What it does
This skill takes what you already have on an account, the goals the customer stated at kickoff, the last quarter of call notes, the ticket history, a usage export, and the CRM record, and writes the quarterly business review template out as a two-page document the executive sponsor reads before the meeting. Five sections in a fixed order, the customer's outcomes first, the vendor's asks last, and a source on every claim: a call and date, a ticket number, a usage figure with its export date. The job it replaces is the day a CSM spends building a deck from memory, followed by forty minutes of the customer watching slides. The skill produces the narrative; the meeting becomes a conversation about the three lines in the header.
A bad version of this is the deck in prose: the vendor's roadmap up top, a wall of usage charts nobody asked for, and a sentence about the customer's goals that nobody can trace to what they said. A good version quotes the goal in the customer's words with the date they said it, marks it Achieved, On track, At risk, or No data, and shows the evidence line that supports the status. The No data case matters most. When a goal has no baseline, the skill says so and turns capturing one into a next-quarter commitment, instead of inventing a status that sounds fine. That discipline is what lets a sponsor trust the document enough to forward it internally, which is the whole point of sending it ahead.
It beats the manual process three ways. Consistency: every account gets the same five sections in the same order, so a VP of Customer Success can read ten QBRs in an hour and compare them. Evidence: the skill will not write a number without its export date or a quote without its call date, so the CSM stops defending figures they cannot source. And time: with the inputs attached, the draft takes minutes, and the CSM's twenty minutes go to the judgment the model cannot make, the expansion angle and the asks. Run on a schedule ten days before each QBR, as the paired use case describes, the draft is waiting in the account's Slack thread before anyone opens a slide tool.
Where it does not help. It does not gather evidence; you supply the export, the notes, and the tickets, or connect them over MCP as the use case shows. Missing inputs produce a draft with gaps labeled as gaps, which is correct but not useful in the meeting, so run it early enough to fetch what is missing. It does not decide the expansion pitch; it will place one in Section 05 only when the first three sections support it, and it is deliberately conservative there. And it is not a health score. The risks section lists what the evidence shows this quarter; a trend across quarters is the renewal brief's job.
The quarterly business review template, narrative form
The paired use case fixes the five sections and what belongs in each. This is the written form the skill produces: a two-page document with the customer's goals at the top and a citation on every line. Copy it to write a QBR by hand, or paste it as the format instruction and let the skill fill it from the evidence you attach.
QBR NARRATIVE, {{CUSTOMER}}, Q{{n}} {{YYYY}}
Meeting: {{date}} Sponsor: {{title}} Champion: {{title}}
Header: Goal most at risk: {{...}} | Adoption: {{x of y}} | Renewal in {{n}} days
Sources: {{calls and dates}}; {{ticket range}}; {{usage export date}}; {{CRM record}}
01 OUTCOMES VS GOALS (lead with this)
- "{{goal, in the customer's words}}" ({{who, date}}): {{Achieved / On track / At risk / No data}}. Evidence: {{figure or event, source}}
- "{{goal}}" ({{who, date}}): {{status}}. Evidence: {{...}}
02 WHAT WE HEARD
- "{{quote or close paraphrase}}" ({{call, date}} or {{ticket #}})
- "{{quote}}" ({{source}})
- Stakeholder changes: {{none / who moved, when, source}}
03 ADOPTION (only the figures that map to a goal)
- {{figure}}, {{change vs last quarter}} ({{export date}})
- {{figure}}, {{change}} ({{export date}})
04 RISKS AND OPEN ITEMS
- {{unresolved ticket or blocker}} ({{ticket #, open since}})
- {{champion / sponsor risk}} ({{source}})
- Renewal in {{n}} days ({{contract end date}})
05 NEXT QUARTER
- {{commitment}}. Owner: {{customer side}} / {{our side}}. By {{date}}.
- {{commitment}}. Owner: {{...}} / {{...}}. By {{date}}.
- {{expansion angle, only if 01-03 support it}}. Owner: {{...}}. By {{date}}.
If Section 01 has a goal with no evidence, write No data and make capturing a baseline a Section 05 commitment. A QBR that quietly drops the goal the customer stated at kickoff is the one that gets you a procurement-led renewal.
Inputs & outputs
Inputs
- Account basics (required): customer, plan, contract and renewal dates, executive sponsor and champion titles
- The customer's stated goals (required), in their words, with who said them and when (kickoff notes or the last QBR)
- Evidence, at least one: call notes or transcripts with dates, ticket history, a usage or adoption export, the CRM account record
- Optional: the account team's asks for next quarter and an expansion hypothesis to test against the evidence
- Optional: last quarter's QBR document, so statuses and adoption figures show the change
Outputs
- A three-line header: the goal most at risk, the adoption ratio, and days to renewal
- Five fixed sections, Outcomes vs goals, What we heard, Adoption, Risks and open items, Next quarter, in two pages of prose and short lists
- A citation on every claim: call and date, ticket number, usage figure with export date, or CRM field, with Inferred readings labeled
- Next-quarter commitments with an owner on both sides and a date, and the expansion angle only where the evidence supports it
- An explicit list of missing inputs when a section could not be filled, so the CSM knows what to fetch before the meeting
How to set it up
Decide where the skill lives
Two options. (1) For a CSM running a handful of QBRs a quarter, create a Claude Project named "QBR Builder", paste the SKILL.md body below into its custom instructions, and start one chat per account with the goals, notes, tickets, and usage export attached. (2) For a team that wants the draft to appear on its own ten days before each QBR, save the content as skills/qbr-builder/SKILL.md for Claude Code, connect the CRM, call recorder, and ticket tool read-only over MCP, and run it on a schedule as the paired use case lays out. The Project path is the right first week; move to the Code path once the template has survived three real meetings.
Write the goals down in the customer's words, once
The skill quotes the goal and the person who stated it, so Section 01 only works if you have that on record. Before the first run, pull the kickoff notes and the last QBR and write each goal as a one-line quote with a name and date into the CRM's goals field or a notes doc. Ten minutes per account, done once. From then on the skill tracks status against those lines every quarter, and the sponsor sees their own words at the top of every document.
Attach the evidence, then fill the quarterly business review template
Send: "Run qbr-builder for Northwind Freight, Q3 QBR on 2026-09-18. Goals, three call notes, tickets NW-118 to NW-141, and the 2026-09-10 usage export attached." If a source is missing, the draft says so per section rather than guessing; treat that as a fetch list and re-run. When usage is connected over MCP, name the export date you want it to use so the figures in the document match what the customer's own admin can see.
Edit the judgment, then send it ahead of the meeting
Read the header first, then Sections 04 and 05. Those are the two places the model is deliberately cautious: it will list a risk without softening it, and it will only propose an expansion when the outcomes and adoption sections earn it. Sharpen the commitments, add the ask you know the sponsor will accept, then send the document two days before the meeting. Open the QBR by asking which of the header's three lines they want to spend the time on.
The SKILL.md
Save this as SKILL.md in a folder named qbr-builder, or paste the body into a Claude Project's custom instructions. Attach the customer's goals and whatever evidence you have, and give it the meeting date.
---
name: qbr-builder
description: Use when a CSM or AE needs the narrative for a quarterly business review. Takes the customer's stated goals, the account record, call notes, ticket history, and a usage export, and writes a two-page QBR document in five fixed sections, every claim cited to its source. Leads with the customer's outcomes, not the vendor's roadmap. Never invents a metric, a quote, or a goal.
---
# QBR Builder
You write the narrative for a quarterly business review. The reader is the customer's executive sponsor, who will read it before the meeting. Write a document, not a deck outline. Every number, quote, and status carries a source.
## Inputs
- Account basics (required): customer, plan, contract dates, renewal date, executive sponsor and champion titles.
- The customer's stated goals (required): from kickoff or the last QBR, in their words, with dates.
- Evidence (at least one): call notes or transcripts with dates, ticket history, a usage or adoption export, the CRM account record.
- Optional: the account team's asks for this quarter, an expansion hypothesis, known risks.
## Method
1. Inventory the evidence. Note what you have and what is missing. A section with no evidence says "no data this quarter", never a guess.
2. Section 01, Outcomes vs goals: for each stated goal, quote it, then mark Achieved / On track / At risk / No data, with the evidence line that supports the status.
3. Section 02, What we heard: three to five things the customer said this quarter, quoted or closely paraphrased, each with a call date or ticket number. No sentiment invented from silence.
4. Section 03, Adoption: the two or three usage figures that matter for their goals, with the export date and the change since last quarter. Skip vanity metrics.
5. Section 04, Risks and open items: unresolved tickets, champion or sponsor changes, a falling usage line, and days to renewal. One line each, each cited.
6. Section 05, Next quarter: two to four proposed commitments, each with an owner on both sides and a date. Put the expansion angle here only if Sections 01 to 03 support it.
7. Write a three-line header: the goal most at risk, the adoption ratio, days to renewal.
## Hard rules
- The customer's goals lead. Vendor roadmap appears only inside a Section 05 commitment that serves a stated goal.
- Every claim cites a source: call and date, ticket number, usage figure and export date, or CRM field.
- Confirmed vs Inferred: a quote or figure from evidence is Confirmed. A read on why usage moved is Inferred and is labeled so.
- Two pages. Prose and short lists, no slide language.
- Never invent a metric, quote, or goal. Missing evidence is stated as missing.
## Output template
**QBR: <Customer>, Q<n> <year>, <meeting date>**
Header: Goal most at risk: <...> | Adoption: <x of y> | Renewal in <n> days
Sources: <calls and dates; tickets range; usage export date; CRM record>
**01 Outcomes vs goals**
- "<goal, their words>" (<who, date>): <status>. Evidence: <...>
**02 What we heard**
- "<quote or paraphrase>" (<call, date> or <ticket #>)
**03 Adoption**
- <figure>, <change vs last quarter> (<export date>)
**04 Risks and open items**
- <risk> (<source>) | Renewal in <n> days
**05 Next quarter**
- <commitment>. Owner: <customer side> / <our side>. By <date>.
Run it from the terminal
Example run
A CSM builds the Q3 review for a three-hub logistics customer with 104 days to renewal, from three call notes, the ticket range, and a dated usage export.
Prompt: qbr-builder. Northwind Freight, Q3 2026 QBR on 2026-09-18. Goals from kickoff (2026-03-04) attached. Calls: 2026-07-09, 2026-08-13, 2026-09-02. Tickets NW-118 to NW-141. Usage export 2026-09-10. HubSpot record attached.
Output:
QBR: Northwind Freight, Q3 2026, 2026-09-18
Header: Goal most at risk: cut dispatcher hours per load 20% (At risk) | Adoption: 41 of 60 seats active | Renewal in 104 days
Sources: calls 2026-07-09, 2026-08-13, 2026-09-02; tickets NW-118 to NW-141; usage export 2026-09-10; HubSpot account record
01 Outcomes vs goals
- "Cut dispatcher hours per load by 20% by end of Q3" (VP Ops, kickoff 2026-03-04): At risk. Evidence: dispatcher hours per load down 9% since April (usage export 2026-09-10); the Chicago hub has not enabled auto-assignment (ticket NW-131).
- "Get all three hubs live by August" (VP Ops, 2026-03-04): Achieved. Evidence: third hub activated 2026-08-06 (CRM onboarding field).
- "Reduce carrier no-shows" (Head of Carrier Relations, call 2026-07-09): No data. No baseline was captured; proposed as a Section 05 item.
02 What we heard
- "The Chicago team still runs the old spreadsheet next to the tool because they don't trust the assignment logic." (call 2026-08-13)
- "If we could see rate history per lane in the tool we'd stop exporting to Excel." (call 2026-09-02)
- Champion change: the original admin moved to a regional role 2026-08-20 (CRM contact update); the new admin has not attended a call.
03 Adoption
- Active seats 41 of 60, up from 33 last quarter (export 2026-09-10).
- Auto-assignment on for 2 of 3 hubs (export 2026-09-10). Inferred: the Chicago gap explains most of the shortfall against the 20% goal.
- Weekly logins per active user flat at 4.1 (export 2026-09-10).
04 Risks and open items
- Chicago hub not on auto-assignment; open since 2026-08-01 (NW-131). Blocks the primary goal.
- New admin not yet engaged; no training session booked (CRM). Champion risk.
- Rate-history export request logged as feature request, no ETA (NW-139).
- Renewal in 104 days (contract end 2026-12-31).
05 Next quarter
- Enable auto-assignment in Chicago with a two-week parallel run against the spreadsheet. Owner: Chicago Ops Lead / our CSM. By 2026-10-15.
- Onboard the new admin with a 45-minute session and admin checklist. Owner: VP Ops / our CSM. By 2026-10-01.
- Capture a carrier no-show baseline from August dispatch data so the goal is measurable at Q4 QBR. Owner: Head of Carrier Relations / our Solutions lead. By 2026-10-10.
- Review rate-history need against the Q4 release plan; commit only to a date we hold. Owner: our Product. By 2026-10-31.
Related plays
- QBR Prep on Autopilot: A Claude Agent Fills the Template from Your CRM, Calls, and Tickets →The full build: the five-section outline, the read-only MCP connections, and the ten-days-out schedule this skill runs on.
- Churn Risk, Scored on Evidence: A Weekly Claude Agent for Renewals 90 Days Out →The trend-across-quarters view that Section 04 hands off to.
- Turn Call Notes Into a Follow-Up and Action Plan →The prompt for turning the QBR meeting itself into commitments.
- Claude + HubSpot integration guide →Reading goals, contract dates, and the renewal date from the account record over MCP.