# The First-Message Send Standard for Passive Candidates

*You can grade any first-touch passive-candidate draft as send, rework, or hold on relevance, length, evidence, ask, and suppression clearance.*

- Canonical URL: https://www.refolk.ai/guides/first-message-send-standard
- Pillar: Recruiting and sourcing
- Format: Standard
- Published: 2026-08-20
- Last reviewed: 2026-08-20
- Reading time: 15 min

A recruiter has drafted the first message to one specific passive candidate and now has to decide: send it, rework it, or hold it. This guide is for in-house recruiters, sourcers, talent leaders, and founders doing their own hiring who want that decision to be repeatable rather than a gut call. It turns the scattered public evidence - platform response-rate floors, length thresholds, personalization uplift, and suppression clearance - into one gradeable checklist your team can adopt as the bar a message must clear before it leaves the outbox.

Every other resource on candidate outreach is a tip list or a template pack that tells you what works on average. That is not the same as a definition of done. This is a standard: five criteria, stated so two recruiters looking at the same draft reach the same verdict, plus the checklist to verify it.

## What "ready to send" means for a first message

A first message is ready to send when it clears five criteria at once: relevance, length, specific-evidence personalization, ask clarity, and suppression clearance. A draft that fails any one of them is not a send. It is a rework or a hold.

The reason to grade all five together is that they trade against each other. Cutting to under 400 characters is easy if you delete the one sentence that makes the message specific to this person - and that trade loses more than it gains. A perfect message to a candidate on a do-not-contact list is still a violation. The standard exists to stop you optimizing one criterion into the ground while another quietly fails.

Three verdicts, defined so they are not opinions:

- **Send.** All five criteria pass. The draft leaves the outbox as written.
- **Rework.** One or more criteria fail on the copy itself - the anchor is generic, the ask is vague, the body is too long. The candidate is still contactable; the message is not ready.
- **Hold.** A gating condition fails: relevance is thin, or suppression clearance is missing or stale. Holds are not fixed by editing the copy. They are fixed by re-checking a fact or walking away.

> **Rule:** A draft passes only if it clears all five
>
> Relevance, length, specific-evidence personalization, ask clarity, and suppression clearance are AND conditions, not a score you average. Any single failure downgrades the draft to rework or hold.

## Why the standard pays off: the 13% seat-level tax

LinkedIn Recruiter enforces a 13% InMail response-rate threshold, and it is a seat-level tax, not a message-level one. That single fact is the operational reason a per-draft gate is worth the two minutes it costs.

Here is the mechanism. If a seat sends more than 100 InMails over a 14-day assessment period and the response rate sits below 13%, a banner appears on the Recruiter homepage warning the seat is at risk of the InMail Improvement Period. During a two-week warning window the rate must climb above 13%. Response rates are calculated 7 days after the warning period ends so recipients have time to respond - a warning on May 1st is re-evaluated on May 21st. If the rate has not exceeded 13% by then, bulk InMail is disabled for 14 days, extended a further 14 if it still does not improve.

**13% - LinkedIn Recruiter InMail response-rate floor**

Triggered at 100+ sends over a 14-day window; below it, a seat risks losing bulk InMail for at least two weeks.

Because the rate is cumulative across the window, one recruiter's lazy bulk sends can restrict a whole seat. Credits soften this a little - seat holders receive a credit back for every sent message that gets a response within 90 days - but a refund does not restore your rate. This is why the send/rework/hold gate pays off: it stops the low-quality drafts that drag the cumulative rate down before they ever count against you. The economics got sharper still when LinkedIn raised the price of additional InMail credits from $3 to $21 per credit in October 2025. Wasted sends now cost more and threaten the seat.

## The length criterion: how short, and what it proves

Aim for under 400 characters, and treat 800 as the outer bound. LinkedIn's own data shows the shortest InMails outperform, and it shows the effect is large enough to be a hard criterion rather than a style preference.

Defining the shortest InMails as those under 400 characters and the longest as those over 1,200, the response rate for the shortest is 22% higher than average, while the longest is 11% below. Messages up to 800 characters still beat average, and messages between 800 and 1,200 are only somewhat below. The table makes the bands and their trade-offs explicit.

| Length band | Response vs average | Share of all InMails |
|---|---|---|
| Under 400 chars | +22% | 10% |
| 401 to 800 chars | +3% to +5% | (46% are over 800) |
| 801 to 1,200 chars | -6% | - |
| Over 1,200 chars | -11% | - |

There is a subtlety worth naming, because it changes how you read the number. Under-400-character InMails get their +22% partly because only 10% of senders write them. The brevity is a scarcity signal as much as a readability one. That does not weaken the criterion for your team today, but it tells you what the criterion proves: a short message signals you edited, which reads as effort and respect for the candidate's time. What it looks like when it lies is a message that is short because it is empty - all length, no anchor. Nearly half of all InMails, 46%, run over 800 characters, so the field is long by default and short still stands out.

> Brevity wins because it is rare, not because it is inherently better, so protect the anchor when you cut.

## The personalization criterion: one anchor that cannot be reused

A message passes personalization when it carries exactly one specific-evidence anchor: a verifiable, candidate-specific detail that could not be pasted into another person's message. A merged first name is not an anchor. A referenced repo, talk, shared group, promotion, or funding event is.

The uplift data justifies the effort, but it also tells you where to spend it. Individual InMails beat bulk by about 15%. A shared group adds 21%. A candidate who already follows your company is 81% more likely to respond. And the status-based signals are bigger than any of these.

| Signal | Response uplift vs average |
|---|---|
| Sent individually vs bulk | +15% |
| Shared group member | +21% |
| Candidate follows your company | +81% |
| Open to Work | +37% |
| Recommended Matches | up to +35% |

Read the bottom two rows carefully. Open to Work candidates respond 37% more often, and Recommended Matches up to 35% more. Both exceed the 15% individual-versus-bulk gain. That means selection beats copy per unit of effort: choosing whom to message is a cheaper win than perfecting the message. The standard captures this by grading relevance and status flags in step one, before a single word of the draft exists.

The anchor test is a single question. Could this exact sentence be pasted into another profile's message without changing anything? If yes, it is not personalization; it is a template with a merge field, and the draft is a rework. Note that the dossier's sources advise referencing a promotion or a funding round but attach no specific uplift percentage to those, so treat them as legitimate anchors without claiming a number for them.

Finding the anchor is the friction the criterion creates - you have to actually read the person's public trail. This is where a plain-English search across public sources removes real work. [Refolk](/) lets you ask for the exact person and the anchor comes back with them, rather than making you assemble it profile by profile.

I ran this search: `DevOps engineers marked open to work who follow our company and have Kubernetes experience` - [see the full result list](https://www.refolk.ai/s/f360hap8p9).

*Returns candidates who already carry two of the highest-uplift signals in the table, so the anchor and the targeting arrive together.*

## Why thin markets raise the stakes on the anchor

In a thin talent market the same candidate is being contacted by many recruiters at once, so a specific-evidence anchor is not a nicety - it is the only thing that separates your message from the other five in the inbox. Refolk's index makes the concentration concrete.

**29.3x - US recruiter/sourcer supply vs UK, in Refolk's index**

21,200 matching US profiles against 723 in the UK, a derived ratio that shows how much thinner some markets are.

In Refolk's index there are 21,200 US profiles matching Technical Recruiter and Sourcer titles, against 723 in the United Kingdom.

| Market | Matching profiles | Ratio (derived) |
|---|---|---|
| United States | 21,200 | 29.3x UK |
| United Kingdom | 723 | 1.0x baseline |

The top US employers of these profiles include Experis, K2 Partnering Solutions, Snowflake, Blue Origin, and Google. The point for your standard is directional, not about recruiters specifically: where a specialist pool is small, every candidate in it hears from many sources, and a generic message is indistinguishable from noise. The thinner your market, the less optional the anchor becomes.

## The procedure: draft to send in eight steps

Follow these eight steps in order. The first two are gating checks that can send the draft straight to hold; the middle four build and sharpen the copy; the last two record the send so the seat's cumulative rate stays visible.

#### From draft to graded send

1. **Confirm targeting and relevance** - Verify the candidate matches the req on the load-bearing must-haves. Done when you can name the specific role and two profile facts that qualify them, and you have flagged any Open to Work or Recommended Matches status.
2. **Clear the suppression and do-not-contact list** - Check the candidate and their employer against opt-outs, prior-contact records, and do-not-contact company lists. Done when a dated clear is logged, or the draft is held.
3. **Draft with one specific-evidence anchor** - Reference one verifiable, candidate-specific detail such as a repo, talk, post, shared group, promotion, or funding event. Done when the anchor could not be pasted into another candidate's message.
4. **Trim to length** - Cut to under 400 characters where possible and treat 800 as the outer bound. Done when the body is under 400 characters or a documented reason for exceeding it exists.
5. **Sharpen the single ask** - State one clear, low-friction call to action. Done when the reader knows exactly what to do next in one sentence.
6. **Grade against the checklist** - Score the draft as send, rework, or hold on relevance, length, evidence, ask, and suppression. Done when the verdict is recorded and two reviewers would reach the same one.
7. **Send or schedule** - Sources disagree on cadence, so pick a team default between 1 and 6 days between active touches. Done when the message is sent or queued in the sequence.
8. **Log for response-rate tracking** - Record the send against the seat's InMail report because the 13% threshold is cumulative. Done when the send is logged and the seat's rate is monitored weekly.

On cadence, the evidence genuinely conflicts, so the standard tells you to set a default rather than pretending there is one right answer. One benchmark advises trying 1 or 6 days between stages: 1 day instills urgency and produces the highest open rates, while 6 days gives candidates time to contemplate and produces the highest interested rates. Practitioners converge on 3 to 5 touches in an active sequence before moving to nurture, where the cadence drops to once every 4 to 8 weeks for high-priority candidates and once every 2 to 3 months for broader segments.

#### The five-criteria gate

1. **Relevance** - Name the role and two qualifying facts, flag status
2. **Suppression** - Dated clear logged, or hold
3. **Anchor** - One reusable-proof detail that fits only this person
4. **Length** - Under 400 chars, 800 outer bound, anchor intact
5. **Ask** - One concrete next step in one sentence

*A draft only reaches send after clearing two gating checks and three copy checks in order.*

## How this goes wrong: the failure modes

Most bad sends pass a casual read and fail the standard. These are the seven ways a draft looks ready and is not, each with the check that catches it. This is the part of the standard that does the most work, because a criterion is only as good as the false positive it rejects.

**Fake personalization.** A template with a merged first name or company name reads as bulk. The false positive is a draft that feels custom but whose anchor fits any candidate. Check: could this exact sentence be pasted into another profile's message? If yes, rework.

**Length gaming.** Cutting to under 400 characters by deleting the specific evidence trades the +22% length gain for the lost +15% to +21% personalization gain. That is a bad trade. Check: the personalization anchor survives the trim.

**Stale suppression check.** A clear from last month is not clear now. Unsubscribes must be honored within 10 business days and do-not-call lists re-scrubbed every 31 days. Check: the suppression timestamp is inside the policy window before send.

**Colleague collision.** The LinkedIn Compose warning only fires if a colleague contacted the candidate in the same calendar month, so cross-month double-contact slips through silently. Check: search the ATS or CRM history, not just the LinkedIn banner.

**Response-rate blind spot.** A seat can pass every single-draft check yet drift under 13% cumulatively and lose bulk sending. Check: review the seat's InMail report weekly, not per message.

**Ambiguous ask.** "Let me know if interested" is not gradeable - it is a polite close mistaken for a call to action. Check: the ask names one concrete next step.

**Over-sequencing into nurture.** Pushing past 3 to 5 touches at active cadence triggers opt-outs, because excessive contact without a specific reason typically drives people to unsubscribe. Check: when the active sequence ends, switch to the 4-to-8-week nurture cadence.

> **Watch out:** The colleague warning has a one-month blind spot
>
> LinkedIn only flags a colleague's contact if it happened in the same calendar month. A teammate who messaged the candidate last month leaves no banner. Always check your own ATS or CRM history before you send.

The matrix below sorts the two variables that most often decide a verdict on a real draft: how specific the anchor is, and whether suppression is clear. It shows why a great message can still be a hold.

#### Anchor strength against suppression status

Horizontal axis runs from Generic anchor to Specific anchor. Vertical axis runs from Suppression unclear or stale to Suppression clear and dated.

| Quadrant | What it means |
| --- | --- |
| Hold and rework | Re-check suppression, then rewrite the anchor |
| Hold | Re-check suppression before touching the copy |
| Rework | Replace the generic anchor with candidate-specific evidence |
| Send | All gates passed, message leaves as written |

*Suppression is a gate; anchor strength decides send versus rework only once the gate is passed.*

## The send checklist and how to keep it current

Run this checklist against every first message before it leaves the outbox. It is written so two recruiters grading the same draft reach the same verdict, which is the whole point of a standard.

#### Before this first message sends

- [ ] I can name the specific role and two profile facts that qualify this candidate.
- [ ] I have flagged any Open to Work or Recommended Matches status, which lift response 35% to 37%.
- [ ] The candidate and their employer are cleared against opt-outs, prior contact, and do-not-contact company lists, with a timestamp inside the policy window.
- [ ] I searched ATS or CRM history for prior contact, not just the LinkedIn colleague banner.
- [ ] The message carries exactly one specific-evidence anchor that could not be pasted into another candidate's message.
- [ ] The body is under 400 characters, or under 800 with a documented reason, and the anchor survived the trim.
- [ ] The ask names one concrete next step in a single sentence.
- [ ] I recorded the verdict as send, rework, or hold, and the send is logged against the seat's InMail report.

A copy-paste rubric your team can adopt as policy, and a message skeleton that fits the length and anchor criteria at once:

**Send/rework/hold rubric**

```
RELEVANCE      pass = role named + 2 qualifying facts        [ ]
SUPPRESSION    pass = dated clear inside policy window        [ ]
ANCHOR         pass = 1 detail unusable on anyone else        [ ]
LENGTH         pass = under 400 chars, anchor intact          [ ]
ASK            pass = one concrete next step, one sentence    [ ]
VERDICT: send (all pass) / rework (copy fails) / hold (gate fails)
```

*Paste into your ATS note field or a shared doc; every criterion is pass or fail, no averaging.*

**First-message skeleton under 400 characters**

```
Hi [name] - saw [the specific thing: their talk / repo / promotion / post].
[One line connecting that to the role, why them specifically.]
Would you be open to a 15-minute call [this week] to hear what we're building?
```

*Replace the bracketed cues with a real, verifiable detail for this one candidate; delete anything generic.*

Keep the standard current by treating its numbers as mechanisms, not constants. LinkedIn can revise the 13% threshold, the length bands, or credit pricing - the credit price already moved from $3 to $21 in October 2025 - so re-read the primary Recruiter help pages when a benchmark stops matching what you see in the report. Verify suppression obligations against the official authorities rather than blog summaries: the FTC for do-not-call rules, and the ICO or EDPB for GDPR guidance, where candidates can request erasure and you must usually respond within one month, with fines up to €20M or 4% of global turnover. Review the seat's cumulative response rate weekly so the message-level gate and the seat-level tax stay connected. The checklist is the policy; the primary sources are how you keep the policy honest.

## Frequently asked questions

### What response rate do I need to keep sending InMail?

LinkedIn Recruiter enforces a 13% response-rate threshold. It is triggered when a seat sends more than 100 InMails over a 14-day assessment period with a rate under 13%. That produces a warning banner; if the rate does not climb above 13% by the end of the grace period, bulk InMail is disabled for 14 days and extended a further 14 if still unimproved. The rate is re-evaluated 7 days after the warning period ends so recipients have time to respond.

### How long should a first outreach message be?

Aim for under 400 characters. LinkedIn data shows messages under 400 characters get a 22% higher response rate than average, while messages over 1,200 characters run 11% below. Only 10% of all InMails are that short, so brevity is partly a scarcity signal. Treat 800 characters as the outer bound, since messages up to 800 still beat average, and never cut your specific-evidence anchor to hit the number.

### Is personalization worth the time at volume?

Yes, but targeting selection is the cheaper win. Sending individually beats bulk by about 15%, a shared group adds 21%, and a candidate who follows your company is 81% more likely to respond. Even bigger, Open to Work candidates respond 37% more and Recommended Matches up to 35% more. Those status-based lifts exceed the individual-versus-bulk gain, so filter for them before you polish copy.

### How often do I need to re-check my suppression list?

More often than most teams do. Under CAN-SPAM, anyone who has unsubscribed must be honored within 10 business days and suppressed from future campaigns. TCPA rules require do-not-call lists to be re-scrubbed every 31 days. Under GDPR, erasure requests must usually be answered within one month. A clearance is only valid inside those windows, so timestamp every clear and re-run it if the stamp is stale.

### Does the LinkedIn colleague-collision warning catch all double-contact?

No. The Compose Message warning only fires if a colleague on your contract contacted the candidate in the same calendar month. Contact from a previous month slips through silently. Before you send, search your ATS or CRM history for the candidate rather than trusting the LinkedIn banner alone.

---

*From the Refolk guide library. I revise these guides rather than replacing them, so the current version is always at https://www.refolk.ai/guides/first-message-send-standard*