Nike · Primly Community

Nike frontend engineer interview: React, performance, and what they actually care about

sre_sol · 4 replies

Just went through the Nike.com frontend engineer interview for a mid-level position (think 3-6 YOE range). This was for the Digital Products org supporting Nike.com, not the apps side. Here's what I found.

The process was: OA, technical phone screen, and then a virtual onsite with three rounds. Total timeline was about four weeks from first contact.

Online Assessment. JavaScript-focused. Two questions: one was a DOM manipulation problem (build a small interactive component from scratch, no frameworks), and one was an algorithm question that was basically array work dressed up with a UI-adjacent scenario. Medium level. The DOM problem was more important -- they're explicitly testing if you can write vanilla JS when needed.

Technical phone screen. An hour with a senior frontend engineer. We talked through my most recent project in depth, then they gave me a live coding problem: build a search autocomplete component with debouncing. I had to handle async fetching, show a loading state, handle errors gracefully, and clear results on empty input. Pretty realistic frontend problem. They watched how I thought out loud more than whether I got it perfect.

Onsite round 1: React + state management. Given a spec, build a small multi-step form. I used hooks, managed state locally, discussed why I wouldn't reach for Redux for this scope. They asked me about controlled vs uncontrolled components, why I made specific hook choices, and what I'd test if this were production. React 18 knowledge is expected; if you haven't looked at concurrent features and Suspense, skim at minimum.

Onsite round 2: Web performance. This round surprised people I talked to who went through the same loop. They gave me a hypothetical: Nike.com product listing page, LCP is 4.2s, what do you do. I walked through image optimization (they use a lot of large hero images and product photography), lazy loading, critical CSS, bundle analysis, CDN edge caching, and component-level code splitting. They pushed back on everything, which was good -- it felt like a real technical conversation, not a quiz.

Onsite round 3: Behavioral. Three STAR questions. One was team conflict, one was shipping under pressure, one was about advocating for a technical decision that stakeholders pushed back on.

Comp for mid-level in Beaverton was around 145-155k base. Heard the Beaverton cost of living adjustment is real but coming from Bay Area that still stings a little.

If I were prepping from scratch: solid React hooks, vanilla JS DOM, one solid web performance study session (Core Web Vitals, LCP/CLS/FID), and behavioral stories. Don't neglect the performance round -- that's the differentiator.

4 replies

backend_bekah

the performance round focusing on nike.com specifically makes sense -- they have massive product images and their page weight has historically been rough. knowing LCP and image optimization for that context isn't generic frontend prep, it's actually domain-specific.

market_realist

145-155 base in Beaverton is the recurring theme across every nike thread. they seem pretty locked in there for mid-level. anyone know if there's room to move on RSUs or is the whole comp package compressed?

frontend_fran

RSUs were in the offer but not huge. i think the total comp is just lower than pure tech companies -- they're a consumer brand, the premium isn't there. i negotiated slightly on base and they moved maybe 5k. didn't push harder because i wanted the role.

mobile_mara

the autocomplete with debouncing problem is a classic and i've seen it at three different companies now. if you can't build that cleanly in an interview you need to drill it until it's automatic. race condition handling is the part that separates good answers from okay answers.