An internship with one review at the end is not an assessment, it is a verdict. The intern learns what they should have done differently on the day it stops mattering, and the company learns whether the internship worked at the point where nothing can be changed.
Two reviews fix this. The midpoint review exists to change the second half; the final review exists to decide conversion and to send the intern away with something true and useful. Both forms are below.
The midpoint review — copy this
MIDPOINT INTERNSHIP REVIEW Intern: ______________ Mentor: ______________ Week: ___ of ___ Purpose: to change the second half. Nothing here decides conversion. 1. THE PROJECT Agreed deliverable: ___________________________________ Where it stands today: ________________________________ On track / behind / ahead (circle): ______ 2. WHAT IS GOING WELL Two specific things, with an example of each: • ______________________________________________________ • ______________________________________________________ 3. WHAT NEEDS TO CHANGE IN THE SECOND HALF At most two things. Behaviour, not personality. With an example: • ______________________________________________________ • ______________________________________________________ 4. WHAT I (THE MENTOR) WILL DO DIFFERENTLY ________________________________________________________ (If this is blank, the review is incomplete. Most intern underperformance in the first half is a scoping or access problem.) 5. WHAT THE INTERN NEEDS FROM US ________________________________________________________ 6. AGREED FOR THE SECOND HALF Deliverable: ____________________________________________ Checkpoint date: ________________________________________ Signed (mentor): __________ Signed (intern): __________ Date: ______
The final review — copy this
FINAL INTERNSHIP REVIEW Intern: ______________ Mentor: ______________ Dates: ______ to ______ Nothing in this review should be new to the intern. 1. DELIVERABLES Agreed at the start: ____________________________________ Delivered: _____________________________________________ Still open, and why: ___________________________________ 2. ASSESSMENT (1–4, same anchors used at interview) Problem solving [ ] Evidence: ____________________ Craft in the role [ ] Evidence: ____________________ Communication [ ] Evidence: ____________________ Ownership [ ] Evidence: ____________________ Coachability [ ] Evidence: ____________________ 3. TRAJECTORY Compared with the midpoint review, this intern has: improved / held / declined on: __________________________ The clearest evidence of change: ________________________ 4. CONVERSION RECOMMENDATION Convert now | Convert to a different role | Re-intern | Do not convert Reason, in one sentence the intern could be shown: ________________________________________________________ 5. WHAT I WOULD TELL THEIR NEXT MANAGER ________________________________________________________ 6. WHAT I WOULD TELL THEM TO WORK ON One thing. Specific enough to act on this month: ________________________________________________________ 7. CERTIFICATE AND REFERENCE Completion certificate issued: Y / N Date: __________ Written reference offered: Y / N Signed (mentor): __________ Signed (intern): __________ Date: ______
How to write feedback an intern can act on
- Behaviour, not personality
- "You went quiet for three days on the blocked ticket" is actionable. "You lack initiative" is a label, and it will be remembered as one.
- One thing, not six
- A first-workplace review with six development areas produces zero changes. Pick the one that would most change their next six months.
- Evidence, always
- Every score needs an example. Without one, the review is an impression with a number attached, and it will not survive being questioned.
- Say the hard thing plainly
- Interns who are told "you did great" and then not converted learn nothing and tell everyone. Kind and clear beats kind and vague, every time.
How the review feeds conversion
The conversion decision should be readable from the two reviews and nothing else. If a decision is being made on impressions that appear in neither review, the process has failed somewhere earlier — usually because the deliverable was never agreed in writing, so there was nothing objective to assess against. Tell interns their outcome within a couple of weeks of the internship ending, either way. The most damaging thing an internship programme does is not saying no; it is saying nothing.
Frequently asked
How many reviews should an internship have?+
Two written ones — a midpoint and a final — plus a weekly one-to-one that is a conversation, not a form. One review at the end is a verdict rather than an assessment, and it wastes the half of the internship where feedback could still have changed the outcome.
Should interns see their reviews?+
Yes, both of them, in writing. A review the intern never sees serves only the company, and the intern learns nothing. The rule worth holding is that nothing in the final review should be new information.
What if the intern is underperforming at the midpoint?+
Ask first what was wrong with the scoping or the access, because that is the cause more often than the intern. Then agree one specific change and one checkpoint. Interns rescued at the midpoint are common; interns rescued in the last week are not.
Should the review include a numerical rating?+
Use the same 1–4 anchors you used at interview, so the two are comparable and you can see trajectory. Avoid inventing a separate percentage or ten-point scale — it adds precision that does not exist and invites arguments about the number.
How soon should we tell an intern about conversion?+
Within about two weeks of the internship ending, either way. Interns who hear nothing assume no and tell their whole cohort — you get the reputational cost of rejecting them without ever having made a decision.