Why recruitment agency founders are still doing BD at 9pm

The agency had its best billing quarter in three years. Six placements in eight weeks, two of them retained. The founder was exhausted, the desk was clear, and the pipeline was empty. So at 9pm on a Tuesday, she opened LinkedIn and started writing outreach messages.

This is not a story about work-life balance. It is a story about a workflow that structurally guarantees the same outcome every cycle.

The received wisdom in recruitment is that BD is a discipline problem. Founders who lose pipeline momentum are told to block time, hire a BDM, or commit to a daily call target. That advice is not wrong exactly. It is just aimed at the wrong diagnosis.

The problem is not that founders lack discipline. The system they are running makes consistent BD structurally impossible. The standard remedies, including the BDM hire, often make the economics worse before they make them better.

"Just hire a BDM" does not pencil out the way people think

The most common advice founders get is to hire someone to own BD. And eventually, at the right scale, that is correct. But the maths at the point most founders are considering it is uncomfortable.

A junior BDM at £30,000 base, with on-costs, desk costs, and a realistic ramp period of four to six months, costs somewhere between £45,000 and £55,000 before they generate a placed fee. In a specialist market where average fees sit at £8,000 to £12,000, that person needs to close five or six placements just to break even on their own cost. In their first year. While learning your market, your tone, your ICP, and your process.

That is not an argument against hiring. It is an argument against hiring as a substitute for having a functioning BD system. A BDM dropped into a broken workflow does not fix the workflow. They inherit it, and they hit the same sequencing problem the founder had: BD stalls when delivery peaks, because delivery always feels more urgent than prospecting.

The half-hire maths gets worse when you factor in what founders actually need at that stage. They do not need someone to generate new contacts from scratch. They need someone, or something, to keep warm contacts warm, to follow up the conversations that went quiet during a busy billing run, and to surface the right names at the right moment. That is a process problem, not a headcount problem.

Headcount solves capacity. It does not solve sequencing.

The pipeline dries up at the worst possible moment by design

Recruitment BD follows a predictable collapse pattern. It looks like this: the founder wins two or three mandates, switches into delivery mode, outreach drops to near zero for six to eight weeks, placements complete, fees land, and then the founder surfaces to find the pipeline is cold. Cue the 9pm LinkedIn session.

This is not bad behaviour. It is a rational response to incentive structures. When you are mid-process on a £15,000 retained search, that is where your attention goes. The problem is that the cost of that attention shift is invisible until the billing run ends and the silence becomes audible.

The pattern is self-reinforcing. Feast and famine cycles are not random. They are the output of a workflow where BD and delivery compete for the same resource: the founder's time and cognitive load. Every agency running on this model is, in effect, scheduling its own dry spells.

The fix people reach for is discipline: calendar blocks, daily call minimums, accountability check-ins. These work at the margin. They do not change the underlying structure, because the structure is that BD requires the founder's active attention, and delivery will always win that competition when fees are on the table.

What changes the structure is separating BD activity from founder availability. Not by hiring someone to do it instead, but by building the parts of the BD workflow that do not require real-time human judgement to run without the founder's active involvement.

What "systematic BD" actually means at agency scale

The common belief is that BD in recruitment is inherently relationship-driven and therefore cannot be systematised below a certain quality threshold. Personal outreach. Warm calls. Founder credibility on the message.

That is true for the conversion moments. It is not true for the forty things that happen before and after those moments.

bd_workflow_audit:
  founder_required:
    - first_call_with_new_contact
    - retained_brief_qualification
    - fee_negotiation
    - relationship_reactivation_after_long_gap
  system_can_run:
    - contact_enrichment_and_segmentation
    - warm_contact_follow_up_sequences
    - job_change_and_trigger_monitoring
    - inbound_lead_routing_and_triage
    - outreach_draft_generation_for_founder_review
    - pipeline_status_reporting
    - lapsed_client_flagging_at_threshold
  common_mistake: >
    Founders treat the entire BD workflow as founder-required.
    The result is that all of it stops when delivery peaks.

The jobs in the "system can run" column are not trivial. Monitoring for job change triggers, flagging lapsed clients at a defined threshold, keeping warm sequences moving on contacts who have gone quiet: these are the activities that maintain pipeline temperature during a billing run. They are also the activities that disappear first when the founder gets busy.

None of this requires a BDM. It requires a workflow design decision: which parts of BD genuinely need a human in the loop at every step, and which parts just need a human to review the output before it goes out.

Most founders have never separated those two questions. They treat BD as one undifferentiated activity, which means it is either fully on or fully off depending on how busy the desk is.

This connects to a broader point about where AI adds value in agency operations. The brief bottleneck problem in agency workflows is structurally similar: the founder becomes the bottleneck not because they are the only person who can do the work, but because no one has designed the workflow to route the non-judgement work elsewhere. The same pattern shows up in BD and in candidate communications.

What to do instead

Stop diagnosing the 9pm LinkedIn session as a discipline failure. It is a workflow output. The system produced it.

The practical starting point is a BD workflow audit: map every recurring BD activity against whether it genuinely requires the founder's real-time involvement, or whether it requires the founder's judgement at one point in the process and then can proceed. Most agencies find that sixty to seventy percent of their BD activity falls into the second category and is currently being treated as the first.

Once you have that map, the sequencing problem becomes solvable. Warm follow-up sequences run during delivery peaks. Trigger monitoring surfaces the right names at the right moment. The founder reviews and approves, but does not have to initiate from scratch at 9pm after a billing run.

The BDM hire, if and when it comes, lands into a functioning system rather than a vacuum. That changes the ramp economics significantly.

If you want to run that audit properly, the AI Workflow Audit at WPS is designed exactly for this: identifying where founder time is being consumed by work that a well-designed system could handle, and building the workflow to separate the two. It is not a product pitch for a specific tool. It is a process design engagement, because the tool choice is secondary to getting the structure right.

And if you want to understand why AI implementations in agencies tend to fail before they get to this point, why most AI pilots fail covers the pattern in detail.

The 9pm outreach session is not inevitable. It is the output of a system that was never designed to run without you. That is fixable.