Allocating dissertation supervision in a business school means converting a cohort of MBA and MSc dissertation topics into a workload each supervisor can actually carry, weighted by project type rather than headcount alone. The seven steps below take a programme from an unweighted list of topics to a signed-off allocation matrix, each with a named owner and a produced artefact.
Step 1: Inventory the incoming cohort by dissertation type, not just by number
Owner: programme administrator. Artefact: a cohort-by-type spreadsheet.
A business school’s dissertation cohort is rarely one population. A typical MSc Management or MBA cohort splits across three project types with very different supervision costs: a company-based consultancy project for a live client organisation, a secondary-data or literature-based dissertation, and a quantitative empirical dissertation using survey or archival data. Before any allocation model runs, list every registered or provisionally registered student against which type they have proposed, because the weighting in Step 3 depends entirely on this split being accurate rather than assumed.
Step 2: Convert supervisor availability into supervision-capacity units
Owner: associate dean for education. Artefact: a workload allocation model (WAM) entry per supervisor.
Most business schools already run a workload allocation model that converts teaching, research and administrative duties into a shared unit, commonly “points” per full-time-equivalent (FTE) academic year. Dissertation supervision needs its own line in that model rather than being absorbed into a general “pastoral and other duties” category, because it is the line accreditation reviewers ask about directly. AACSB, EQUIS and AMBA accreditation reviews all expect a business school to be able to show how supervision workload is planned and monitored, not just that dissertations get supervised somehow. A common practical benchmark used in workload models is a base capacity of six to ten supervisees per FTE academic per annual cycle for a taught masters programme, before the type-weighting in Step 3 is applied — treat this as a starting point to calibrate against your own school’s historical hours-per-dissertation data, not as a fixed rule.
Step 3: Weight the allocation by project type, not by headcount
Owner: dissertation coordinator. Artefact: a weighting table appended to the WAM.
A company-based consultancy project carries meaningfully more supervisor time than a secondary-data dissertation: it typically requires an employer-liaison call before topic approval, an access and confidentiality agreement, at least one site or virtual client check-in during the project, and often a data-sharing or ethics review that a literature-based project does not trigger. A defensible weighting scheme assigns each project type a multiplier against the base unit from Step 2 — for example, treating a company-based project as 1.4–1.6 times the supervision cost of a secondary-data dissertation, and a quantitative empirical dissertation requiring statistical method supervision at roughly 1.2–1.3 times. The exact multipliers should come from your own supervisors’ logged hours in the previous cycle wherever that data exists; treat the ranges above as a starting hypothesis to test, not an imported constant.
Step 4: Run a preference-and-match allocation round
Owner: dissertation coordinator. Artefact: a signed allocation matrix.
With weighted capacity established, match students to supervisors on expertise fit first and remaining capacity second. A simple, auditable method that avoids first-come-first-served disputes is a two-round preference process: students submit a ranked shortlist of three supervisors by topic area, supervisors independently rank the proposals that fall inside their expertise, and the coordinator runs a matching pass that respects both rankings up to each supervisor’s remaining weighted capacity. Company-based projects with an already-secured external partner should be matched first, since the client relationship constrains which supervisor can realistically take the project regardless of preference.
Two situations need a pre-agreed tie-break rule rather than an ad hoc decision on the day: two students ranking the same supervisor first with that supervisor at exactly one remaining capacity unit, and a student whose top three preferences are all already full. For the first, a documented tie-break — commonly proposal submission date, or where the programme runs one, an academic-merit ranking from the preceding term — keeps the process defensible if either student later queries the outcome. For the second, the coordinator should have standing authority to route the student to the overflow pool from Step 5 rather than forcing a fourth-choice match with a supervisor outside the student’s stated interests, since a poor topic fit tends to cost more supervision time over the life of the project than the mismatch it was meant to avoid.
Step 5: Build a named overflow pool before you need it
Owner: associate dean for education. Artefact: a standing list of qualifying adjunct or industry-practitioner co-supervisors.
Business school cohorts fluctuate year to year more than most faculties, driven by MBA intake cycles and corporate-sponsored programme numbers that are set outside the academic calendar. Rather than discovering a capacity shortfall after Step 4 has already run, maintain a standing list of qualifying adjunct faculty or senior industry practitioners approved in advance to co-supervise company-based projects specifically, since that is the project type most likely to spike unpredictably when a corporate partnership brings in a larger-than-expected cohort. Pre-approval matters because most institutions require a research-integrity and safeguarding check before anyone touches student supervision, and that check should not be the rate-limiting step in an already-tight allocation cycle.
Step 6: Set an explicit cap and an escalation path
Owner: programme director. Artefact: a one-page cap policy.
Publish the maximum weighted load any individual supervisor can carry, and name who a supervisor escalates to when their queue would exceed it — typically the dissertation coordinator in the first instance and the associate dean where a rebalancing across the whole cohort is needed. Without a published cap, the practical ceiling on any one supervisor tends to be set informally by whoever is least willing to say no, which is neither equitable across faculty nor visible to an accreditation reviewer asking how overload risk is managed. This is the same underlying problem addressed programme-wide in how to train supervisors to use and govern AI writing tools, where consistent supervisor practice is likewise built on an explicit, published standard rather than individual habit.
A one-page cap policy needs only four fields to be usable in committee. Template language:
Maximum weighted supervision load: [X] units per FTE per annual cycle
Escalation route: supervisor to notify the dissertation coordinator in writing
within [N] working days of any assignment that would exceed the cap
First-line resolution: dissertation coordinator reassigns within the
overflow pool or adjusts the weighting table for the affected project
Second-line resolution: associate dean approves a cohort-wide rebalancing
where the shortfall exceeds [Y] units across the programmeKeeping the policy to these four fields is deliberate: a longer document tends to accumulate exceptions that quietly erode the cap, whereas a short one is easy enough to actually follow during a busy allocation cycle.
The same documentation also does double duty at accreditation review. AACSB’s standards on faculty sufficiency and deployment, EQUIS’s process documentation for programme quality assurance, and AMBA’s criteria on faculty-student engagement each ask, in different language, for evidence that supervision workload is planned rather than absorbed informally. A dated cap policy, a weighting table with its calibration source noted, and a rebalancing log from the most recent cycle answer that question directly, which is a lighter lift during a self-study than reconstructing the same evidence retrospectively from individual supervisors’ recollection.
Step 7: Rebalance at census date and feed actuals back into next year’s model
Owner: programme administrator, with sign-off from the associate dean. Artefact: a rebalancing log and a revised WAM formula for the following cycle.
Late registrations, withdrawals and topic changes mean the allocation matrix from Step 4 is rarely still accurate by the programme’s census date. Run one scheduled rebalancing pass at census, log every change and the reason for it, and — critically — feed the actual hours supervisors logged against each project type back into next year’s weighting table from Step 3. A workload model that is never recalibrated against its own prior-year actuals drifts further from reality every cycle it runs.
Where this connects to the wider capacity plan
Business schools with a large taught-masters intake are exactly the population that produces the seasonal capacity spike described in the masters dissertation season capacity cliff: a single submission date, an entire cohort, and a support establishment sized for term-time demand rather than a peak. The allocation model above manages supervisor-side load; the staff-to-student denominator it is built on, and how that denominator compares against sector figures, is set out in how many postgraduates and how many staff to supervise them. Programmes further along in institutional AI adoption should also connect the supervision-capacity model to the departmental pilot process in how to run a departmental pilot of an AI writing tool and the documentation standard in how to write a departmental writing standard for theses, since a consistent writing standard reduces the variance in supervision time that Step 3’s weighting table has to absorb.
If you would like help building this allocation model against your own business school’s cohort data, request an institutional evaluation and we will work from your actual workload figures rather than the illustrative ranges above.
Frequently asked questions
How many dissertations should one business school supervisor take on?
There is no universal number; it depends on the supervisor’s other workload commitments and the project-type mix. A common starting benchmark is six to ten supervisees per FTE academic per cycle before type-weighting, calibrated against your own school’s historical supervision hours.
Why should company-based consultancy projects count for more than a literature-based dissertation?
They typically add employer liaison, an access or confidentiality agreement, at least one client check-in, and sometimes an ethics review — time costs a purely literature-based dissertation does not carry. A defensible model weights for this rather than counting every dissertation equally.
Do accreditation bodies actually check dissertation supervision workload?
AACSB, EQUIS and AMBA reviews all expect a business school to demonstrate how supervision workload is planned and monitored as part of faculty sufficiency and quality assurance, which is why an explicit, documented allocation model is worth having beyond its day-to-day operational value.
What happens when a supervisor’s allocated load exceeds the published cap?
The supervisor should escalate to the dissertation coordinator first, and to the associate dean if a cohort-wide rebalancing is needed, following the escalation path set out in the cap policy rather than absorbing the overload informally.
Should adjunct or industry co-supervisors be approved in advance?
Yes. Pre-approving a standing pool avoids the research-integrity and safeguarding check becoming the rate-limiting step during an already-tight allocation cycle, particularly for company-based projects where intake can spike unpredictably.
When should the allocation matrix be rebalanced?
At the programme’s census date, after late registrations, withdrawals and topic changes have settled, with every change logged and the actual hours fed back into the following year’s weighting table.
Is a spreadsheet enough to run this, or does it need dedicated software?
A spreadsheet is sufficient for most single-programme allocations. Dedicated workload software becomes worthwhile once a school is running the model across multiple programmes with shared supervisor pools, where manual reconciliation of overlapping capacity becomes error-prone.
How does this differ from allocating supervision in a smaller, research-heavy department?
A smaller doctoral-heavy department carries a continuous, individually supervised load rather than a seasonal cohort peak, which is a fundamentally different shape of demand than a business school’s taught-masters intake typically produces.
How should two students who both rank the same supervisor first be resolved?
Use a pre-agreed tie-break rule — commonly proposal submission date or a documented academic-merit ranking — rather than a case-by-case decision, so the outcome is defensible if either student queries it later.
What should go in a one-page cap policy?
Four fields are enough: the maximum weighted load per FTE, the escalation route and timeframe, the first-line resolution owned by the dissertation coordinator, and the second-line resolution owned by the associate dean for cohort-wide rebalancing.