Guide · · 2 min read
Intern surveys that lead to decisions
Most intern surveys get read once. How to tie every survey question to a decision or a follow-up, and connect answers to offer outcomes.
Jacob Moeckler
Co-founder & CEO
Most programs survey interns two or three times a summer, and the results get read once. Usually the questions were never tied to a decision anyone planned to make.
Map every question to a decision you'll make or a person you'll follow up with. If you can't name either, cut the question.
Three moments, three jobs
| Moment | What it's for | Example decision |
|---|---|---|
| Before day one | Readiness and logistics risk | Who needs a call about housing or onboarding |
| Midpoint | Manager relationship, workload, intent | Which manager needs a nudge; which intern is at risk |
| Exit | Experience, likelihood to accept, recommendation | Program changes for next year; offer risk by name |
The midpoint survey is the one most teams skip and the one with the most room to act. By the exit survey, the summer is over.
Questions that earn their place
| Question | The decision it informs |
|---|---|
| If offered, how likely are you to accept a full-time role? | Who needs attention before offers go out |
| How often do you get feedback from your manager? | Which managers need coaching mid-summer |
| How meaningful is your project to your team? | Which projects to redesign next year |
| How likely are you to recommend this program to a classmate? | Campus reputation, by school |
A few design rules
- Mostly closed questions. Scales you can compare across cohorts and years. One or two open questions at most.
- Under five minutes. Interns are busy, and long surveys get rushed answers.
- Decide on anonymity up front. Anonymous surveys get candor; identified ones let you act on a specific person. You can't follow up with someone you can't see, so many teams identify the intent questions and keep the experience questions anonymous.
- Close the loop. Tell interns what changed because of what they said. It's the best way to raise response rates next time.
Connect answers to outcomes
The most useful analysis happens later. Did low midpoint scores predict declined offers? Did certain projects correlate with acceptance? That only works if survey responses stay connected to each person's outcome.
From Jacob: One student told me her company's recurring surveys felt like sending feedback into a void, especially after her project was canceled when her manager was laid off. Another had the opposite experience: daily check-in surveys for the first five weeks, then weekly, plus regular one-on-ones with her manager. But she was clear that it came from one engineering manager, not the company. Students can tell when a survey leads somewhere. The good news is that it's a design choice, not luck.
FAQ
How many surveys is too many?
Three a summer is plenty. More than that and response quality drops.
Should managers see the results?
Aggregated by team, yes. Individual responses depend on the anonymity choice you made up front.
How teams run this in Springbird
Springbird keeps every survey tied to the people who answered it and what happened next.
- Pre-start, midpoint and exit surveys sent to the group you choose, with reminders until each one closes.
- Responses linked to each person's outcome, so you can see what actually predicted an accepted offer.
- Follow-up groups that keep themselves current, like "low midpoint score, no manager check-in."
- Scales you can compare across cohorts and years on one dashboard.