The application season for research internships is approaching. These roles pay well (often three to four times a PhD stipend), connect students with people rarely encountered in academia, provide resources and platforms beyond what universities can offer, and often allow them to spend the summer in a new city for a refreshing break from the routine of PhD life. Most importantly, they give students an edge when entering the full-time industry job market.
During my PhD, I completed five research internships at Adobe (2x), Google (2x), and Meta. In my first year as a research scientist at Adobe, I’ve interviewed, made offers to, and mentored four PhD interns. Having experienced internships from both sides, as a mentor and a mentee, I am writing to share some advice on how to land and excel in a research internship. My experience is mostly with HCI interns, but I believe many of these insights generalize to other subfields of CS.
Except in rare cases, there’s no guaranteed way to land a research internship, but you can increase your chances by optimizing a simple formula:
<aside> 🔥
Likelihood = Attractiveness × Visibility
</aside>
Attractiveness is approximately a weighted sum of
(W₁ × track record) + (W₂ × research fit) + (W₃ × interview performance).
The weights vary across hiring teams. Some hire for pre-defined projects where research fit matters most; others prioritize strong track records, e.g., publications, awards, prior internships, and simply hire the best, figuring out what to work on later. Interview performance also matters, but it’s mainly about gauging your vibe and communication style rather than testing skills, since research ability can’t be measured in a one-hour interview. For that reason, this post won’t focus on interviews.
In general, I’d advise focusing on building a strong track record. A student with several first-author CHI/UIST papers under their belt naturally appears more ready for an internship than one without, especially since internships typically require completing a project within 12–14 weeks.
Visibility is the other multiplier. Even with great credentials, if hiring managers don’t know you exist, your chances are close to zero.
To boost visibility, two paths:
I’m a PhD student at [University] working on [Topic]. I’m very interested in the position at Adobe Research on your team and wonder if we can connect to discuss potential opportunities.
These get ignored. Instead, write with purpose: show who you are, why you’re interested, and what you could work on together:
I’m a PhD student at [University] and see strong overlap between my research on [Topic] and your recent work on [Topic]. I’d be very interested in exploring a potential internship opportunity with you at Adobe Research. I’ve published [N] papers on [ABC] and have hands-on experience with [DEF].
If I were to work with you, I see three possible directions we could explore together:
- [Direction 1]
- [Direction 2]
- [Direction 3]
Of course, don’t use this template verbatim. What I want to illustrate is that a thoughtful, personalized email that clearly articulates why you want to work with the researcher(s) and how your work connects to theirs shows credibility, interest, and initiative—all of which help you stand out. Still, don’t expect replies from everyone; researchers receive many such emails and typically respond only to the most relevant ones.