The problem: A staffing firm buys applications by the hundreds but cannot say which channels produce people who start work.
The solution: Follow candidates from source to start date and margin, so recruiting spend follows placements instead of clicks.
The math
A channel costing $1,200 a month needs about six starts worth $2,500 in gross margin each year just to cover its $14,400 annual spend.
The cheapest application is often the one that costs the recruiter the most time. It arrives without the required experience, goes unanswered after the first call, or accepts a different shift. The marketing dashboard still counts it as a lead.
For a staffing firm, an application is inventory with an uncertain chance of becoming a start. The question is where the candidates who actually fill orders come from.
Count starts, not forms
Give each candidate a first known source: a job board, referral, local campaign, prior applicant list, or direct visit. Keep the source when the person is interviewed, submitted, offered, and started. Then compare sources on the measure the firm earns from: gross margin on assignments or placements, after the direct cost of acquiring the candidate.
Do not hide the recruiter effort. A source that produces twice as many applications but takes four times as many screening calls can be more expensive than its invoice suggests. Candidate quality is not a mysterious score; it can be observed as qualification, response, start, and staying long enough for the assignment to pay.
A look at a staffing firm
Consider a $7 million staffing firm paying $1,200 each month for one candidate source. It receives a steady stream of applications and assumes the spend is working.
If that channel produces six additional starts over a year and each start yields about $2,500 of gross margin, it produces roughly $15,000 against $14,400 of spend before recruiter time. It has barely cleared the first hurdle. If a referral program with a smaller budget produces the same six starts and requires fewer screening calls, it may deserve the next dollar instead.
Those numbers are illustrative. The point is to make the firm's own cost per start and margin per start visible. A candidate source should be judged by the work it fills, not the stack of resumes it creates.
Let the weak stages identify the fix
If candidates from one source are qualified but stop responding, the problem may be slow contact. If they respond but fail screening, the message may be attracting the wrong people. If they accept but do not show, the firm should inspect the shift, pay, and confirmation process before blaming marketing.
AI can organize source data and identify where each cohort drops out. A recruiter should review the reasons before cutting a channel that serves a hard-to-fill specialty.
The four-step check, in your business
- Name every source. Use a small, stable set of source labels and keep an unknown option.
- Track each stage. Record qualified, contacted, submitted, offered, and started for each candidate.
- Price the result. Compare annual spend and recruiter hours with gross margin from starts by source.
- Test one shift. Move a modest budget amount to the better source and watch starts, not application volume.
A firm that can trace source to start can buy fewer applications and fill more of the orders that matter. That is a better marketing result than a cheaper form submission.
