Requirements differ by scheme — read the specific call, and involve both research offices earlier than feels necessary.
You spot a call with an international collaboration component, your Vietnamese partner is enthusiastic, and there are eight weeks. Six weeks later the science section is strong, the budget is not agreed, and nobody has started the institutional sign-off.
This is the normal failure mode, and almost all of it is schedule.
Decide three things before writing
Who leads, and what that means. Most schemes require the lead to be based in the funder's country, which fixes who holds the budget and who signs. Say this out loud rather than leaving it implied — it shapes everything downstream.
Who assembles the final document. One named person, holding the master copy and owning the deadline.
Who writes what, with individual deadlines. Write this down and circulate it. Most delay comes from each side assuming the other is drafting a section.
Plan backwards, and allow for their process
The step routinely underestimated is institutional approval on the partner side.
Ask early and specifically: what does your institution need to sign this, who signs, and how long does it take? Expect longer than your own office. Some institutions require sign-off from a level above the department, and that queue is not one your partner controls.
A workable schedule: final document one week before the deadline; full draft to both research offices two weeks before; budget agreed three weeks before; complete first draft of all sections four weeks before; structure and roles fixed six weeks before.
If the call leaves less than six weeks and you have no prior joint material, consider waiting for the next round. A rushed joint proposal is usually a weak one, and a partner pushed through an unfamiliar internal process under time pressure is a poor start to a relationship.
Write the partner's role so reviewers value it
Reviewers of collaboration schemes look for whether the partnership is real. A section describing the Vietnamese group as providing site access and sample collection signals that it is not.
Three things to do instead:
Give them intellectual ownership of specific work packages, including analysis and writing — not only data collection.
State concretely what they bring that the project cannot proceed without: a long-running dataset, access to a population or field system, methodological expertise in that context, an existing relationship with a ministry or local authority.
Let them write their own section. They know their institutional capabilities and their track record; a version you draft on their behalf will be generic, and it removes the ownership reviewers are looking for.
Then read their draft and ask for specifics where it is vague — many partners write modestly about their own contribution, and a direct request for concrete detail usually produces a much stronger section.
Budget: coordinate, do not assume
Four things to establish early:
What the scheme permits for the partner share, and whether there is a cap.
What their institution's rules allow — rates, permitted categories, whether indirect costs can be claimed.
The real costs on their side, including ones they may not think to raise: coordination time, instrument or bench fees, translation, bank charges and exchange differences, and any institutional levy.
A cross-check so nothing is double-counted or falls between the two budgets.
Do this alongside the writing, not at the end. Budget work exposes problems in the work plan, and finding them early fixes both.
Practical coordination
One shared document, not files exchanged by email. Version confusion costs more than it seems.
Comment rather than overwrite on sections someone else owns.
Account for the time difference in the schedule. A question sent at the end of your day is answered at the end of theirs, so each exchange costs a day. In the final fortnight, agree one overlapping hour per week for anything that needs a real conversation.
One person reads the whole thing at the end for consistency of voice. A proposal that reads as four separate documents is immediately recognisable.
What commonly goes wrong
Silence read as agreement. If your partner has not commented on the work plan, do not assume they accept it. Ask directly whether the timeline is feasible on their side.
Paperwork left to the end. Institutional letters, legal entity details, CVs in the funder's format, financial documentation — ask for the list in week one.
The partner's costed effort set too low. A partner who cannot fund adequate staff time will under-deliver, and it will be visible in the mid-term report rather than now.
No agreement on what happens if funded. Authorship, data rights and decision-making are far easier to settle while the outcome is still uncertain — and a short memorandum agreed during proposal writing costs an hour.
If it is not funded
Success rates on these schemes are low, so this is the likely outcome.
Keep the material — most of it transfers to the next call. Read any reviewer feedback together rather than forwarding it, and agree what to change.
And tell your partner promptly either way. A partner who invested weeks and then hears nothing for months learns something about the collaboration that is hard to unlearn.
What is most often underestimated?
Institutional approval on the partner side — ask early what their institution needs to sign, who signs, and how long it takes.
How should the partner's role be written?
Give them intellectual ownership of specific work packages including analysis and writing, state concretely what only they bring, and let them draft their own section.
What should the budget coordination cover?
Scheme rules on the partner share, their institution's permitted rates and categories, real costs they may not raise, and a cross-check against double-counting.
What should be settled while writing?
Authorship, data rights and decision-making — far easier to agree while the outcome is still uncertain, and a short memorandum costs an hour.
