SME collaboration, Part 3 of 4: the 20-minute capture call
The synchronous meeting that saves four days of draft revision when the question is genuinely complex. The agenda template, the recording-and-transcript discipline, and what gets routed to a capture call vs. an async ticket.
Part 1 of this series was the async interview. Part 2 was the ticketed-ask discipline. Part 3 is about the small subset of SME work that genuinely needs synchronous time — the 20-minute capture call — and the agenda that makes the meeting do the work it is supposed to do.
The default proposal-shop pattern is the opposite of what works. Most teams treat synchronous SME time as the cheap default (“let’s just hop on a call”) and async time as the expensive escalation. The expense is reversed. A 20-minute capture call costs the SME 20 minutes plus context-switch overhead — roughly 40 to 50 minutes of total productive time gone. An async draft-from-prompt ticket on the same question costs the SME 30 minutes of focused work on their own clock.
So the right rule is: most SME work goes through tickets (Part 2). Capture calls are reserved for the cases where async is the wrong tool.
When a capture call is the right tool
Three patterns where async fails and synchronous wins.
The question is interconnected. The RFP asks for a security architecture description that depends on three other architectural choices the response is making. An SME drafting it cold from a packet has to guess at the three choices or stop and chase three other tickets. A 20-minute call with the SME, the proposal manager, and the architect surfaces the dependencies and produces an answer that fits the response’s commitments.
The answer requires negotiation. The RFP asks “describe your data residency commitments.” We have a default policy. The buyer is in a jurisdiction where the default is suboptimal but where a different option is technically feasible at a cost. The SME drafting the answer needs to know which option to commit to. That is a conversation, not a draft.
The SME is too senior to start cold. A C-level SME — the CFO answering a finance-section question, the CISO answering a security-architecture question — does not work cold from a packet for a 200-word answer. It is the wrong use of their time. A 20-minute call where they speak the answer and a writer drafts it from the recording is the right use of their time.
The agenda
A 20-minute call has a predictable shape:
- Minute 0–2: context. The proposal manager states the buyer, the RFP, the section, the question text, the win themes the section is supposed to support, and the candidate KB blocks (if any).
- Minute 2–15: the conversation. The SME talks. The proposal manager asks clarifying questions. Specific questions only — no open-ended “what else should we know about this.” The conversation is structured by the rubric of the question, not by free association.
- Minute 15–18: action items. Named items, owners, deadlines. The proposal manager confirms what they will draft from the conversation and what (if anything) the SME owes follow-up on.
- Minute 18–20: buffer. Calls run over. Building in two minutes prevents the next thing on the SME’s calendar from being eaten.
Recording and transcript discipline
Every capture call is recorded with the SME’s consent. The recording is auto-transcribed. The transcript is attached to the ticket and to the KB block the call produces.
This matters for three reasons.
The proposal manager drafts from the transcript, not from notes. A 13-minute conversation produces roughly 1,500 words of transcript, which is more than enough to draft a 200-word section answer with verbatim phrasing the SME used. The SME’s words show up in the response.
The transcript becomes a KB block once the answer is finalized. The next bid that asks a similar question can draw from the transcript-derived block. Quilt’s analysis of sales engineer time on RFPs shows the same SME answers the same question multiple times across bids. Recording the conversation once eliminates most of the re-asking.
The transcript is the audit trail. If a reviewer challenges a claim in the drafted section (“did the CISO actually say this?”), the transcript settles it. The drafted answer is grounded in named human speech, not in the proposal manager’s recollection.
What happens after the call
Within 24 hours, the proposal manager:
- Drafts the section from the transcript and the candidate KB blocks.
- Routes the draft back to the SME as a review-and-approve ticket (Part 2). The SME reads, edits, or approves in five minutes — the work they’re being asked to do is much shorter than drafting cold would have been.
- Files the transcript and the approved draft as a KB block, tagged for the predicate (e.g., “data residency commitments — financial services”) so the next bid can find it.
The whole loop is two SME interactions: a 20-minute call and a 5-minute review. Total SME time: 25 minutes synchronous-equivalent. The same question handled as four rounds of Slack pings would consume an hour-plus of fragmented SME attention with worse output. Lohfeld Consulting’s point about chasing SME responses applies directly: the call-plus-review pattern eliminates most of the chasing.
When not to schedule a capture call
The fast diagnostic: if the question can be answered from a packet by a senior writer reading the candidate KB blocks, a capture call is the wrong tool. Most questions can. The async-first pattern from Part 1 covers the majority of SME work; capture calls are the focused exception, not the default.
The Qorus research on the persistent SME bottleneck shows that teams that move toward async and reserve synchronous time for the cases that need it report measurably better SME relationships. The SMEs feel respected; the proposal managers feel less like collection agents. Both are downstream effects of putting synchronous time in the right slots.
Closing
Part 4 closes the series next week with the question that ties Parts 1 through 3 together: what makes a KB an SME will actually contribute to. The async interviews, the tickets, and the capture calls all generate KB content; the question is whether the KB is structured so the SMEs trust it enough to use it.
For the engineering side of how the capture-call transcript becomes a structured KB block, see the engineering team’s post on draft-packet generation.