How long should I wait before following up?

Ten to fourteen days, then one short follow-up that repeats the ask in a single line and adds anything new. If that produces nothing, stop. Academic inboxes are large and no reply is usually workload rather than a judgement, but a third message changes how you are remembered.

Should I propose a joint grant application in the first email?

Rarely. A grant application asks a stranger for weeks of unpaid work before any trust exists, which is why those emails go unanswered. Propose a call about one idea, and if the exchange goes well raise the application at the end of the call or in the follow-up, with a realistic account of the deadline and the effort.

What if the collaborator wants a bigger role than I expected?

Treat it as useful information arriving early rather than as a problem. Ask what they would want to lead and why, then decide whether the project you had in mind still exists inside that shape. Reshaping it now is cheap; discovering the mismatch after a year of shared work is not, and this is exactly what the pre-project discussion checklist is for.

Is it acceptable to approach several groups about the same project?

Yes, provided each message is genuinely individual and you are honest if two conversations converge. What is not acceptable is letting two groups invest effort in the same role without telling either. If you have to choose, say so early and plainly.

How do I approach a much more senior researcher?

The same way, shorter. Seniority does not change the structure of the message, it only reduces the amount of attention available, so the specificity has to be higher and the ask smaller. A precise question about one paper is often the best opening, and a referral from someone they know is worth more than any amount of polish.

How to Approach a Potential Research Collaborator

You have a name and a reason. This page covers what happens next: the size of the first ask, the call that turns interest into a plan, and the errors that quietly end most exchanges.

Finding someone is the easy half. The approach fails far more often than the search does, usually because the first message asks for too much, offers nothing specific, or arrives ten days before a deadline. This page is about the sequence from first contact to a working agreement to do something together. Where to find candidates in the first place is in how to find research collaborators, the copyable wording is in the outreach templates, and what to settle before the work starts is in the discussion checklist.

Decide the size of the ask before you write

Most rejected approaches are not rejections of you, they are refusals of the amount of work requested by a stranger. Asks come in a ladder, and the rule is to start one rung lower than feels natural.

A question: one specific technical thing only they can answer. Costs them ten minutes, opens a channel, and commits nobody.
A call: twenty five minutes to discuss one idea. This is the right first ask in most cases, and the one most likely to be accepted.
A small joint piece: a shared analysis, a comment, a workshop session, a co-supervised student. Enough to find out whether working together is pleasant, small enough that failure costs a month.; A full project or a grant application, which asks for weeks of unpaid work before any trust exists. Almost never a viable first ask, and the single most common reason a first email gets no reply.; Say which rung you are on, explicitly. An email that does not state what it wants leaves the reader to assume the largest possible version and decline it.

What the first message has to do

The wording is in the templates. The structure matters more than the phrasing, and it has four jobs, in this order.

Prove you read their work. One sentence about a specific finding, method or limitation in a named paper, saying something a keyword search could not produce.; Name the gap. What your project cannot do without them, stated in one sentence and in their vocabulary rather than yours.
Say what you bring: data, a cohort, an instrument, funding, a student, the analysis, the writing. This is the part people leave out, and its absence turns the message into a request for charity.; Make one clear, small ask with a defined cost in their time, and end there. No attachments longer than a two-page summary, and no full proposal in the body of a first email.; Be honest about deadlines up front. If a call closes in three weeks, say so in the first message rather than after they have expressed interest, because discovering it later is what actually damages the relationship.

Running the first call so it ends in a decision

A first call that ends with "let's stay in touch" has failed politely. Twenty five minutes is enough to reach a real answer if you arrive with an agenda, and sending that agenda in advance is normal rather than presumptuous.

Two minutes of context, then the idea in three sentences. Do not spend a quarter of the call on your background.; Ask what they would want out of it. Their motivation is usually different from yours and you cannot design a workable split without knowing it.
Put the awkward things on the table while the stakes are low: who would lead, roughly who would be first and last author, whose funding pays for what, and what the timeline looks like.
Agree the next concrete step and who owns it: a shared document, a data sample, a pilot analysis, a second call with two more people. A step with a name and a date is the difference between a collaboration and a nice conversation.
Send a five-line written summary within 24 hours: what you agreed, what each side does next, and by when. This one habit prevents most of the misunderstandings that appear six months later.

From interest to a real commitment

Enthusiasm on a call is cheap for both sides. What converts it into a collaboration is a small piece of shared work with a deadline, done before anyone commits to anything large.

Pick a first deliverable that takes weeks rather than months: a pilot analysis on a data subset, a joint abstract for a conference, a short methods comparison.; Use it to learn the practical things that decide whether a bigger project would work: how fast they reply, whether deadlines are met, how disagreement is handled, whether their student is supported.; Only after that, discuss the full project, the funding application or the joint grant. Ordering it this way also makes the eventual application stronger, because you can point to work already done together.; Settle roles, authorship and data before the work starts rather than after, using the discussion checklist. The conversation is easy while nothing has been produced and hard once there is a draft.; Check with your research office early if money, personal data, materials or an industry partner are involved, because those need a written agreement that is not yours to draft.

The mistakes that end exchanges early

These recur constantly and are all avoidable. Each one has the same root: treating a stranger's time as free.

"I would love to collaborate" with no gap, no ask and no specifics. There is nothing to answer.

Asking for data or an instrument while offering nothing, which is a service request dressed as a partnership.

Arriving with a deadline hidden. Ten days before a call closes, with the deadline mentioned in the third email, is the version people remember.

Sending the full proposal, the dataset and the draft in the first message. The volume itself is the refusal.

Promising authorship, position or funding you do not control. Anything you cannot deliver alone should be described as something you will propose, not as something you are offering.

Writing to a postdoc or student to get around their principal investigator. It puts them in an impossible position and it will be mentioned.

Copying the same message to several people in one group, which is visible immediately and costs you all of them at once.

When to stop, and how to leave the door open

Most approaches do not turn into collaborations, and that is the normal rate rather than a failure. What matters is that the ones that end still end well, because fields are small and the person you did not work with this year reviews your grant next year.

No reply after two weeks: send one short follow-up. No reply after that: stop. A third message converts a busy inbox into an unfavourable impression.; A polite no is a complete answer. Thank them, and ask whether they can suggest someone else, which is a request people say yes to far more often than the original one.; If the fit is wrong, say so first. Withdrawing early and clearly costs you nothing and is remembered positively.; Keep the note. A no in March and a funded call in October are different conversations, and a short message referring back to the earlier exchange starts much warmer than a new cold email.

Frequently asked questions

Four things: one specific sentence proving you read a named paper of theirs, the gap in your project that they would fill, what you bring to it, and one small ask with a defined cost in their time, normally a short call. Keep it under 200 words, avoid attachments beyond a two-page summary, and state any deadline in the first message rather than later.