A dashboard redesign costs $3,000 to $8,000 for a single surface and takes three to five weeks of design. A product with several role-specific dashboards runs $10,000 to $30,000 over eight to twelve weeks. Those are my working numbers for 2026, and they surprise people in both directions: higher than a restyle should cost, lower than rebuilding the product. The reason is that you are not paying for charts. You are paying to decide what the screen is for, and that part does not get cheaper with fewer widgets.
This is how that number is built, what pushes a quote past the top of the range, what brings it down, and how to tell whether you need a redesign at all.
The short answer
| Scope | Range (USD) | Timeline |
|---|---|---|
| Audit: what to keep, cut and fix | $1,500 to $4,000 | 1 to 2 weeks |
| One dashboard surface, one role | $3,000 to $8,000 | 3 to 5 weeks |
| One surface serving two or three roles | $6,000 to $12,000 | 4 to 7 weeks |
| Several dashboards across a product | $10,000 to $30,000 | 8 to 12 weeks |
| Front-end build on top | Add 40 to 60 percent | Adds 3 to 6 weeks |
These are independent senior rates, hiring directly. An agency quoting the same scope will usually land two to three times higher. For how dashboards sit inside a whole product budget, see what B2B SaaS design costs.
Where the hours go on one surface
A single-surface redesign at around $6,000 breaks down roughly like this. The proportions matter more than the total, because they show where a cheaper quote has been cut.
- Discovery, about a fifth. Watching three or four people use the screen and naming the decision each of them makes with it. This is the step that tells you what to cut.
- Structure, about a fifth. The audit of every widget against a decision, then the hierarchy and the wireframe. Most of the value is settled here, before anything looks finished.
- Screens and states, about two fifths. The surface itself, the table under it, and the states nobody puts in a brief: empty, loading, stale, partly loaded, failed and overloaded.
- Prototype and handoff, about a fifth. A clickable version for review, specs the engineers can build from, and answers to their questions while they do.
A quote that skips discovery is a restyle with a dashboard’s name on it. It will look better and nobody’s morning will change, which is the failure I walk through in this dashboard redesign, step by step.
What pushes the cost up
More than one role on the same screen
Every role that looks at a dashboard makes different decisions with it, at a different frequency. Operations staff checking forty times a week and a manager checking on Mondays cannot share one layout without one of them losing. Two or three roles on one surface means different default views built from the same components, which is why it prices at $6,000 to $12,000 rather than double.
States, not charts
Clients count widgets. I count states. A chart that can be empty, loading, stale since this morning, missing one of its three sources, or showing forty exceptions where the design assumed three is six designs, not one. This is the largest single gap between two quotes for the same brief, and it is almost never itemised.
How live the data is
A dashboard refreshed overnight is a report. One that updates every minute has to show when each number was last true, what happens mid-refresh, and how to flag that a source has gone quiet. Real-time data adds about a week to most projects, so it is worth asking whether the decision behind the screen actually needs it.
Filters, date ranges and drill-down
Every control multiplies the states above. A date picker, three filters and a drill-down into each row add up to a dozen paths through the screen, and any of them can end somewhere empty. Saved views, comparison periods and exports each add to it again.
Customer-facing rather than internal
An internal dashboard needs to be clear. A dashboard your customers log into also has to survive their data, their branding if you white-label, and their first week, when every chart is empty. Expect the customer-facing version of the same surface to land towards the top of its range.
What brings it down
- Access to users. Two hours watching four people work shortens every later step. Teams who cannot arrange it pay for the guessing instead.
- A working component library. A real one takes 20 to 30 percent off. A folder of inconsistent Figma frames described as a design system does not.
- Usage data before we start. How often the screen is opened, how often people route past it, and how many exports it causes. That baseline turns arguments about what to cut into a short conversation.
- One person who can say yes. Every extra round of “let’s see another option” is real hours, and it is the one variable entirely under your control.
- Starting with an audit. It answers a $2,000 question before anyone commits $8,000: whether the problem is the design, the data, or a process nobody has written down.
How the three to five weeks run
The first week goes on discovery and the decision map, the second on the audit and the wireframe. The third and fourth cover the surface, the table and the states, with one review round each, and the fifth the prototype and handoff. What stretches it is almost never the design work. It is waiting for access to the people who use the screen, and stakeholder rounds that reopen decisions already made.
If build is included, the front end runs alongside from the third week, and I stay on it until the states nobody demoed are working in production. That is where a dashboard design usually gets quietly compromised.
When you do not need a redesign
Some dashboards need a week of tuning rather than five weeks of redesign. If people open the screen and act on it, the job is right and the details are wrong: labels, order, a missing filter. If people route past it to the table underneath, the screen has the wrong job and needs deciding again, which is a redesign. And if the numbers are late or wrong, no design fixes that. You have an engineering problem, and paying for screens first is the most expensive way to find out.
The eight principles I apply to the ones that do need redesigning are in B2B dashboard design, and the ordering dashboards in the goodport case study show what the result looks like in a real product.
Frequently asked questions
How much does a dashboard redesign cost in 2026?
For one surface, $3,000 to $8,000 of design over three to five weeks, working with an independent senior designer. Two or three roles on the same screen take it to $6,000 to $12,000, and several dashboards across a product run $10,000 to $30,000. Agencies usually quote two to three times these figures for the same scope.
Is redesigning a dashboard cheaper than building a new one?
Less than people hope. Discovery, the states and the table cost about the same either way, because they are decisions rather than drawing. The saving comes from a data model that already works. If it does not, a redesign turns into a rebuild in week two, and it is better to know that before the quote.
Do you price per chart or per screen?
Neither. A screen with fourteen charts can be a smaller job than one with four, if ten of the fourteen are being cut. I price by the decisions the surface serves and the roles that make them, which is also what the finished dashboard should be judged on.
What should I send to get an accurate quote?
A screenshot of the current dashboard, who uses it and how often, what they decide with it, and whether the build is included. Usage numbers help if you have them. With that I can usually give you a range in one reply and tell you what would move it.
Should we redesign the dashboard or the whole product first?
Usually the dashboard. It is contained, it is often where the complaints are, and its results are easy to measure: time to the first action, exports to spreadsheets, and whether people open it at all. A good result there makes the case for the rest of the product with real numbers instead of opinions.
Getting a number for your dashboard
Send me a screenshot and tell me who uses it and for what. I will tell you whether it needs tuning, a redesign or an engineer, and roughly what each would cost. This is the work I do as dashboard design, and there is more of it in my case studies.
Planning a website or product design project?