Results
Every rate here shows the count it came from.
If you came to this page first, you are the reader we wrote it for. Its job is to be more honest than you were expecting, including where that costs us the better-looking number.
First, the definition
What we mean by retention.
Programme reporting usually leaves this undefined, which is what lets a rate move depending on who is asking. Ours is fixed, and it is the definition used everywhere else on this site.
- Still in role — employed by the absorbing firm, in the position they were absorbed into.
- Still in sector — employed in engineering practice, though not necessarily by the absorbing firm.
- Retention, as reported below, means still in role — the strictest of the three, and the one a funder actually cares about.
Someone who leaves the absorbing firm for another engineering practice is counted as a loss in the table below, even though the country keeps the engineer. We would rather understate it than argue about it later.
The numbers
Retention by milestone, with denominators.
A milestone only appears once a cohort has actually reached it. Nothing below is projected forward from an earlier milestone.
retention overall — [CLIENT: number still in role] of [CLIENT: total placed to date] people placed, still in role.
| Milestone | Retention | What the cohort at this milestone is |
|---|---|---|
| 3 months | [CLIENT: %][CLIENT: n] of [CLIENT: N] | Everyone placed at least three months ago. |
| 6 months | [CLIENT: %][CLIENT: n] of [CLIENT: N] | Everyone placed at least six months ago. |
| 9 months | [CLIENT: %][CLIENT: n] of [CLIENT: N] | Everyone placed at least nine months ago. |
| 12 months | [CLIENT: %][CLIENT: n] of [CLIENT: number past 12 months] | Too few people to draw a conclusion from. Read the note below before using this row. |
Our twelve-month retention data covers [CLIENT: number past 12 months] people. That's a small number and we're not going to dress it up. The rest of the cohort hasn't been placed long enough to tell you anything about twelve months. What we can show you is the pattern at three, six and nine months, and it's holding.
If a programme shows you a twelve-month figure off a cohort this size without saying so, the figure is not the problem — the reporting is, and you should assume the same treatment has been applied to everything else they showed you.
Employers
Firms that have absorbed our people.
Named references, with permission, and contactable. A reference you cannot phone is a testimonial, and testimonials are worth what they cost to write.
[CLIENT: named employer reference 1]
[CLIENT: role and contact, with written permission]
[CLIENT: named employer reference 2]
[CLIENT: role and contact, with written permission]
[CLIENT: named employer reference 3]
[CLIENT: role and contact, with written permission]
The people
Three of them, in their own words.
Named, with written consent, and photographed the way a firm photographs its own staff — because they are engineers, not case studies.
[CLIENT: learner story 1 — name, qualification, absorbing firm]
[CLIENT: their account, with written consent]
[CLIENT: learner story 2 — name, qualification, absorbing firm]
[CLIENT: their account, with written consent]
[CLIENT: learner story 3 — name, qualification, absorbing firm]
[CLIENT: their account, with written consent]
Learner personal information is published only with documented written consent and is processed in line with POPIA. See the privacy notice.
What we cannot show you
The list we would want if we were buying this.
These are open questions about our own model. They will come up in the conversation anyway; putting them here saves you having to extract them.
- Twelve-month retention rests on [CLIENT: number past 12 months] people. It is indicative at best.
- We have never run a cohort at 200. The largest we have run is 30, and scaling has failure modes we have not met yet.
- Accreditation is under application, not granted. We say so on every page.
- External assessment outcomes are set and moderated independently. We prepare for them; we cannot promise them.
- Employment seats are commitments under agreement between the parties, not guarantees of employment for any learner.
- Engineering placement to date has run largely with one consultancy partner. Breadth across firms is something we are building, not something we have.
2027 seat allocation
Ask us the question this page didn't answer.
We will bring the funnel with its actual counts, the full retention working, and whichever of the limitations above matters most to your board.