// Recruitment Technology, Business Development
Spec Sending Is a Billing Engine Run on Memory and Outlook
Published: 02 October 2026,
9 min to read
The bottom line
Spec sends drive a meaningful share of billings on most desks, yet almost nobody tracks them. Run as a repeatable campaign, with scheduled follow-up and recorded outcomes, candidate-led marketing is the cheapest revenue most agencies have. The agencies that systematize it book far more client meetings from the same candidate pool.
Candidate marketing is the revenue line nobody owns
Ask a recruitment director which activity returns the most per hour across their desks. Spec sending comes up inside the first minute. Ask where it is recorded and the room goes quiet. On most desks a spec send lives in two places: the recruiter’s memory and the sent items folder. Neither of those is a system.
Candidate marketing is the one revenue activity that agencies still run on recall. A strong candidate gets pitched to a handful of clients late on a Thursday. One client replies. The rest are never chased. The campaign ends without anyone deciding to end it, and no record exists, so nobody can say what it earned.
That matters more now than it did five years ago. Clients brief more slowly and brief harder roles. A warm introduction to a known quantity outweighs another CV against a live requisition.
Why does spec sending collapse the moment the desk gets busy?
Spec sending collapses under pressure because it is the only revenue activity on the desk with nobody chasing it. A live role has a client asking for an update. It has a candidate asking where they stand. A spec campaign has neither, so it loses every scheduling contest against work somebody is waiting on.
The result is a recruitment process that exists in principle and vanishes in practice. Atlas research found that 22.22% of agencies have defined processes that are followed inconsistently. Spec sending is usually the first thing to drop out of the consistent column. Applicant tracking systems record the live role meticulously. They record the speculative send nowhere, which tells your team how much the activity is valued.
Data decay compounds the problem. A speculative CV is only as good as the information behind it, and availability moves fast. Six weeks after a conversation, a passive candidate’s salary expectation has shifted. They may already have accepted something else. Agencies without real-time updates on role changes end up marketing people who are no longer marketable. Clients learn to skim past the next send.
What does candidate marketing look like when it runs as a campaign?
It looks like outreach with a defined audience, a scheduled sequence of touches, a named owner, and a recorded result. The shift is from sending a CV and hoping to running a short campaign per marketable candidate. That campaign needs a target list of clients who hire the profile. Each client needs a written reason to care. The second contact goes in the diary before the first one sends, and a third follows if nobody answers.
The commercial case for proactive pitching is stronger than most agencies price in. ManpowerGroup surveyed more than 40,000 employers and found that 75% of employers report difficulty in filling roles. Clients who cannot fill roles from their own pipeline are rarely irritated by a credible profile. They are being handed a solved problem early. That is why clients with no immediate vacancies still take the call.
A recruitment marketing campaign worth running has four parts an ad hoc send never has:
Running that structure by hand is where it dies, so the operational question is what your system does for you. Atlas, an AI-powered recruitment platform that uses agentic AI to strip admin out of the desk, builds speculative campaigns around each candidate. It drafts the outreach in the recruiter’s own voice, drawing on past interviews, calls, notes, and emails. The send is backed by polished candidate reports assembled from the same material. Multi-touch outreach then carries the sequence across channels, so the later contacts happen whether or not the recruiter remembers.
- A target client list built from who actually hires that profile, not from who the recruiter spoke to most recently
- Targeted messaging that says why this candidate matters to this client, drawn from what the candidate said rather than what the CV lists
- Scheduled follow-up touches across email and LinkedIn, diarized at the point of the first send rather than decided later
- A recorded outcome for every client contacted, so the next campaign starts from evidence instead of instinct
How much billing is sitting in the follow-up you never sent?
More than most agencies would care to measure. Follow-up on a spec send tends to happen when the recruiter next thinks about the candidate. That is often the week somebody else places them. The first send starts a conversation. The second and third convert it, and those are the ones that reliably go missing.
Recruiters already know where the fix sits. Atlas research found that 50% of recruiters using AI agents apply them primarily to outreach and follow-up sequences. That is the largest single use case by a wide margin. The work is repetitive and rule-bound, which is exactly what agencies hand over first.
The output is measurable. Ethea Solutions rebuilt its outbound process on Atlas. The firm recorded a 500% increase in BD outreach while spending less time on it. One sequence converted a pre-IPO FinTech prospect into six figures of billings. That volume came from removing admin around the campaign rather than from adding hours to the week.
Where the volume lands matters as much as the volume itself. Tracking marketable candidates alongside live deals in one pipeline changes that. Candidate engagement becomes a stage a manager can read, rather than activity scattered across individual inboxes.
Which numbers tell you a spec campaign is actually working?
Four of them: CVs sent on spec, client replies, meetings booked, and placements traced back to a send. Read together they give you the cost per client conversation on your cheapest channel. Most agencies cannot produce that number and would act on it immediately if they could. Without measurable outcomes, a candidate marketing strategy is a preference rather than a plan.
The gap has commercial consequences in a crowded market. Industry research reports that over 65% of staffing firms name client acquisition as their biggest obstacle. Yet the activity with the shortest path to a new client relationship is the least instrumented thing most agencies do. A recruitment team buying lead lists while leaving its spec data unrecorded is paying for prospects it already owns.
Instrumenting it is straightforward once the activity is captured as it happens. Live dashboards on recruiter activity and conversion show which consultants are speccing, which accounts respond, which profiles move, and which campaigns stall. That evidence changes behavior. A desk that can see its send-to-meeting ratio starts protecting the time that produces it. It also moves candidate-led recruitment business development from a founder’s belief to a line anyone can audit.
Frequently asked questions (FAQs) on candidate marketing and spec sends
Candidate marketing is the practice of presenting a strong candidate to clients who have not briefed a live role. In agency language it is spec sending. It runs on behalf of the candidate rather than against a requisition. Done as a campaign, it opens client conversations a job-led approach never reaches.
Enough to make the candidate’s market visible. In practice that means a target list in the low double digits rather than two familiar accounts. Build the list from who hires that profile rather than who you spoke to recently. The quality of the reason you give each client matters more than the number of sends.
At least twice after the first send, with both touches scheduled when the campaign starts. Most agencies stop after one send. That is where the majority of the available revenue is lost. Spacing contacts across channels lifts reply rates without increasing the total volume of messages.
It does if the automation has nothing to work from. Recruiters consistently say they want AI to draft outreach for review rather than send it unsupervised. That model keeps human judgment at the point of contact. Drafts assembled from real interview and call history read as specific because they are.
The consultant who owns the candidate relationship, with the campaign visible to their team lead. Full-desk recruiters deprioritize speccing under delivery pressure. The fix is making the activity appear in the same pipeline review as live roles. Visibility usually works better than a mandate.
Yes, and the economics are often better. Rebate exposure is lower, and the same person can be marketed repeatedly between assignments. Finishers coming off contract are the strongest spec material in your talent pool, since availability is known. The requirement is knowing precisely when each contractor becomes available.
The cheapest revenue you already own
Most agencies already accept that candidate marketing works. The open question is whether to keep running it on memory, with the compounding cost of forgotten follow-up staying invisible. The alternative is to make it a tracked campaign and find out what it has been worth. The first option gets more expensive every month the database grows.
An agency that records every spec send, schedules the follow-up, reads the conversion, and feeds the result into the next campaign will find revenue it already paid to generate. Making that the default means putting the activity on a platform built to run it in the background. That is the role agentic AI plays inside Atlas: the campaign admin runs itself while consultants keep the client conversation. Worth a look if your best candidates are reaching three clients and stopping there.
