Most internship programmes are really a series of one-off internships. They work when a motivated manager runs one, and they stop when that manager is busy — because nothing was written down, so the next cycle starts from zero.
The difference between a programme and a series of internships is about a dozen artefacts. This is the list. If you want the reasoning behind the design choices rather than the list itself, the programme design essay in Recruiter Intelligence covers it.
Decide these before anything else
Documents to have before cycle one
Roles that must have a name against them
| Role | What they own | Failure if unfilled |
|---|---|---|
| Programme owner | The cycle calendar, the artefacts, the cohort data. | The programme becomes a series of one-off internships again. |
| Executive sponsor | Budget, conversion seats, and defending the programme at planning time. | Programme is cut in the first tight quarter. |
| Mentor (per intern) | Scope, weekly one-to-one, reviews, the certificate line. | The single most common cause of a failed internship. |
| Buddy (per intern) | The questions the intern will not ask their assessor. | Slower ramp, and problems surface late. |
| IT owner | Access provisioned before day one. | Interns lose their first week waiting. |
Running the cycle
Measure these, and only these
| Metric | Why it earns its place |
|---|---|
| Deliverable completion rate | Tells you whether projects were scoped realistically. The first thing to fall when they are not. |
| Mentor hours actually delivered | Predicts almost every other outcome. Ask the interns, not the mentors. |
| Conversion rate, and offer-accept rate on conversion | Two numbers, not one. A high conversion rate with low acceptance means you waited too long to decide. |
| Intern NPS or "would recommend to a friend" | Your campus reputation, measured before it reaches campus. |
| Cost per converted hire | The number to take to the budget conversation. Cost per intern is not. |
Frequently asked
How many interns should a first cycle have?+
Enough for a cohort, few enough to run properly — five to ten is a good first cycle for most companies. One or two interns never becomes a programme because nothing gets written down; thirty in cycle one overwhelms an owner who is still learning what breaks.
Who should own an internship programme?+
One named person with time actually allocated. Programmes owned by "HR" collectively, or by a manager as an added duty, run once well and then quietly stop when that person is busy.
What is the difference between an internship programme and just hiring interns?+
Written artefacts and a repeatable calendar. Hiring interns is a decision each manager makes; a programme is a set of documents and dates that let the next cycle start from where the last one finished.
Should every team be allowed to host an intern?+
Only teams that can name a mentor with real capacity and produce a written scope. Making those two things a precondition is the single highest-leverage rule in a programme, and it is the one that gets waived first under pressure.
How do we know whether the programme is working?+
Five metrics: deliverable completion, mentor hours actually delivered, conversion and conversion-acceptance, intern would-recommend, and cost per converted hire. Ask interns for the mentor-hours number rather than mentors — the two answers routinely differ.