Real-time client dashboard for a SaaS: cost in Miami (2026)

The verdict in three sentences

A real-time client dashboard built into your SaaS costs 33,000 to 82,000 USD (about 30,000 to 75,000 EUR) of development and 330 to 1,650 USD per month of infrastructure depending on message volume. Embedded analytics tools (Metabase, Power BI Embedded, Luzmo) are faster to plug in but rarely refresh below 30 to 60 seconds and often bill per viewer. If second-level data is a selling point of your offer, custom development is usually the only clean way to deliver it.

The reference architecture and its cost

An IoT platform that pushes live readings to the browser rests on four building blocks. Each has a build cost and a running cost.

Two choices weigh heavily on the bill: upstream aggregation (continuous aggregates per minute and per hour, so millions of points are never sent to the browser) and retention (raw data 30 to 90 days, aggregates for several years). Properly tuned, they cut storage costs by 5 to 10 times.

In Miami, a senior developer at a local agency bills 120 to 180 USD an hour in 2026, so the same 50 to 110 day project can cost 50 to 100% more than with a nearshore team. The ranges above reflect a nearshore or mixed team.

Embedded analytics or custom: the limits to know

Embedded tools are attractive for their time to launch. Their limits show as soon as real time or full white-labelling become requirements.

A mixed approach works well: a custom real-time module for operational screens, and an embedded tool for rarely viewed historical reports.

Mini case study

Carlos, CTO of a 45-person IoT scale-up in Miami, monitors 2,000 industrial sensors emitting one reading per second, for 120 customers. An embedded solution would cost about 1,650 USD per month, plus 16,500 USD of integration, i.e. about 76,000 USD over 3 years, with a one-minute refresh his customers will not accept.

The custom build is quoted at 60,000 USD, plus 880 USD per month of infrastructure, i.e. about 92,000 USD over 3 years. The 16,000 USD gap is covered by a premium live monitoring option at 55 USD per month, taken by 40 customers: 26,400 USD of recurring revenue per year, or about 79,000 USD over 3 years. The dashboard becomes a revenue line rather than a cost centre.

FAQ

WebSocket or Server-Sent Events?

Server-Sent Events are enough when data only flows from server to browser, which covers 80% of dashboards. WebSocket is needed if users also send commands to devices, for about 10 to 20% extra development cost.

How many concurrent connections can one server handle?

A well-configured Node.js or Go server holds 10,000 to 50,000 open connections. Beyond that, a managed service (Ably, Pusher) costs about 55 to 550 USD per month depending on volume.

Which time-series database should we choose?

TimescaleDB fits if your team already knows PostgreSQL and stays under a few billion rows. ClickHouse becomes relevant above 100,000 messages per second, often with 10 times better compression.

Where should customer data be hosted?

For US customers, a US region of AWS, Azure or Google Cloud; for European customers, an EU region to comply with GDPR without transfers. Running two regions adds roughly 15 to 30% to infrastructure costs.

How long until a first screen is in production?

Allow 6 to 8 weeks for a first live dashboard on a limited set of sensors. The full version with alerts, permissions and exports arrives in 10 to 16 weeks.

Let's scope your project. We price your real-time dashboard from your message volume, your customers and your current stack, with an indicative budget of 33,000 to 82,000 USD and a first screen in production in 6 to 8 weeks. Detailed quote within 48 h. WhatsApp +221 77 596 93 33.