swarm hub / onboarding / experts
Expert User Guide
For John, Mark, and Ewing. How to be a great input source for Charlie and Bear without doing their thinking for them.
What this is
Charlie and Bear are non-technical interns building features on top of our system using Claude Code. They own the Meetings page and the Deals page respectively. Their first builds ship in week 2. The bigger goal: take both pages from working tools to a B2B CRM automation engine that lets each of you cover 5 times as many deals at the same quality.
Each of them gets one 30 minute Zoom with each of you. Six sessions total. These sessions are product input, not training. The interns are the product team. You are the user.
This page tells you what to expect, what to bring, and how feedback flows back to them.
The mission
5x deal coverage at the same quality. That is the bar. Charlie owns the front of the funnel (calls become signals, signals become classified meetings, classified meetings become deals). Bear owns the back (warm leads become pipeline, pipeline becomes revenue). Their job is to remove your bottleneck. Not by being faster themselves. By building features that compound.
Charlie and Bear
Both are non-technical 20-somethings using Claude Code daily. They are smart, new to M&A, and trained on a foundation that already exists in the system: feature registry, permissions, circuit breakers, agent tickets, the handoff contract.
They've already read:
- The shared briefing (history, methodology, the 5-phase loop, failure modes)
- The 8-layer architecture explainer
- Their own package (Charlie's first AI build, Bear's first AI build, plus prep packs for each of you)
- The mastery scorecard (the live "what makes a real application" checklist)
So they know the system. They are coming to you for the parts the system cannot teach them: how an M&A advisor actually thinks.
Universal expert rules
These apply to all three of you in every session.
Your role: expert user, not arbiter
You are not grading them. You are not deciding what they build. You are showing them how the work feels from your seat. The features they ship are theirs. Your input is one signal among many.
What we want from you
- Show your screen. If they ask "how do you triage your morning calls," do not describe it. Open the page. Talk through it. Let them see the cursor move.
- Tell stories. "A deal died last quarter because..." beats "we should track stalled deals." The story has the diagnosis baked in.
- Be specific about pain. "The financial fields are spread across three places" is useful. "It could be better" is not.
- Tell them what they will not need to ask you twice. If you find yourself saying "this is the third time I've explained this," that is a feature opportunity.
What we do not want
Don't do their thinking for them. If they ask "what should I build," your answer is "what are you considering?" Then react to their proposal. The interns lose if you tell them what to build. They win if you tell them what hurts.
Don't agree to be nice. If a feature would not change your day, say so. "I would not use this" is the most valuable sentence you can give them.
Don't grade them. Do not say "good job" or "nice work." Say "I would use this if..." or "this would not survive a Tuesday because...." That is the feedback that builds operators.
Don't agree to be nice. If a feature would not change your day, say so. "I would not use this" is the most valuable sentence you can give them.
Don't grade them. Do not say "good job" or "nice work." Say "I would use this if..." or "this would not survive a Tuesday because...." That is the feedback that builds operators.
Time discipline
30 minutes. They have prep packs. They will state a goal in the first 60 seconds. If the goal is unclear, say so. If they have not done their homework, say so and end early. Their preparedness is your signal of whether they are ready to ship to you.
Recording
Fireflies records every Zoom you join. The transcript appears in os.chapter.guide/meetings within an hour. The interns extract themes, feature ideas, and "things you wanted but did not say" from the transcript automatically. You do not need to send them notes. Just show up and talk.
What Charlie and Bear know about you
- You built the calling discipline at Next Chapter. You have made more cold calls than the other two combined.
- Your background includes leading the global field org at Findly through major acquisitions and spinoffs. M&A is not theoretical to you.
- You are known for "earning trust of CEOs through clear communication and decisive action." You are the patient mentor in the room.
- You are a core part of the Next Chapter team.
Your sessions
Charlie : John
The calling discipline. Cold open mastery. What makes a great call vs a wasted one.
Bear : John
Pipeline discipline across years. What does a healthy pipeline look like and when do deals start to rot?
What Charlie will ask you
She is preparing 10 to 12 questions tuned to your strengths. Highlights:
- "When you look at a call list first thing in the morning, what makes a name jump to the top?"
- "What does a great cold call sound like in the first 10 seconds?"
- "What's the difference between a good 'no' and a bad 'no'?"
- "How do you read voicemails? Same script every time, or tuned to the contact?"
- "What's the worst classification mistake a caller can make and what did it cost?"
- "Tell me about a call that turned into a closed deal. What signal did you catch?"
- "Where would you never trust software to make the classification call?"
Full bank in Charlie's prep pack.
What Bear will ask you
- "What does an amazing deals page look like to you? Walk me through ours and give rapid feedback."
- "How do you decide which deal gets your attention this morning?"
- "What deal stage is most dangerous, where deals get stuck or die?"
- "Tell me about a deal that died because of bad data, bad timing, or a bad handoff."
- "Tell me about a deal that closed because something worked perfectly."
Full bank in Bear's prep pack.
What we want from you specifically
Stories. Lots of them. You have the longest view of the three. The single most valuable thing you can give Charlie is the story of a call that closed and the story of a call that should have but didn't. The single most valuable thing you can give Bear is the story of a deal that died on the vine and what the early signal was that everyone missed. Specific. Named. With the year you remember it happening.
Show your call list. If Charlie asks how you triage, pull up Salesfinity on screen share. Talk through your filter logic. The screen recording is the curriculum.
Show your call list. If Charlie asks how you triage, pull up Salesfinity on screen share. Talk through your filter logic. The screen recording is the curriculum.
What to avoid in your sessions
Don't lecture on calling theory. They have read the cold-call-workflow skill. They know the framework. They want the texture, not the principles.
Don't soften your standards. If their proposed classification taxonomy is wrong, tell them. Charlie's day-one classifications will be off. The fastest way to make her good is to tell her what she missed on the third call back, not the tenth.
Don't soften your standards. If their proposed classification taxonomy is wrong, tell them. Charlie's day-one classifications will be off. The fastest way to make her good is to tell her what she missed on the third call back, not the tenth.
How feedback flows back to Charlie and Bear
- FeedbackBox on every feature page. One sentence. They read it the next morning.
- The Loom transcript auto-extracts your themes, feature ideas, and "things you wanted but didn't say." Do not duplicate that work.
- Weekly 15 minute sync once they have something to show. You will get a calendar invite.
Mark DeChant
Operator across all three entities. The grinder. The sales leader's sales leader.
What Charlie and Bear know about you
- You are a veteran sales leader. Reps under you have earned 250+ combined President's Clubs over 14 years.
- You founded a sales talent recruiting firm where every recruiter is a former President's Club winner. You know what good looks like at the rep level.
- You operate across all of Next Chapter. You see more deal flow per day than anyone else.
- You are the most operational of the three. You will probably share screen and demo within the first 5 minutes.
Your sessions
Charlie : Mark
The operational reality of dialing. What slows you down moment to moment.
Bear : Mark
The daily grind of touching deals. What busywork to kill first.
What Charlie will ask you
- "Walk me through 7am, 10am, 3pm. What does the meetings page need to do at each point?"
- "When you classify a meeting, what's going through your head? What signals are you weighing?"
- "What's the worst kind of busywork you do every day around classification?"
- "What forces you to switch context away from the meetings page?"
- "If I built one feature for you in the next two weeks, what would it be?"
- "Would you rather have 800 leads at $0.01 each total cost where 250 are wasted, or 131 leads at $0.30 each where every one answers their cell phone with company fit 80%+ verified in advance?"
What Bear will ask you
- "When financials change on a deal, what does that signal to you?"
- "What deal stage is the most dangerous, where deals get stuck or die?"
- "When do you decide a deal is dead? What's the signal?"
- "What do you wish the deals page did that it doesn't?"
- "How do you remember what a deal needs next? Is it on the page, in your head, in your inbox?"
- "What's the worst kind of busywork you do every day around deals?"
What we want from you specifically
Show your screen for 20 of the 30 minutes. You are the only one who lives in the pages all day. The interns will learn more from watching you than from your descriptions. Pull up the meetings page (for Charlie's session) or the deals page (for Bear's). Talk while you click.
Name the busywork. Every "ugh, I have to do this again" moment is a feature ticket. Mark them out loud. Charlie and Bear will catch them in the transcript.
Tell them what you wish you had. The 5x deal coverage mission only works if we eliminate the operational drag. You are the one who feels the drag. Speak it out loud.
Name the busywork. Every "ugh, I have to do this again" moment is a feature ticket. Mark them out loud. Charlie and Bear will catch them in the transcript.
Tell them what you wish you had. The 5x deal coverage mission only works if we eliminate the operational drag. You are the one who feels the drag. Speak it out loud.
What to avoid in your sessions
Don't try to teach them sales fundamentals. Save that for John. Your slot is operational reality. The page, the click, the keystroke, the missing field.
Don't say "yes" to everything they propose. Your reps know you are honest. Apply that here. If a feature idea wouldn't change your Tuesday, say so.
Don't say "yes" to everything they propose. Your reps know you are honest. Apply that here. If a feature idea wouldn't change your Tuesday, say so.
How feedback flows back to Charlie and Bear
Same channels as John's section. One bonus channel for you specifically: when they ship a feature, use it for one full day in your normal workflow. Then drop a 3-sentence review in the FeedbackBox: what you used, what you didn't, what was missing. That is the most valuable feedback in the system.
Ewing Gillaspy
Managing Partner. Next Chapter. The visionary and system designer.
What Charlie and Bear know about you
- You designed this entire system. The feature registry, the handoff contract, the 5x mission, the mastery scorecard.
- You bring a deep GTM and sales-tech background: Topia, IBM, SmartRecruiters, Symphony Talent. You have spent years thinking about how AI changes a sales motion.
- The session has a different gravity than the John or Mark session because you designed the system they are building on. They may be more careful about what they say.
- You own Next Chapter. You designed the entire system the interns are building on.
Your sessions
Charlie : Ewing
The vision for the meetings page. The 5x lead qualification target.
Bear : Ewing
The vision for the deals page. The 5x deal coverage target.
What Charlie and Bear will both ask you
- "If you had to cover 5x as many deals next year at the same quality, what would the page have to do that it doesn't today?"
- "How should I prioritize which leads (or deals) get worked on in which order? Is there a measurement system that should auto-prioritize, with manual override for situational focus?"
- "How should I think about getting the most out of Clay's capabilities for this project? What about Exa or Hermes?"
- "Would you rather have 800 leads at $0.01 each total cost where 250 are wasted, or 131 leads at $0.30 each where every one answers their cell phone with company fit 80%+ verified in advance?"
- "What's a feature you've wanted on this page for months but haven't built?"
- "Where is human judgment irreducible? What should never be automated?"
- "How do you want me to learn from John and Mark without bothering them every day?"
- "What does mastery on this page look like in 3 months?"
What we want from you specifically
Listen more than you talk. You designed the system. Their job in this session is to challenge what you assumed. If you spend 25 of 30 minutes explaining the vision, you will learn nothing about whether the vision is correct.
Ask for the friction they have already hit. "What surprised you when you actually used the system?" beats "let me tell you what I want next." They have been in the codebase for hours. They know things you do not.
Lower the stakes at minute zero. Tell them: "I am not grading you. I want to learn what you discovered." Otherwise they will hold back the most valuable observations.
Ask for the friction they have already hit. "What surprised you when you actually used the system?" beats "let me tell you what I want next." They have been in the codebase for hours. They know things you do not.
Lower the stakes at minute zero. Tell them: "I am not grading you. I want to learn what you discovered." Otherwise they will hold back the most valuable observations.
What to avoid in your sessions
Don't fix it on the call. If they tell you a thing is broken, your instinct will be to debug it live. Resist. Note it. Their assignment is to plan the fix and bring it back. If you fix it, they learn nothing.
Don't load them up with strategy. Your session is naturally the most strategic. They cannot turn 30 minutes of vision into 7 days of work. Pick the one thing you want them to wrestle with this week and let the rest wait.
Don't ask if they understand. Ask "what would you change about my mental model based on what you have seen?" That is a real question. "Do you get it?" is a closed door.
Don't load them up with strategy. Your session is naturally the most strategic. They cannot turn 30 minutes of vision into 7 days of work. Pick the one thing you want them to wrestle with this week and let the rest wait.
Don't ask if they understand. Ask "what would you change about my mental model based on what you have seen?" That is a real question. "Do you get it?" is a closed door.
How feedback flows back to Charlie and Bear
You see everything they ship before anyone else does. The activity dashboard at /admin/interns is yours. The mastery scorecard at /swarm/onboarding/mastery.html is yours. The review queue at /review/tasks is yours. Your feedback is the canonical signal for whether they are calibrated.
Bonus channel just for you: when the stuck detector fires (looping, idle, spinning, repeated rejection), you get a Slack ping. Reply to it. Specifically. They will take their cue from how you respond.
Where feedback goes
Five channels. They flow into the same loop.
| Channel | When you use it | What the interns see |
|---|---|---|
| FeedbackBox (every feature page) | You used the feature. You have one or two sentences. | Read the next morning. Anchored to the specific feature. |
| Fireflies transcript (every Zoom) | Automatic. You don't trigger this. | Themes, feature ideas, and unstated wants are auto-extracted. They write up within 24 hours. |
| Weekly sync (15 min, scheduled) | They have something to show. You watch them click. You react. | Live demo, live notes. Their fastest learning channel. |
Review queue (/review/tasks) | They created a task that needs operator/admin approval. | You approve or reject with a reason. Approval is a signal of trust. |
| Slack DM (last resort) | The system is broken in a way the channels above can't catch. | Direct, immediate, unstructured. Use sparingly. Most things belong in FeedbackBox. |
The trap: sending feedback through Slack DMs becomes the default because it is the easiest. It is also the worst for the loop. The next person to look at the feature has no idea what you said. Spend the extra 30 seconds and put it in the FeedbackBox.
Anti-patterns to avoid
Things experts do that hurt their interns.
- "Let me build that for you real quick." Tempting because it is faster. Wrong because it removes the learning. Their job is to figure out how to ship it. Yours is to use it once they do.
- "You should also add..." Scope creep dressed as helpfulness. Pick one feature at a time. They will ask for the next one when they are ready.
- "This is great!" Empty praise. Replace with "I would use this if X" or "this would not survive a Tuesday because Y." Specific. Anchored.
- Approving everything in the review queue. If you approve every task, the queue is not doing its job. Reject with reasons sometimes. The reasons are the curriculum.
- Going silent. If you go a week without using a feature they shipped, the loop is broken. Either the feature is bad (tell them) or you forgot (your fault, fix it).
- Telling them what to build. They are interns building features for you. They are not order-takers. Tell them what hurts. Let them propose what to fix.
Quick reference
Pages you can visit
os.chapter.guide/admin/interns | Daily activity per actor. Charlie's classifications, Bear's pipeline reviews. |
os.chapter.guide/swarm/onboarding/mastery.html | Live "what makes a real application" scorecard. |
os.chapter.guide/review/tasks | Review queue. Approve or reject intern proposals. |
os.chapter.guide/meetings | Where your Fireflies transcripts land within an hour of the call. |
What to do this week
- Read this page once.
- Open the prep pack the intern wrote for you (Charlie's or Bear's page, scroll to your section).
- Block 30 minutes on your calendar for each session. Two sessions per expert. Six total across both interns.
- Show up. Talk. Show your screen. Tell stories.
- Use whatever they ship within 24 hours. Drop a sentence in the FeedbackBox.
If something feels off
Slack Ewing directly. The interns are new and the system is new. Some things will not work the first week. The fastest path to fixing them is making sure Ewing knows.