 03Build in sprints→ 04Rollout and management
No-obligation conversation

Introduction

Who uses the dashboard, which decisions it needs to support, and which sources sit underneath. We listen to your situation and provide direction where it helps.

Workshop with your team

KPIs and data model

Together we map out the KPI definitions, source systems and visualisation requirements. Output: a workable data model and a prioritised list for the build.

Fortnightly sprints

Build and validation

A working build every sprint. You test alongside the actual users. Feedback goes straight into the next sprint, so the dashboard grows with what your team actually uses.

Ongoing

Rollout and management

A phased rollout to the first group of users, handover to your management team and further development in subsequent sprints. Monitoring and security patches continue throughout.

Tech stack we work with.

Per project we choose what suits your existing landscape, your real-time requirements and the level of visual control you need. The frontend relies on React or Next.js with D3, Recharts or Visx for visualisation. The backend runs on Node.js or Python, depending on your situation. For the data warehouse we work with Snowflake, BigQuery, ClickHouse or PostgreSQL, depending on data volume and query patterns. For real-time we use WebSocket or Server-Sent Events. Authentication via SSO with Auth0, Azure AD or your own IdP, with role-based access per user or organisation. We keep performance in check through caching layers, materialised views and incremental loads. In short: technical provisions that don't exist in a drag-and-drop BI tool but are routine in custom development.

Frontend and visualisation
ReactNext.jsD3.jsRechartsVisxWebSocket / SSE
Backend and data
Node.jsPythonPostgreSQLSnowflakeBigQueryClickHouse
Auth, infrastructure and integration
Auth0Azure ADRBACGCP / AWSAirbyte / Fivetrandbt

Frequently asked questions.

Do you replace Power BI, Tableau or Looker?
No. Power BI, Tableau and Looker are excellent standard BI tools and we don't replace them. We build custom solutions where standard tools reach their limits: missing connectors, real-time requirements, embedded or white-label dashboards, public-facing KPI pages, or calculations that don't fit DAX or LookML. Often our dashboards run alongside an existing Power BI or Tableau environment rather than instead of it.
What is the difference between real-time and scheduled dashboards?
A scheduled dashboard is refreshed at fixed times, usually overnight via an ETL job. A real-time dashboard receives updates the moment the source data changes, via WebSocket or Server-Sent Events. Real-time is essential for operational dashboards (production floor, logistics, customer support) where minutes matter. For reporting dashboards where daily figures are the unit, scheduled is usually sufficient and cheaper.
Can a dashboard be white-labelled or embedded?
Yes, and that is one of the most common reasons to choose custom development over a standard BI tool. White-labelling means the dashboard runs entirely in your own brand identity, without Microsoft, Tableau or Looker iconography. Embedding means we place the dashboard in your existing application or customer portal via an iframe, web component or SDK, with the same login and permissions as the rest of your product.
How do you handle access and role-based access?
Through SSO with your existing IdP (Azure AD, Auth0, Okta or a custom solution), with role-based access so every user only sees what they are allowed to see. With multi-tenant dashboards (a dashboard that shows data to multiple customers or organisations), data separation is enforced at database level, not just at UI level, so customers can never see each other's figures. Audit logs are a standard part of delivery.
Can the dashboard connect to our ERP or CRM?
Yes, and in almost every project that is the substantive work. We have experience with SAP, Exact, AFAS, Microsoft Dynamics, NetSuite, HubSpot, Salesforce, Pipedrive and custom enterprise software. When a standard API is missing, we rely on database replication or smart API integrations. We preferably recommend a data warehouse as an intermediate layer, which relieves the source systems and makes the dashboard's performance independent of them.
Can we build custom visualisations?
Yes. That is one of the reasons to choose custom development. Using D3.js, Recharts or Visx, we build visualisations that don't exist in a standard BI tool: tailored heatmaps of a production floor, custom Sankey diagrams, geo-overlays with live IoT data, or a specific industry visualisation your sector will recognise. We build them as reusable components so they can be deployed across multiple dashboards.
What about hosting and security?
Hosting can run in your own cloud (Azure, AWS, GCP, or a private cloud) or with us. We work with encryption in transit and at rest, audit logs as standard, a DPIA for every major project, and penetration tests in the final sprint. Where compliance requirements apply (ISO 27001, NIS2, GDPR) we build the architecture around them from day one, which is far cheaper than bolting compliance on afterwards.
What determines the cost of a custom dashboard?
The biggest cost drivers are the number and complexity of the source integrations, the choice between real-time and scheduled data, the number of roles and the multi-tenant nature, and the degree of custom visualisation. A simple marketing dashboard with three sources and a handful of charts is a project of a few sprints. A multi-tenant white-label dashboard with real-time data and its own calculation engine is a larger project spanning several sprints. After the introductory call, we give direction, with an open book on what scope choices mean.
Can we start small and build from there?
Almost always that is the smartest route. We start with a first version that covers the most valuable KPI set and source integration, and lay down an architecture that can grow with you. Extra sources, roles, dashboards and clients are added in later sprints without having to rebuild. This way the dashboard delivers value straight away while you learn what a second version needs.

Talk to us about your KPI dashboard.

A free, no-obligation half-hour introductory call. We listen to who will use the dashboard, which data sources sit beneath it and which decisions it needs to support. Would you rather first explore the wider context of custom software development? That is possible too.

Edit content