Matthew Cassella
About Work Craft AI Leadership Career Get in touch
Home About Work Craft AI Leadership Career Get in touch
All work / MNTN Select / Budget & Goal
Case study · MNTN Select

Budget & Goal.

The campaign step where advertisers schedule their budgets and choose the performance goal their plan will optimize for.

Role
Design lead and designer
Product
MNTN Select
Step
2 of 4 in campaign setup
Status
Working prototype
Where I led
  • Defining what this step is for and what it leaves to Audience and Media Plan
  • Getting the goal options and edge cases agreed before build
What I designed
  • The budgets table, goal picker, drawer, alerts and every empty or blocked state
  • A working HTML prototype in light and dark
  • All of the hands-on design on MNTN Select, this feature included
The problem

Two decisions, one screen.

Budget and goal shape everything that follows: which deals the plan recommends and how money is split. But advertisers often arrive unsure what a goal means, and budgets in the real world run across months, pause, and overlap.

The step had to make both decisions feel simple, without hiding the detail that sophisticated buyers need.

What's in the design

Every state, designed.

The prototype covers the happy path and the awkward ones, so engineering gets one spec instead of a list of follow-up questions.

Budgets table

Schedule, type, budget and live spend per row, with sort, filters and paging.

Goal picker

Four goals with value entry, validation and a recommended range.

Side drawer

Create or edit a budget without leaving the page.

Alerts

A critical alert when a start date falls in the last week of a month.

Empty and blocked

No budgets yet, and no targetable audience, each with a clear next step.

Light and dark

Both themes, built on the same tokens as the rest of Select.

Try it

Experience the working prototype.

Use the settings icon in the header to switch to the empty and blocked states, add a budget, then pick a goal. Try a start date in the last week of a month.

Design decisions

How it works, and why.

Budgets are a schedule, not a number. Each budget has a start, an end, and a type, so advertisers plan across months and see what is live, upcoming or paused in one table.
Goal comes after budget. The goal control stays locked until a budget exists, with a plain message saying why, so nobody sets a target for money that is not there yet.
Four goals, each explained. Cost per acquisition, return on ad spend, cost per completed view and cost per visit, each with a one-line outcome and a longer description, plus a recommended starting range for ROAS.
Edge cases are first-class. No targetable audience, no budgets yet, a budget starting in the last week of the month, and an items-to-check review are all designed states, not afterthoughts.
Spend is visible while you decide. The table shows current campaign spend next to each budget, so changes are made with the latest data in view.
How I worked

Prototype first, then review.

I built this as a working prototype on the Select design system rather than static frames, so product and engineering could click through the real behavior. The same review format as the Flight Plan empty state carried it to a decision.

See the AI workflow behind it, or the FlightPlan case study.

Want to talk about designing complex setup flows? Get in touch.
Get in touch