How long should a collaboration outreach email be?

Around 200 to 250 words for the cold and conference variants. The project variant is necessarily longer because it has to state the role, the effort, the budget share and the deadline, but even there the reader should be able to decide whether to reply within the first two paragraphs.

Should I attach a proposal or a paper to the first email?

No. Offer a one-page summary instead and send it when they ask. Attachments in unsolicited email go unopened, increase the chance of a spam filter, and make the message feel like a demand for hours of reading. Offering also gives you a natural reason to write a second time.

When should I mention a grant deadline?

In the first paragraph, and in the subject line if it is close. Deadlines are normal in collaborative funding and partners work with them constantly. What damages the relationship is discovering the deadline after they have expressed interest, because it makes the earlier friendliness look like a tactic.

Should I discuss authorship in the first email?

Not in a cold first message. Raise it on the first call, once there is a concrete piece of work to allocate, and settle the rule rather than a fixed list. In the consortium variant it is different: the role and the work package leadership belong in the email, because they are what the person is being asked to agree to.

Does the collaborator finder send these emails?

No. It reads the document you upload and suggests researchers whose published work sits close to it, each with a short reason. It does not show email addresses, does not draft messages and never contacts anyone. Everything after the shortlist is yours, which is what these templates are for.

Research Collaboration Outreach Emails

Three variants for the three ways a collaboration actually starts, each annotated with what makes it work, plus the lines to leave out.

A collaboration request succeeds or fails on two things: whether the reader believes you have actually read their work, and whether the ask is small enough to accept without a meeting. The three variants below cover a cold approach after reading a paper, a follow-up after meeting someone at a conference, and the case where you already have a defined project and a deadline. The reasoning behind the structure is in how to approach a potential collaborator, and what to settle once someone says yes is in the discussion checklist.

Variant 1: after reading their paper

The cold approach, and the one with the highest failure rate when written badly. Everything depends on the second paragraph: a specific engagement with their work that could not have come from an abstract. The ask is a call, nothing larger.

Cold outreach after a paper

Subject: [Their method] applied to [your problem], possible collaboration

Dear Dr [Surname],

I am [role] at [institution], working on [one line: the problem, not your biography].

I read your [year] paper on [specific topic]. The part I keep coming back to is [specific finding, method choice or stated limitation], because we are running into the same thing from the other side: [one or two sentences on what you have and where it stops].

We have [the concrete asset: a cohort of n participants, three years of instrument data, a validated model, ethics approval for x]. What we do not have is [the specific capability they hold]. Combining the two would let us test [the one specific question], which neither of us can answer alone.

Would you have 25 minutes in the next few weeks for a call to see whether this is worth pursuing? I am happy to send a one-page summary in advance if that is useful.

Best wishes,
[Full name]
[Role, institution] | [ORCID or group page]

Engage with one specific element of the paper, ideally a limitation the authors named themselves. Nothing else distinguishes your message from the generic ones they receive.

State what you bring before you state what you need. A collaboration email that only asks is a request for a favour, and it is read as one.

Name the question that neither side can answer alone. That sentence is the entire proposition, and if you cannot write it, the collaboration is not yet an idea.

Ask for 25 minutes. A specific, small number is far easier to accept than an open-ended request to discuss possibilities.

Offer the one-page summary rather than attaching it. Attachments in a first email go unopened, and offering it gives them a low-cost way to say yes.

Do not mention authorship, budgets or grant applications in a first email. Those belong on the call, once there is something to allocate.

Variant 2: after meeting at a conference

The warmest of the three, and the one most often wasted by being sent a month too late. Write within a week, while the conversation is still recoverable from memory, and be concrete about which conversation you were.

Conference follow-up

Subject: Following up from [conference], [the specific thing you discussed]

Dear [First name],

It was good to talk after your [talk or poster] at [conference] last week. You made the point that [the specific thing they said], which I have been thinking about since.

You mentioned [the problem they described or the tool they use]. On our side we have [the complementary thing], and the obvious experiment is [one sentence].

I would like to try a small version of this first: [a pilot that takes weeks rather than months, for example running your method on a subset of our data, or a joint abstract for [next meeting]]. If it works we can decide what a real project looks like.

Are you free for a call in the next two weeks? I can send the data description beforehand.

Best,
[First name]

Send it within a week. After a month, a conference follow-up is a cold email with a claim of acquaintance the reader cannot verify.

Quote something they said. It proves which conversation this was, which matters because they had thirty of them that week.

Propose a pilot rather than a project. A small first piece of work is the fastest way to learn whether a bigger collaboration would function.

Use first names if the conversation did. Reverting to formal address after an informal meeting reads as distance.

Name a concrete forum or deadline, such as an abstract for the next meeting, which gives the collaboration a natural first milestone.

If a student or postdoc introduced you, mention them by name and copy them in. Cutting out the person who made the introduction is noticed.

Variant 3: a defined project, or a call with a deadline

Here you are asking for real commitment, so the message has to carry more information: what the project is, what role they would hold, what it requires of them, and what the deadline is. The rule that matters most is that the deadline appears in the first paragraph, not the last.

Project or consortium approach

Subject: [Funder] [call name], partner in [role], deadline [date]

Dear Prof. [Surname],

We are preparing a proposal for [call name, funder], which closes on [date]. I am writing because the consortium needs [the specific role: a partner in an eligible country, a group with x capability, a clinical site] and your group is the strongest fit we have identified.

The project in three lines: [objective]. [Approach]. [Expected outcome].

What we would be asking of you: [work package or role], roughly [effort: n person-months over m years], with an indicative budget of [amount or share] to your institution. Your group would lead [specific part], and we would expect [named person] to be a work package leader.

What is already in place: [coordinator, other partners, letters of support, prior data]. What is not yet fixed: [the parts still open, so they can shape them].

I know [date] is close. If this is of interest, could we speak this week? I can send the draft outline and the partner template immediately.

With thanks,
[Full name]
[Role, institution]

Put the deadline in the subject line and the first paragraph. Discovering a two-week deadline in the fourth message is the single most damaging thing you can do to a new professional relationship.; Say why them specifically, and be honest if the reason is eligibility rather than science. Partners know how consortia work and would rather hear it.
Quantify the ask: effort, duration, budget share. A vague invitation to join a proposal is impossible to evaluate and is usually declined by default.; Say what is already in place and what is still open. Being able to shape a work package is often the deciding factor, and a proposal presented as finished is a request for a signature rather than a partnership.; Name the intended role and the person who would lead it. Roles decided later become authorship disputes later.; Copy in your research office early. Consortium calls usually require an institutional agreement between partners before the grant starts, and that paperwork is not yours to draft.

What to leave out, and the one follow-up

The failure modes are consistent across all three variants, and so is the follow-up rule. None of this is about politeness; it is about not spending a stranger's attention on things that do not help them decide.

Leave out your biography. Two lines of context is the maximum before the reader needs to know why you are writing to them.; Leave out attachments in a first message. Offer them instead, which also gives you a natural reason to write again.; Leave out "I would love to collaborate" with no gap and no ask. There is nothing in that sentence to reply to.
Leave out promises you do not control: authorship position, funding, a place for their student. Describe them as things you would propose.; Leave out the same message sent to several people in one group, and never put multiple recipients on one email.; Follow up once after ten to fourteen days, in the same thread, adding one new thing and offering an easy no. After that, stop and write to the next candidate.

Frequently asked questions

Four elements: one specific sentence about a named paper of theirs that proves you read it, the gap in your work that they would fill, what you bring to the exchange, and a small ask, normally a 25 minute call. Keep it under 250 words, do not attach anything, and do not raise authorship or budgets before there is anything to allocate.