Why do agencies neglect their own pipeline?
Because the people who could build it are billing. New-business research and first-touch outreach are done in bursts between client work, and the pipeline empties whenever the agency is busy, which is exactly when it should be filling. An SDR Digital FTE on ftes.ai does the research and drafting daily for the agency itself, so the pipeline builds during busy months and the partners spend their time on the conversations rather than on the lists.
The line is drawn at the first reply: the FTE researches and drafts; the partner has the relationship.
Who is on the protected list, and why does it matter more for an agency?
Every current client, every former client, every prospect a partner is already talking to, and every contact who came through a referral or a partner relationship. Anyone at the agency can add to the list; only the owner can remove a name, with a reason. No name on the list receives automated outreach at any autonomy level. For an agency the list matters more than for most businesses, because a cold email to a current client's colleague, or to a prospect a partner is courting, damages a relationship the agency lives on.
The list is written before the first draft, with every partner in the room.
- Current and former clients, and their staff
- Prospects any partner is already in conversation with
- Referral and partner-sourced contacts
What does a first-touch draft contain?
The prospect's fit against the agency's ideal client profile with the criteria scored and evidenced; the public sources the research came from, named; a draft message in the agency's voice from the Brand Brain, referencing something specific and sourced about the prospect; and nothing inferred about the prospect's problems or budget. The partner approves, edits, or sends it back with a reason, and the reasons tune the profile and the voice. Cold first-touch to prospects outside the protected list can earn autonomy within bounds from a clean record at the second review; most agencies keep it gated because the partner's edit is part of the craft.
A draft that guesses at a prospect's situation is sent back; the fact gate applies to claims about the prospect as it does to claims about the agency.
How are inbound enquiries qualified?
Against the agency's own criteria, budget range as stated, services requested, timeline, company size, fit with the agency's sector focus, each scored with the evidence from the enquiry and from public sources, and routed to the right partner with a summary. Enquiries the criteria capture poorly are flagged rather than scored. The partner challenges a criterion with a reason rather than overriding the verdict, which is how the criteria improve. The FTE never replies to an enquiry beyond an approved acknowledgement template, and never quotes a fee or a scope.
The effect is that partners take the calls that fit and answer the rest promptly with a drafted, approved decline.
Where is the hard line?
At the first reply. Every response from a prospect goes to the owning partner with the thread attached; the FTE does not continue the conversation, does not propose a fee, does not describe a scope, and does not book a call except into a connected calendar the partner controls. It never contacts anyone on the protected list, never deletes a CRM record, and never edits a partner's notes. CRM hygiene, next steps, stage corrections, duplicate merges, is proposed with the activity behind it and confirmed by the partner in seconds.
The agency's pipeline stays the partners' pipeline; the FTE builds the top of it.
What does an agency measure, and when?
Qualified conversations started per month against the baseline; partner hours recovered from research and CRM work; time from an inbound enquiry to the right partner's first touch; the share of opportunities with a current next step; and the approval rate of first-touch drafts with send-back reasons, which shows whether the voice has transferred. Measure after a full agency sales cycle, which is usually a quarter or more, not a month. Outreach sent is not a measure.
By the end of the first cycle the agency knows whether its pipeline built during a busy month for the first time.