Ask what a custom dashboard costs and you will get a number that answers a different question. The quote covers the build: the visible weeks of connecting, transforming, and rendering. But a dashboard is not a deliverable, it is a small piece of infrastructure with a pulse, and infrastructure has an operating cost that outlives every invoice. I build these for clients, so this article prices my own work the way I wish everyone in this market priced theirs: all three lines, including the two that quotes omit.
Line one: the build (the line you expected)
What actually drives a dashboard build's size, in order: the number of data sources, the messiness of each source, and the transformations between raw data and the numbers you want on screen. The charts themselves are nearly free; rendering a clean table of monthly revenue takes an hour once the data underneath is clean, and that "once" is the entire project.
Concretely: "GA4 plus our shop backend, weekly" is a small build. "GA4, three ad platforms, email, and the CRM, with cost data reconciled to revenue per campaign" is a real data project wearing a dashboard costume, because reconciliation is a discipline of its own, as anyone who has read the reconciliation how-to knows. I scope Custom Build engagements per project off exactly this inventory: sources, mess, transformations. Any fixed price you see quoted before that inventory happens is a guess wearing confidence.
Line two: the pipes, forever
Every source connection is a promise made to an external company's API, and external companies break promises on their own schedule. Fields get renamed, auth methods get deprecated, quotas change, exports move. When that happens, your dashboard does not crash honestly. It does something worse: keeps rendering with stale or partial data, looking exactly as authoritative as yesterday.
This line is why "we built it, we are done" is the most expensive sentence in custom tooling. The pipes need monitoring (does the data arrive, is it fresh, do totals still reconcile to sources) and they need an owner who reacts when a check goes red. The monitoring itself is cheap to build during line one. The owner is the actual cost, and it is line three.
Line three: maintenance, the line nobody budgets
Someone must own the dashboard: apply the API fixes, adjust definitions when the business changes what "conversion" means, add the metric the new channel needs, answer "why does this number look weird" with evidence instead of a shrug. Budget a small but real monthly retainer of attention, whether internal or external. The alternative is well documented: an unowned dashboard drifts quietly wrong, the team notices before anyone says it aloud, trust dies, and everyone returns to exporting spreadsheets, at which point the build cost bought a monument. The mechanics of that decay are in why your single source of truth keeps lying.
A useful planning heuristic: over a dashboard's life, budget for the operating side (lines two and three) to rival the build cost. If that ratio kills the business case, the business case was never alive.
The benefit line: what it must beat
A custom dashboard's competition is not "no dashboard." It is Looker Studio, which is free and covers the display job for most businesses. Custom earns its three cost lines only when the job is something the free canvas cannot do: reconciled cross-source metrics with real transformation logic, workflows attached to the numbers (alerts, approvals, client portals), row-level access by franchise or client, or embedding into your product. "Our Looker Studio got messy" is not on that list; mess ports beautifully into custom, at custom prices. Clean the definitions first, then re-ask whether anything is still missing.
The decision, stated plainly
Run the same honest sequence as every build vs buy call: name the job the free tool provably cannot do, inventory sources and transformations, price all three lines, and compare against the workflow cost the dashboard removes. If the math clears, custom tooling is genuinely great; owning the exact instrument your operation runs on is the good version of the custom-build temptation. If it clears only by ignoring lines two and three, you have priced a third of a dashboard, and the missing two-thirds will collect themselves. When you want the inventory done by someone who prices this for a living, that is a Clarity Call; the wider worldview is in the build vs buy guide.