Decision framework
Weight the indices against a life, not against a ranking.
Word list
- Decision framework
- The rules for reading the order on the homepage. It is a weighting for one applicant, not a quality ladder of the six schools.
- Indices
- Five scores from 1 to 10: livability, job security, future proof, selectivity, and ROI. Each score compares these six programs only.
- Visa prior
- A planning range for OPT and for whether employers in that market sponsor. It is not an admission rate and not a promise.
- Future proof
- How directly the curriculum treats human-AI work, systems, and the mix of research, craft, and technical skill after 2026.
- Job security
- How usable the path is into product design, UX research, or product management: brand, network, internship, and where alumni land.
- All-in sketch
- Tuition plus a living estimate for the length of the degree. Not an invoice. Fees, flights, and aid sit outside it unless a line says so.
- Holds
- Where the program is strong for this decision.
- Costs you
- What you give up, or the risk you take, if you pick it.
- Full freight
- You pay the published tuition. Do not assume a scholarship or an assistantship will cut it.
The working rank on the homepage is for a product-design move, from India, with tuition that has to be survived. If that is not you, change the weights. The scores will not be offended.
How to use a 1–10
Livability
City-level quality of life for an international graduate student: rent, safety, transit, climate, and how livable the day is outside class.
Job security
Strength of a path into product design, UX research, or product management: brand, network, internship conversion, and where alumni actually land.
Future Proof
How directly the curriculum treats human–AI interaction, systems, and the mix of research, craft, and technical fluency you will need after 2026.
Selectivity
Relative admissions difficulty across these six programs: published rates where they exist, cohort size, and how hard the typical admitted profile is.
ROI
Total cost of attendance set against placement strength, time to a salary, and whether aid is real or theoretical.
Visa chances
A percentage range, not a score. It is a prior about OPT and sponsorship, not about getting in. Read the limit before you treat 65–80% as a plan.
Full criteria and the things these numbers cannot see are on the methodology page.
A weighting you can steal
| If this is the real constraint | Turn up | Turn down |
|---|---|---|
| Craft gap in the portfolio | Future proof, and the curriculum page | Selectivity as a quality signal |
| Need a sponsor-capable job | Job security and the visa range | Livability, until you have an offer |
| Debt you cannot carry | ROI and the cost table | Brand, including CMU at rank 1 |
| You can fund one intense year | Speed: CMU or UW | Two-year flexibility you will not use |
| You want research credentials | Michigan and Georgia Tech methods | One-year studio speed |
Selectivity is a planning input: how many applications, how finished the work must be. A higher score is a harder door. It is not a better education.
Five honest forks
01
The portfolio is the gap
Weight craft over brand. Start with UW versus CMU. Washington will make you design every week. Carnegie Mellon will make you ship a seven-month client project and will not teach you form if you arrive without it. Georgia Tech’s industrial design track is the third door, and only if you apply to that track.
Compare UW and CMU on curriculum02
The target job is product management
Put Berkeley first. Michigan’s new PM pathway is the serious alternative if you want the skill without the Bay Area burn. CMU can reach PM through strategy and the capstone, and it is still a design-and-research degree. Do not pick a studio program and hope the title changes.
Compare the PM three03
You want research, not another mockup class
Michigan’s UX pathway is built for methods. Georgia Tech Psychology is the methods track inside a more technical school. CMU will give you research inside a product team, on a clock. If your portfolio has never survived contact with a user who disagreed, fix that before you optimize prestige.
Compare Michigan and Georgia Tech04
Cash is the constraint
Price UT Austin first: $9,893 per nine-hour nonresident semester on the Fall 2026–Spring 2027 information-school table. Then ask whether a one-year program returns you to a salary soon enough to beat two cheap years. UW at about $64k tuition can win that math. Georgia Tech’s professional rate is not the bargain people remember from in-state rumors.
Compare the cost three05
The visa is the constraint
Bias toward STEM-designated degrees (confirm the CIP on the I-20, especially at Michigan) and toward employers that have sponsored before. Berkeley’s range is the highest prior in this set because PM and engineering-adjacent roles sponsor more often than design studios. A beautiful portfolio aimed only at studios is a weak visa plan at every school, including CMU.
Compare visa priorsA 90-second comparison
- Pick two programs, or three if one is a spoiler on cost.
- Read Cost. If the invoice is impossible, stop. Taste cannot refinance it.
- Read Outcomes, including who the employment number actually describes.
- Read Indices last, as a check on the story you just told yourself.
- Open essay notes only for the finalists. Writing six statements is how January disappears.