How to Research a Candidate Before Reaching Out

Kome Team · Workflow

Researching a candidate before reaching out does not require a twenty-minute deep dive. Spend five focused minutes across three places, their profile, one thing they actually wrote or built, and any recent news about their current company, and you have enough for one specific, true sentence in your first message. That one sentence is the difference between a message a candidate reads and one they scroll past. Everything beyond it is diminishing returns: more open tabs, not more signal.

If you send forty outreach messages a week, you already know the math does not support a deep background check on every candidate. What it supports is a fast, repeatable pass that gets you one good detail, every time, without eating your whole morning.

What "good enough" candidate research actually looks like

Recruiters who are good at outreach are not doing more research than everyone else. They are doing less research, better aimed. The goal of this pass is not to build a full profile of the candidate before you have even spoken to them. It is to find one true, specific thing you can mention that proves you read past their job title.

A generic message says "I came across your profile and was impressed by your background." A researched one says "the API rate-limiting approach in your talk from last quarter is close to a problem we're actually stuck on." Same amount of typing. Completely different read for the person receiving it.

The five-minute research pass before you reach out

1. Scan the profile for one real signal. Skip the job titles and skip straight to the summary, featured section, or pinned post. You are looking for what they chose to highlight about themselves, not what a résumé parser already told you.
2. Find one thing they made public. A blog post, a conference talk, a GitHub repo, a shipped feature they announced. This is almost always the strongest thing to reference because it is specific, verifiable, and something they are proud of.
3. Check their current company for recent news. A layoff round, a funding announcement, a pivot, or a product launch changes what kind of message actually lands. Someone at a company that just cut a team wants a very different first line than someone at a company that just raised.
4. Save what you found, with a one-line note on why. Not a bookmark titled with their name. A note that says what the detail was and why it mattered, so you can write the message later without re-reading the whole page.
5. Write the message from the note, not from memory. Memory compresses specifics into vague impressions within a day. The note keeps the actual detail intact.

Where to look, and what it is actually good for

Where most outreach research falls apart

The failure mode is not skipping research. It is doing research that never turns into anything. A recruiter opens a candidate's LinkedIn, their GitHub, their old company's About page, and two personal blog posts, reads all of it, closes every tab at the end of the day, and sends the same template they always send because nothing from that reading got written down anywhere.

The reading was not wasted exactly, but it produced nothing you can reuse. The fix is not reading less. It is capturing the one detail worth using the moment you find it, instead of trusting that twelve open tabs will still mean something four hours later when you finally sit down to write outreach.

Turning research into a message someone actually answers

This is where a browser extension built for exactly this earns its keep. Kome lets you bookmark a candidate's page or post the moment you find something worth using, and attach a quick note explaining why it mattered, right there on the page, without switching to a separate doc or spreadsheet. That note is searchable later, so "the API talk from Q3" still turns up a candidate's name six weeks after you first saved it.

When you sit down to actually write the message, Kome's compose feature can turn that saved page and note into a first draft of an outreach email or LinkedIn message, built from the specific thing you found rather than a blank cursor. You still edit it, you still add your own voice, but you are not starting from nothing after doing all that reading.

LinkedIn's own data backs up why this is worth the extra few minutes: personalized InMails sent one at a time outperform bulk sends by roughly 15 percent, and messages under 400 characters get about 22 percent more responses than the average message. Specific and short beats long and generic, consistently.

How to keep candidate research from disappearing between reqs

The other failure is losing research you already did. A candidate you researched for one req resurfaces for a different one three months later, and none of what you found the first time is reachable, because it lived in a tab that got closed or a note in an app you have since abandoned. Folders do not fix this, because you have to remember which folder and which req before you can find anything. A searchable note attached to the actual page a candidate's work lives on holds up much better, the same way taking notes directly on a page beats copying quotes into a doc you will never reopen.

A candidate research checklist you can actually keep to

• One profile scan for what they chose to highlight, not their job history
• One piece of their own writing, code, or talk, read for real
• One check of their current company for anything from the last 90 days
• One saved note explaining the detail you plan to use, written the moment you find it
• One message written from that note, edited in your own voice, kept under 400 characters where the channel allows it

None of this scales to a hundred candidates a day. It does not need to. It scales to the ten or fifteen candidates a week you actually want to hear back from, which is the only number that was ever going to move your pipeline.

Frequently asked questions

How long should candidate research take before I send outreach?

Five to ten minutes is enough for a single, specific first message. Scan their profile, find one thing they wrote or built, and check for recent news at their current company. Past that point you are mostly stalling, not gathering anything you will actually use in the message.

What should I look at first when researching a candidate?

Start with whatever they chose to make public and specific: a project, a talk, a blog post, a GitHub repo, a LinkedIn post with real detail in it. Their job title and years of experience are already in your ATS. Look for the thing that tells you what they care about, not what they were hired to do.

Is it strange to mention something specific I found in a message?

Not if it is something they posted publicly and you reference it accurately and briefly. Referencing a public blog post or a shipped project reads as attentive. Referencing something obscure or personal, like a comment on a friend's post, reads as unsettling. Stick to what they clearly meant for a professional audience to see.

How do I keep candidate research from getting lost between reqs?

Save the page and the reason you saved it the moment you find something worth using, rather than trusting yourself to remember it later. A searchable note attached to the page beats a browser tab or a mental note, both of which disappear the moment your laptop restarts or your attention moves to the next req.

Can an AI tool write the outreach message for me?

It can draft one from what you already found and saved, which is faster than starting from a blank template. It cannot do the research for you. The personal detail that makes a message land still has to come from a real page you actually read, not from a tool guessing at what a candidate might like to hear.