Every business-school dissertation cohort starts at the same chokepoint, and it is not supervision — it is getting a proposal approved in the first place. A committee built to review a handful of well-formed proposals a week is instead reading a stack of first drafts that are not yet reviewable, plus a subset of company-based consultancy proposals stuck waiting on an external employer’s sign-off nobody in the university controls the timeline for.
The proposal-review committee is doing two jobs it was never resourced for
A proposal-review committee’s actual job is a quality judgement: is this research question answerable in the time available, is the proposed method appropriate, does the student understand what they are committing to. What it spends most of its time on instead, at the volume a taught-masters or MBA cohort produces, is a triage job it was never resourced for — reading proposals that are incomplete, vaguely scoped, or missing the specific information the committee needs to make that judgement at all. A committee meeting budgeted to clear thirty proposals in two hours slows to a crawl when half of them need a round of clarifying questions before the actual quality judgement can even start, and that round-trip, not the substantive review itself, is where most of the calendar time disappears.
Company-based consultancy proposals add a second, external bottleneck
The allocation model already set out in how to allocate supervision in a business school notes that a company-based consultancy project typically requires an employer-liaison call and an access-and-confidentiality agreement before topic approval — and that step sits entirely outside the university’s own review calendar. A proposal-review committee can clear its own queue efficiently and still have a meaningful share of the cohort stuck waiting on an external company’s legal or HR team to countersign a confidentiality agreement, a delay the committee has no lever to accelerate. The practical fix most business schools under-use is decoupling the two approvals: give a company-based proposal conditional academic approval on its research design while the external agreement is still in progress, rather than holding the whole proposal in limbo until both pieces land simultaneously. That does not remove the external delay, but it stops it from also blocking the academic review calendar.
The scale of this problem tracks the proportion of the cohort choosing company-based projects, which varies significantly by programme — an MBA cohort with strong industry ties can see a majority of students pursuing a live-client consultancy project, while a more research-oriented MSc cohort may see only a handful. A committee that has not measured its own split between company-based and secondary-data or literature-based proposals is working from an assumption about where its bottleneck actually sits, rather than evidence; the fix in each case is different enough — conditional approval for the company-based share, readiness checklists for the rest — that the split is worth measuring directly from the last two cycles’ proposal records before deciding where to invest committee time.
What “ready to review” actually requires
Most of the proposals that stall a committee are missing the same small set of things: a research question specific enough to be answerable rather than a broad topic area, a stated method with enough detail to judge feasibility (not just “qualitative interviews” but how many, with whom, accessed how), and for a company-based project, confirmation that data access has at least been discussed with the employer contact, not assumed. A published proposal-readiness checklist, distributed before students draft rather than discovered when a proposal bounces, moves this triage work upstream to the student and their personal tutor, so what actually reaches the committee is substantially more often review-ready on the first pass. This is the same logic already applied to writing-quality calibration generally in how to write a departmental writing standard for theses: publish the standard before the work is produced, rather than correcting it after submission.
A worked example: where the hours actually go
Consider a business school running a proposal-review committee of five academics, meeting weekly, each meeting budgeted at two hours to clear roughly thirty proposals — four minutes of substantive discussion per proposal, which is tight but workable for a well-formed submission. Now suppose 40 percent of submitted proposals arrive missing a clearly answerable research question, a specified method, or, for company-based projects, any confirmation that data access has been discussed with the employer contact. Each of those does not just take longer in the room; it generates a clarification request, a wait for the student’s response, and a second read at a future meeting — effectively doubling the committee’s real workload on that 40 percent, which in practice consumes the majority of the meeting’s total time even though it is a minority of the proposals. Multiply that across a 200-student annual cohort and the arithmetic is stark: roughly 80 proposals require a second read that would not have been necessary with a clearer first submission, consuming committee hours equivalent to running an entirely separate review cycle that exists only to compensate for underspecified first drafts. The committee’s own sense that “proposals are taking too long” is accurate, but the fix it usually reaches for — adding more meeting time — treats the symptom. The actual lever is upstream: reducing how many proposals arrive in the 40 percent category in the first place.
Common failure modes in how business schools try to fix this
Three fixes recur, and only one of them reliably works. First, adding more committee meetings or members increases capacity to process the same proportion of underprepared proposals, but does nothing to reduce that proportion — the bottleneck moves, it does not shrink. Second, a stricter rejection policy (bouncing anything incomplete without discussion) shifts the triage cost from the committee to the student and their personal tutor, which can work, but only if the readiness criteria are published clearly enough that a rejection is unambiguous and a student can self-correct without a second round-trip through the committee to find out what was actually missing. Third — the fix that actually addresses the root cause — is moving the readiness check earlier, into the drafting process itself, so a proposal is substantially more likely to already meet the bar by the time it reaches committee. The first two redistribute the existing cost; the third reduces it.
Sizing the problem against your own cohort
A business school running the seasonal intake pattern described in the masters dissertation season capacity cliff faces the proposal-review bottleneck at its sharpest in exactly the same narrow window the rest of the support system is already strained — proposals typically cluster in the weeks immediately after topic areas are announced, well before the write-up-stage pressure the capacity-cliff piece describes. A committee sized to clear proposals at an average weekly rate across the term will still bottleneck badly if 70 percent of a 200-student cohort’s proposals arrive in the same three-week window, which is the more common real pattern than a smooth weekly trickle. Checking your own committee’s historical proposal-submission dates against its actual meeting calendar — not the assumed even distribution — is usually the fastest way to find out whether the bottleneck is capacity or scheduling. Sizing this against the wider staff-capacity picture set out in how many postgraduates and how many staff to supervise them is also worth doing before concluding the fix is more committee members: if the ratio of academic staff to postgraduates is already stretched sector-wide, adding standing committee capacity competes directly with the supervision hours the same staff owe once proposals are approved, which is exactly the workload already accounted for in the allocation model referenced above.
Where Tesify fits
The proposal-readiness gap described above is exactly the kind of repetitive, structural writing problem a supervised drafting tool is built to close. The Tesify AI Writing Editor for Institutions walks a student through a structured proposal draft — research question, method, scope, and for a company-based project, a data-access statement — inside a workspace the personal tutor can see before the proposal is submitted to committee, so the readiness checklist above is enforced by the drafting process itself rather than relying on students to self-check against a document they may not reread. Tutors catch an underspecified method or a missing data-access statement at the drafting stage, where a five-minute comment fixes it, instead of at committee, where it costs a full review cycle. The same Tesify for Institutions workspace a student uses for the proposal carries through to the dissertation itself, so a proposal’s approved research question and method are the same document the student continues drafting against, rather than a proposal artefact that gets filed and forgotten the moment committee approves it. A free departmental pilot lets one intake test this against your committee’s own proposal volume before any procurement decision.
Frequently asked questions
What does this cost?
The departmental pilot is free and scoped to one proposal cycle. Institutional licensing beyond the pilot is priced against your cohort size once the pilot shows a measurable reduction in proposals needing a clarification round before committee.
Where is student data processed, and does that raise GDPR or FERPA concerns?
Data residency and processing terms are set out in the institutional data processing agreement provided before any pilot begins. A proposal drafted on the platform is treated with the same data-protection rigour as any other student academic work — confirm the specific residency terms relevant to your jurisdiction with your procurement or data-protection office before the pilot starts.
Does this replace the committee’s judgement on whether a proposal is approved?
No. It structures the student’s draft against a published readiness checklist so the committee spends its time on the substantive quality judgement — feasibility, method fit, scope — rather than on identifying what information is missing.
What is the integration effort for one intake to pilot this?
Minimal. A departmental pilot does not require SSO or LMS integration to begin; students access the workspace directly, and a wider institutional rollout, if the pilot succeeds, is the point at which SSO and LMS integration become relevant.
Does using a structured drafting tool for a proposal raise an academic-integrity question?
No. The tool prompts the student to specify their own research question, method and scope; it does not generate the proposal’s substantive content, and every draft is visible to the personal tutor exactly as any other supervised drafting stage would be.
Does this help with the external employer-agreement delay on company-based proposals?
Not directly — that delay sits outside the university’s process. What it can do is make the academic half of a company-based proposal (research question, method, scope) ready for conditional approval while the external agreement is still in progress, rather than the whole proposal waiting on both pieces at once.
How is this different from the supervision-allocation model already covered on this site?
The supervision-allocation model addresses workload once a project is approved and a supervisor is assigned. This addresses the earlier gate — getting the proposal approved at all — which is a different bottleneck with a different fix.
Should the committee simply reject incomplete proposals without discussion?
It can, and some schools do, but only fairly if the readiness criteria are published clearly enough that a student can identify exactly what was missing and resubmit without a second committee cycle. A rejection against unpublished or informal criteria just moves the frustration downstream rather than fixing the throughput problem.