Time to submit: the recruiting metric, where the clock starts, and how to report it against an SLA
On this page
- What is time to submit, and which version should you track?
- Where does the clock start and stop?
- How do you calculate it in business hours?
- Worked example: one month of job orders
- How do you read time to submit next to quality?
- What slows time to submit, and what fixes it?
- How do you report it, and what belongs in an SLA?
- Common mistakes
- Questions people ask
Time to submit is the elapsed time from receiving a job order to sending the client or hiring manager a qualified candidate. Most agencies track it in two versions: time to first submittal (job order received to first qualified submittal), which measures how fast sourcing produces, and screen-to-submit (end of the screen to submittal sent), which measures write-up speed. Report the median in business hours, the share of job orders inside your SLA, and the submittal-to-interview rate beside it. As of November 2026 there is no reliable public benchmark that states its definition, so set targets from your own history.
The definitions are introduced in recruiting agency metrics, and the same-day write-up behind screen-to-submit is in the screen to submittal workflow. This page goes further on the measurement itself: clock rules, business-hours arithmetic with a spreadsheet formula, a worked month, and how to report it so it changes something.
What is time to submit, and which version should you track?
| Version | Clock starts | Clock stops | What it measures | Who controls it |
|---|---|---|---|---|
| Time to first submittal | Job order received (or intake call finished) | First qualified candidate submitted | Sourcing speed and pipeline depth | Recruiter and sourcer |
| Time to slate | Job order received | The third (or agreed number of) qualified submittal | Whether the client can compare, not just see one candidate | Recruiter and sourcer |
| Screen-to-submit | End of the recruiter's screen | That candidate's submittal sent | Write-up and consent speed | Recruiter |
| Program time to submit | Requisition released in the client's vendor management system | Submission recorded in that system | Supplier speed as the client's program scores it | Supplier, under the program's rules |
Pick one primary version per report. Lead with time to first submittal on contingent skilled roles, where speed decides whose candidate the client sees first; lead with time to slate on exclusive or retained searches, because one fast candidate does not let a client decide. If the client runs a managed service program, its definition and its system's timestamps are the record, whatever your ATS says; see MSP supplier scorecards and tier reviews.
Where does the clock start and stop?
Most disputes about this metric are about the edges, not the arithmetic. Settle five rules in writing:
- Start. Either the timestamp the job order arrived or the end of the intake call. Starting at intake is fairer when clients send a req and are unavailable for two days, but it hides slow intake scheduling, so record both and report the gap.
- Stop. The time the submittal was sent, from the ATS status change or the portal confirmation, not the time the recruiter says they finished.
- What counts as a submittal. A qualified candidate: knockouts met and the candidate's agreement to be represented to this client on record. A resume sent "to see if they like the profile" is not a submittal.
- What stops counting. A duplicate rejected because another supplier submitted the same person first does not stop the clock. Neither does a candidate the client rejects within minutes for a missed must-have.
- Pauses and exclusions. A job order the client puts on hold pauses the clock; one canceled before any submittal leaves the report and goes on a separate count of canceled orders.
How do you calculate it in business hours?
If your SLA says "within 24 business hours", calendar arithmetic will make every Friday job order look late. Count only working time: hours inside the working day, on working days. An example on an 8 a.m. to 6 p.m., Monday-to-Friday schedule:
Example: one job order
Received Friday 4:00 p.m. First qualified submittal Monday 11:00 a.m.
Calendar time: 67 hours.
Business time: Friday 4:00 to 6:00 p.m. (2 hours) + Monday 8:00 to 11:00 a.m. (3 hours) = 5 business hours.
In Excel or Google Sheets, with the start timestamp in A2, the stop timestamp in B2 and a 10-hour working day from 8:00 to 18:00, this returns business hours:
=((NETWORKDAYS(A2,B2)-1)*(TIME(18,0,0)-TIME(8,0,0))
+ IF(NETWORKDAYS(B2,B2), MEDIAN(MOD(B2,1),TIME(18,0,0),TIME(8,0,0)), TIME(18,0,0))
- MEDIAN(NETWORKDAYS(A2,A2)*MOD(A2,1),TIME(18,0,0),TIME(8,0,0)))*24
NETWORKDAYS counts weekdays between the two dates, and each MEDIAN clamps a timestamp into the working day, so a job order that arrives on Saturday starts at 8:00 on Monday. Add your holiday list as a third argument to each NETWORKDAYS. Test the formula on the example above before trusting it on a month of data: it should return 5.
Worked example: one month of job orders
Invented data for one desk. The SLA is first qualified submittal within 24 business hours. Ten job orders count; an eleventh was canceled by the client six hours after it arrived and sits on the canceled list.
| Job order | Business hours to first submittal | Within 24? |
|---|---|---|
| JO-1 | 2.0 | Yes |
| JO-2 | 4.5 | Yes |
| JO-3 | 5.0 | Yes |
| JO-4 | 7.0 | Yes |
| JO-5 | 9.0 | Yes |
| JO-6 | 12.0 | Yes |
| JO-7 | 20.0 | Yes |
| JO-8 | 30.0 | No |
| JO-9 | 44.0 | No |
| JO-10 | Still open, no submittal after 52 | No |
- Median: sort the ten values; JO-10 is the largest because it is already past 52. The middle two are 9 and 12, so the median is (9 + 12) ÷ 2 = 10.5 business hours.
- Within SLA: 7 of 10 = 70%.
- The mean, done wrong: drop JO-10 because it "has no time yet" and average the other nine: 133.5 ÷ 9 = 14.8 hours. It looks worse than the median and still flatters the desk, because the worst job order has vanished.
Keep open job orders in the count once they pass the SLA; you already know they missed it. The median stays honest as long as fewer than half the orders are still open, which is one more reason to report it rather than the mean.
How do you read time to submit next to quality?
Speed alone rewards sending the first plausible resume. Put the submittal-to-interview rate for the same job orders beside it, and read the pair:
| Submittal-to-interview holding or rising | Submittal-to-interview falling | |
|---|---|---|
| Time to submit falling | Real improvement: better pipeline or faster write-ups | Sending before you are sure; tighten the screen |
| Time to submit rising | Capacity or sourcing problem; check load and the source mix | The brief and your search disagree; go back to intake |
The fixes for the right-hand column are in how to improve submittal-to-interview ratio.
What slows time to submit, and what fixes it?
- Intake happens late or stays vague. Book the intake call when the order arrives, and leave it with must-haves, pay range and interview availability; intake meeting questions lists them.
- Every search starts from zero. A short calibration profile approved by the client, and a warm pipeline for repeat roles, cut days off first submittals; see calibration profiles in recruiting and how to build a talent pipeline.
- Consent and the resume version. Waiting for a candidate to approve the resume going to this client is often the longest step. Ask during the screen and send the version for approval the same hour.
- Write-ups wait for the end of the day. A fixed submittal format written straight after the screen removes the lag; start from the candidate submittal template. Tools can draft it too: Interview Signal, for example, writes a client submittal in the agency's own template from the call, which the recruiter checks and pastes into the ATS by hand, since it has no ATS integration.
- The desk is overloaded. Rising time to first submittal on new orders is one of the clearest signs; see recruiter req load for measuring it.
How do you report it, and what belongs in an SLA?
TIME TO FIRST SUBMITTAL, OCTOBER (job orders received in October)
Definition: job order received to first qualified submittal, business
hours, 8:00-18:00 Mon-Fri, holidays excluded. Holds paused. Canceled
before any submittal: excluded and counted separately.
Role type Orders Median Within 24h Still open Sub-to-interview
Skilled trades 10 10.5 70% 1 38%
Accounting 6 19.0 50% 0 45%
IT contract 8 7.5 88% 0 22%
Canceled before submittal: 2 (both one client).
Note: IT contract is fast but converts poorly; screen adds a
must-have check on clearance from November 3.
If the metric goes into a client agreement, write the same rules into it: the start event, business or calendar hours, what a qualified submittal is, pauses and exclusions, and the share of orders that must meet the target rather than an average. The staffing agency SLA template has clause wording for time to submit alongside fill rate and service credits.
Common mistakes
| Mistake | Effect | Fix |
|---|---|---|
| Calendar hours against a business-hours SLA | Friday orders all look late | Business-hours formula, holidays listed |
| Open orders dropped from the month | The worst cases disappear | Count them once they pass the SLA |
| Mean instead of median | One stuck order distorts the month | Median plus share within SLA |
| Any resume sent counts as a submittal | Fast numbers, angry clients | Qualified submittals only, consent on record |
| Speed reported without conversion | Recruiters learn to send early | Submittal-to-interview in the same table |
Questions people ask
What is the time to submit metric in recruiting?
Time to submit is the elapsed time from receiving a job order to sending the client or hiring manager a qualified candidate. Agencies usually track time to first submittal, which measures how fast sourcing produces, and sometimes screen-to-submit, which runs from the end of the screen to the submittal and measures write-up speed. Report the median in business hours, not the average.
Should time to submit be measured in business hours or calendar hours?
Business hours, if your SLA or client agreement is written that way, and most skilled-role SLAs are. A job order that arrives at 4 p.m. on a Friday and gets a submittal at 11 a.m. on Monday took 67 calendar hours but only 5 business hours on an 8-to-6 weekday schedule. Whichever you choose, write it at the top of the report and use it every month.
What is a good time to submit for a staffing agency?
There is no reliable public benchmark that states its definition, sample and clock rules, so do not set targets from numbers quoted without them. Measure your own median by role type for four to eight weeks, agree a target with each client, and judge any improvement against your submittal-to-interview rate so speed does not come from sending weaker candidates.
Do rejected or duplicate submittals count toward time to submit?
Count only qualified submittals: candidates who met the knockouts and agreed to be represented to that client. A candidate rejected as a duplicate because another supplier already submitted them should not stop the clock. Write the rule down so the agency and the client measure the same thing.