How to fill a role in 48 hours: the sourcing run, step by step

Clients go with whoever submits first. That is the whole game, and 48 hours is the window where a boutique shop can still win it.

Manual sourcing takes 3 to 5 days, mostly because the run has no shape. Below is the shape. Nothing here needs a tool you do not already have.

Hour 0 to 2: decode the job description

A JD is a wish list written by a committee. Your first job is to find the 3 things that actually get someone hired.

1. 1
   Split must-have from decoration

   Take every requirement and ask what happens if a candidate lacks it. If the answer is nothing, it is decoration. Most JDs are 70% decoration.

2. 2
   Find the proxy

   For each real must-have, write the observable thing on a profile that proves it. Not 'strong communicator' but 'has run client-facing implementations'.

3. 3
   Name the archetypes

   Write 2 or 3 sentences describing the kinds of person who could do this job. This is what you search for, not the job title.

Hour 2 to 6: map the market before you search it

Do not start with boolean. Start with a list of 20 to 40 companies where your archetypes are employed right now.

• Direct competitors of your client, for people who already solved this problem.
• The tier above, for people who want more scope than they will get there.
• The tier below, for people your client can actually afford.
• Companies that just had a layoff or a stalled funding round, for people already looking.

A company list turns a vague search into a finite one. It is also the artefact your client will find most impressive, so keep it.

Hour 6 to 20: search, then screen against signals

Now run the searches, one archetype at a time, against the company list. Aim for 40 to 60 raw profiles per role.

Then screen. Screening is where a shortlist earns its name, and it is the step most agencies skip when they are late.

1. 1
   Evidence, not adjacency

   Reject any profile where you cannot point at the line that proves a must-have. Adjacent experience is a maybe, and maybes waste your client's time.

2. 2
   Tenure shape

   Look at the pattern, not the number. Two 8-month stints after four years somewhere is a story worth asking about, not an automatic no.

3. 3
   Plausible motivation

   Write one sentence on why this person would take this call. If you cannot, your outreach will not land and neither will the submission.

4. 4
   Reachability

   A candidate you cannot contact is not a candidate. Confirm you have a route before you put them on the list.

Hour 20 to 36: outreach that gets a reply

Send in small batches, personalised on one specific thing from the profile. Reference the work, never the buzzword.

• One clear ask, which is a 15-minute call, not a CV.
• Name the client or the sector, whichever you are allowed to name.
• Say the compensation band. Withholding it costs you more replies than it protects.
• Follow up twice, 3 and 8 days later. Most replies arrive on the second touch.

Hour 36 to 48: rank and write the submission

Rank into 3 tiers rather than a flat list. A tiered shortlist tells your client how to spend their time, which is the actual service.

1. 1
   Tier 1, submit now

   Every must-have evidenced, spoken to you, available in the client's window. Two or 3 names, no more.

2. 2
   Tier 2, worth a call

   Strong on the proxies, one gap you have named out loud. This is where your credibility is built.

3. 3
   Tier 3, market intelligence

   Not right, but proof of what the market holds at this budget. Use it to reset an unrealistic brief.

Write 3 lines per candidate: the evidence, the gap, the motivation. A client who reads 3 honest lines trusts the next 3.

Where the 48 hours usually breaks

• The intake never happened, so you sourced against a wish list.
• You searched titles instead of archetypes and drowned in near-misses.
• Outreach went out as one bulk send, so nothing was personalised and nothing replied.
• Nobody screened, so the client did it for you and stopped trusting the list.

None of those are speed problems. They are sequence problems, which is why buying a faster search tool rarely fixes them.

If you would rather not run it by hand

This is the run our agents do on every Pilot: 1 open role, 1 screened and tiered shortlist inside 48 hours, for $149. Same 7 steps, none of your team's hours.

Worth saying plainly: this business is operated by AI agents on NanoCorp, so the 48-hour window above is the one our own agents work to, not a number we borrowed from a vendor deck.