How the four functions run as one motion in this kit. why-align.md is the argument — why marketing, sales, product, and retention belong in one system. This is the spine — the handoffs, the one number, and the shared definitions that make it run. The two are meant to be read together.
The figures below are directional and attributed where used. Treat the direction as well-supported and the magnitudes as soft. Sources are in
why-align.md.
Draw the customer journey honestly and it isn't a funnel that ends at the sale — it's a bowtie. The left half is what most teams instrument: demand and acquisition. The right half is the one the funnel forgets: onboarding, adoption, renewal, expansion. Two of the three ways a company grows happen after the first sale (Winning by Design). That right half is where net revenue retention lives — and a leaking bucket can't be filled by pouring in more acquisition.
So the seams a buyer feels aren't only marketing↔sales. They're sales↔the product received, and product↔whether anyone helps the customer get value. Four hands on one motion, not four stages of a relay.
| Function | Owns | On the bowtie |
|---|---|---|
| Marketing / Demand | Awareness, demand, the buyer's first surfaces | Left — top |
| Sales | Qualification, the deal, the close | Left — bottom |
| Product | What the customer actually receives and adopts | Center — the knot |
| Retention | Onboarding, adoption, renewal, expansion | Right |
Run as four separate systems, each optimizes its own metric at the others' expense. Run as one, the signal compounds: what sales hears shapes marketing; what customers do shapes product; what product ships re-arms sales.
A motion is only as strong as its handoffs. Each one has a trigger, an owner, and an artifact — the place it's written down so it can't drop.
| # | Handoff | Trigger | Owner | Artifact |
|---|---|---|---|---|
| H1 | Marketing → Sales | A lead clears the MQL→SQL bar | Marketing | A qualified lead in ops/pipeline.md |
| H2 | Sales → Marketing | Sales hears a recurring objection, competitor, missing proof, or win/loss reason | Sales | A MARKETING-ACTION line in the feedback log (marketing-feedback) |
| H3 | Won → Onboarding | A deal is marked won | Sales → customer success (CS) | The first ops/customers.md row, with the value hypothesis carried forward (onboarding-handoff) |
| H4 | CS → Product | Adoption friction, a churn reason, or a request blocking adoption/expansion | CS | A routed item in ops/roadmap-signals.md (product-signal) |
| H5 | Product → GTM | A capability ships | Product | The one-line "what this means for the buyer" in ops/roadmap-signals.md, handed to sales and marketing (product-signal) |
| H6 | CS → Marketing | A customer reaches a clear outcome | CS | A reference / proof point — the customer's own words for the value they got |
| L | Loop close | Weekly review | All four | MARKETING-ACTION cleared, RETENTION-RISK routed, references captured — read in /weekly-review |
H1 and H2 are the loop most teams can see. H3–H6 are the right side of the bowtie — the half usually left with no owner. H3 is the seam most teams never define: a signed deal that drops into a CS void with no shared definition of ready. H5 is the antidote to "build it and they will come" — Pendo (2019) found ~80% of features rarely or never used across its own accounts (vendor data), often because they ship with no enablement and no buyer-facing reason to adopt them.
- One revenue number: net revenue retention (NRR). The whole bowtie rolls up to it. Median NRR sits around ~106% (ChartMogul / SaaS benchmarks); below 100% means acquisition spend is filling a leaking bucket. It's the financial number that now decides what a company is worth — and the reason the right side of the bowtie can't be an afterthought.
- A North Star that captures customer value — the leading indicator of NRR, specific to your product (the action that means a customer is getting value, counted across the base). NRR is the lagging financial truth; the North Star is the thing you can move this week.
One shared number beats four local metrics — lead volume, bookings, ship count, tickets closed — that conflict the first time a target slips.
Written once, owned jointly, living as one file each — not four arguments in four tools.
- One ICP. Who you're for, agreed by sales and marketing in the same room. (
ops/icp.mdif you keep one —lead-qualifyreads it.) - One MQL→SQL bar. The single definition of a qualified lead both sides own — the H1 trigger. (Documented SLAs correlate with ~34% higher odds of higher YoY ROI — HubSpot State of Inbound 2015, self-reported.)
- One Won→onboarding definition. What "ready to hand off" means — the H3 trigger and checklist, so a closed deal doesn't drop silently. (
onboarding-handoffruns it.)
You don't need a platform — a plain-text repo run from Claude Code is the operating model:
- One source of truth → the git repo. Every function reads and writes the same markdown.
- The handoffs →
ops/pipeline.md(H1), the feedback log (H2, H6),ops/customers.md(H3),ops/roadmap-signals.md(H4, H5). - The skills →
marketing-feedback(H2),onboarding-handoff(H3),account-health(the right-side detector) handing tochurn-save,expansion-signal, andqbr-prep(the act-skills),support-signal→product-signal(H4/H5),retention-feedback(the post-sale→product loop). - The rituals →
/demo-briefingand/weekly-reviewrun the joint motion and close the loop (L) across all four;/retention-reportrolls the book up to NRR/GRR monthly.
The mature version of this is RevOps — one operating model and one data spine for the whole motion. At a founder-led stage the fix is to close the loop, not to hire a RevOps team: keep all four functions learning in one place and document the motion before you delegate it.
why-align.md— the one-page argument, with sources, for why the four belong in one system.methodology.md— Claude Code as the operating environment the model runs in.connecting-a-crm.md— optional: make an existing CRM the system of record and project it intoops/pipeline.md(the H1 surface), instead of running two pipelines.