Designing a scheme, starting with whether to run one
Designing a mentoring programme means fixing five things before any recruitment starts: purpose, population, dose, duration and exit.
Anyone at the point where a mentoring programme is still an idea on a page, before money is spent and before a volunteer has been told anything.
The people who would be referred into it. Design errors are cheap to fix now and are paid for later by somebody who had no say in them.
The scheme is designed backwards from the funding: purpose written to match the grant, duration set by the length of the funding round. That is how a twelve-month relationship ends up with a six-week budget behind it.
Write those five down and get them signed off, and everything downstream follows from them, from the advert to the closing letter. Leave one to be settled later and it will be settled by accident, by whoever is nearest when the first match goes wrong.
Whether a one-to-one volunteer scheme is the right instrument at all, rather than the one that happens to be fundable, is a prior question and not a design question: it is settled before any of the five below, and the tests that settle it belong with running a scheme.
Purpose and population: what the scheme is for, and who for
The purpose is one sentence naming who the scheme serves, what changes for them and how the change would be visible. Youth mentoring gives a referred young person a consistent non-family adult across a school year, and the visible change is attendance, engagement, or who the young person says they would go to with a problem. Befriending gives an older person regular company, measured on loneliness. Goals are then written so that somebody who dislikes the scheme could check them:
- Hold 25 active matches by month nine, with no coordinator over the stated maximum.
- State the proportion of youth matches the scheme intends to keep running to twelve months, hold the scheme to that figure at review, and record the reason for every early close.
- Complete a loneliness measure at the start, at six months and at close.
Population is written as an eligibility rule somebody in the referring service could apply without telephoning the coordinator: the age range or circumstance that brings a person in, the geography, the referral route, and, said out loud, who the scheme is not for. Most designs name who is in and leave who is out to be discovered at the first difficult referral. Write the exclusions down instead, whether that is acute crisis, a need for personal care or a person who cannot safely be seen alone, because an exclusion held only in the coordinator’s head arrives as a refusal that looks arbitrary.
Check what already reaches that population first. Age UK and other national charities run befriending services across the country, and health services, councils and community organisations run their own. Where a population is already served, the honest contribution is often a referral route rather than a second scheme.
Programme and program are the same word; program is the American spelling.
Dose: how much contact, how often
Dose is the amount of contact the design commits to, and it decides whether the scheme helps or harms. Set it at a level volunteers can sustain for the whole duration, because a dose quietly halved after four months is a broken promise, not a flexible one. Set it as a floor and a ceiling rather than an aspiration: the contact below which the coordinator intervenes, and the contact above which the volunteer has moved into something the scheme has neither trained nor insured them for.
| Type of scheme | Standard dose |
|---|---|
| Youth mentoring | About 1 hour a week |
| Befriending, face to face | 1 hour a week or a fortnight |
| Telephone befriending | 20 to 45 minutes, weekly |
Duration and exit: how long, and how it stops
Duration and exit are designed together, because deciding how long a match runs has already decided when it ends. Youth mentoring asks for 12 months, and that is the ask for a reason: matches ending before three months are associated with outcomes worse than no match at all, published by Grossman and Rhodes in 2002. Befriending commonly runs 6 to 12 months and is frequently open-ended for older people, which turns the design problem into a scheduled review both people know about.
Write exit as a procedure: a named end date given at the first meeting, a final session that is marked, a separate conversation with each person afterwards, and a written record of why the match closed. Design it for the bad cases too: the volunteer who moves away in month four, the person who stops answering the door in month seven.
Name the exit criteria too, so that ending a match is a decision anybody can check rather than a judgement made under pressure. Four cover most schemes: the planned duration is reached, the purpose is met early and both people agree to stop, the person supported asks to stop, or the scheme ends the match on safeguarding or capacity grounds. Anything outside those four is an unplanned close, and unplanned closes are the thing to count.
The funding is for six weeks and the referrals are children
The coordinator does not redesign the scheme to fit the money and no volunteer is recruited to deliver it. A six-week mentoring block for a referred child ends well inside three months, and matches ending before three months are associated with outcomes worse than no match at all. Good practice inside those weeks does not compensate for the ending. Ask the funder for a duration that matches the intervention, or spend the money on something that works at that length.
Frameworks, including GROW, and what they do not do
Frameworks such as GROW, which walks a conversation through goal, reality, options and will, are session tools borrowed from coaching. GROW suits workplace and career mentoring, where the mentee arrives with an objective, and fits badly in befriending, which has no goal. It fits badly too in the first months of youth mentoring, where goals demanded before a relationship exists make the hour feel like another appointment.
A framework is a tool inside a session, not a scheme design. What moves the effect of mentoring is programme practice: matching, training, supervision and match length. A well-run scheme beats a badly-run one by more than mentoring beats no mentoring, which is the practical meaning of the small but real effect, around 0.2 of a standard deviation, that DuBois and colleagues found in 2011.
What the design commits each side to
The mentor or befriender
- Accepts the dose and duration as written, before being matched rather than after
- Says early if the commitment has become impossible, so the ending can be planned
- Works to the scheme’s purpose rather than a private one
The young person or the person visited
- Is told the duration and the end date at the first meeting, in words they can use
- Is told what the scheme is for, so they can judge whether it is what they wanted
- Can decline it, and stop it, without affecting any other service they receive
The scheme owns the design and cannot amend it quietly. If dose, duration or exit change once matches are live, both people are told in writing and the change is recorded with its reason. Design that drifts is the ordinary route by which a twelve-month promise becomes a four-month relationship.
Sizing the scheme before you build it
Size the scheme against coordination capacity first and money second, because capacity binds: one full-time coordinator holds 30 to 50 active matches, fewer where referrals are complex. The calculator below applies the working cost band and that capacity range to a target number of matches. It is arithmetic on published ranges, and the width of every band comes from referral complexity, coordinator hours, volunteer travel and expected match length.
Enter a number of matches and choose a scheme type.