TL;DR: Program prestige matters less to your actual PhD experience and outcomes than advisor fit, funding security, and lab/department culture — three things that are harder to research but far more predictive of whether you finish, and finish well. A structured comparison across a small set of concrete factors beats going with a gut feeling about "ranking."
Why ranking alone is a weak decision tool
Program rankings are built from aggregate departmental metrics — average faculty citation counts, overall placement rates, research funding totals. None of that tells you what your specific day-to-day experience will be with a specific advisor, in a specific lab, for the 5+ years you'll actually be there. Two students in the same "highly ranked" department can have wildly different experiences depending entirely on advisor and lab fit.
The factors worth weighing, roughly in order of predictive value
1. Advisor fit. This is consistently cited by graduate students as the single biggest factor in whether a PhD experience is positive or miserable — more than program prestige. Talk to current and former students of the specific advisor, not just the department. Ask directly about mentorship style, availability, and what happened to students who left the lab early.
2. Funding security and structure. A 5-year guaranteed funding package is a materially different risk profile than a 2-year guarantee with "expected but not guaranteed" funding after that. Ask explicitly what happens if your advisor's grant funding runs out.
3. Lab and department culture. Ask current students (ideally without faculty present) about work-life balance expectations, how conflicts get resolved, and whether people generally seem to like being there.
4. Research fit and resources. Does this lab or department actually have the equipment, data access, or computational resources your specific research direction needs — not just "does this field exist here."
5. Placement outcomes for people with your specific goals. A department's overall placement rate matters less than where students who wanted what you want (academia, industry, a specific subfield) actually ended up.
6. Cost of living and stipend, adjusted together. Compare the actual purchasing power of each stipend in its city, not just the headline number (see Grad Housing Math for a way to think about this, and the comparison tool for stipends and local costs side by side).
7. Program prestige and ranking. Real, but least predictive of your day-to-day experience — worth weighing last, not first, despite often being the first thing applicants fixate on.
A practical exercise
Score each offer 1–5 on each of the seven factors above, based on what you actually learned from current students and your own conversations — not assumptions. Weight advisor fit and funding security more heavily than the others (for example, double their score) before totaling. This won't produce a mechanically "correct" answer, but it forces a comparison based on the factors that actually predict outcomes, rather than the factor (prestige) that's easiest to compare on paper but least predictive in practice.
The question to ask every current student you talk to
"If you were choosing again today, knowing what you know now, would you pick this advisor and this program?" — phrased this way rather than "are you happy," it tends to get a more honest answer, including hesitation that a more generic question wouldn't surface.