# Usage: curl -sSL https://seed.show/sales.pipeline.crm | bash -s <install-path>
# <install-path> is the directory where the file should land.

set -euo pipefail
[ -z "${1:-}" ] && {
  echo "install requires a path: curl -sSL https://seed.show/sales.pipeline.crm | bash -s <install-path>" >&2
  exit 1
}
TARGET="$1"
mkdir -p "$TARGET"
DEST="$TARGET/seed-fold.Lhs4f6.folded.md"

cat > "$DEST" <<'PORTDOWN_9F36F4F7'
<!--
This is a `.folded.md` archive — a directory packed into one markdown
file. The markers below are load-bearing; don't edit them directly.

To unpack (requires bash — if you have no shell, tell the user):
  1. curl -sSLf https://fold.dom.vin/skill | bash -s <INSTALL_DIR>
  2. <INSTALL_DIR>/fold/scripts/unfold <this-file>
     (or: unfold <this-file>  if fold/scripts is on your PATH)
-->

---
fold: true
marker: 6e1d87
at: 2026-05-07T16:16:24Z
root: seed-pack.b3V2bL
---

<!--fold:6e1d87@file path="README.md" mode="644"-->
# sales.pipeline.crm

Sales pipeline and CRM hygiene context for agents. What to know before touching a pipeline, running a forecast, or logging an activity.

## What a pipeline actually is

A pipeline is a **probability-weighted map of future revenue** — not a list of deals the seller feels good about. Every snapshot should reflect the real state of buyer decisions, not seller optimism. The gap between these two things is why forecast calls disappoint.

Three properties define a healthy pipeline:

**Stage integrity.** Each stage is defined by a buyer action, not a seller action. A deal advances when the buyer has done something — agreed to evaluate, introduced the economic buyer, confirmed budget exists, signed off on legal. Not when the seller has sent a proposal or had a good call. Without buyer-milestone stage definitions, stages drift optimistic and the forecast lies.

**Coverage discipline.** Pipeline coverage ratio — total pipeline value divided by the revenue target for the period — is the leading indicator of forecast attainment. The right multiple varies by sales motion and win rate (3x–5x is a common range for B2B; the actual target is calculated from historical win rate and average sales cycle). Coverage without the math behind it is a feeling, not a metric.

**Velocity completeness.** The pipeline has enough records with enough fields populated to calculate pipeline velocity: how fast deals move, where they stall, which segments produce revenue fastest. A pipeline that can't answer velocity questions is missing data, not missing deals.

## Mental model: stage is a hypothesis, not a milestone

The most important reframe — and the one agents most consistently miss.

A stage answers: "What has the buyer done or agreed to that makes us believe they are here?" A stage does **not** answer: "What has the seller done?"

| Wrong framing (seller action) | Right framing (buyer milestone) |
|---|---|
| "Discovery call completed" | "Buyer confirmed they have the problem we solve" |
| "Proposal sent" | "Buyer agreed to evaluate our solution against stated criteria" |
| "Contract sent" | "Buyer confirmed purchase intent; legal review underway" |
| "Following up" | Not a stage. This is an activity. |

When stage definitions are seller actions, deals sit in "Proposal Sent" for months while the seller follows up and the buyer has gone quiet. The pipeline looks populated; the forecast misses.

**Buyer journey, sales stage, and forecast category are three separate dimensions.** Conflating them produces unreliable forecasts.

- **Buyer journey**: where the buyer is in their decision process — Awareness, Consideration, Decision, Implementation.
- **Sales stage**: milestone within the seller's pipeline, anchored to buyer milestones — Qualified, Discovery, Evaluation, Proposal, Commit, Closed.
- **Forecast category**: seller-applied confidence bucket used in rollups — Pipeline, Best Case, Commit, Closed.

A deal at stage "Proposal" might be in forecast category "Best Case" or "Commit" depending on MEDDIC completion and Economic Buyer access. Conflating stage and forecast category produces numbers that mean nothing.

## Time-in-stage: the leading indicator agents skip

Most CRMs store `stage_start_date`. Most agents never touch it.

Time-in-stage tells you:
- Whether a deal is progressing or stalled (compare actual vs. average time for the stage and segment)
- Whether stage definitions are enforced (if every deal sits 90 days in "Evaluation," the definition isn't working)
- Which deals need attention in pipeline review (long tenure in early stages rarely recovers; new arrivals at late stages need immediate care)
- Where sales cycles are elongating — a forecast problem that time-in-stage surfaces weeks before the miss

A pipeline review that doesn't surface time-in-stage outliers is not a real review.

## What agents get wrong

**1. Confusing activity with progress.**

CRM data is dense with activity: calls logged, emails sent, meetings scheduled. These are effort signals, not progress signals. A deal with 40 logged calls in "Qualification" for six months is stalled — regardless of rep sentiment. Activity volume and stage advancement are orthogonal. When reviewing pipeline, look at stage advancement and buyer-milestone evidence. Ignore activity counts.

**2. Treating stage names as universal.**

"Qualification," "Proposal," and "Negotiation" mean something specific to each company's sales motion. A MEDDIC-based enterprise process has different stage criteria than a transactional SMB motion. When working in any pipeline, read the stage definitions the company has configured — don't import generic mental models. If stage definitions aren't documented, that is the first problem to fix. All downstream analysis is wrong without them.

**3. Generating forecasts from stage alone.**

Stage × close date × amount is not a forecast. It is an optimistic projection. A real forecast incorporates MEDDIC completion (especially Economic Buyer access and Decision Criteria), time-in-stage vs. historical average, pipeline coverage against quota, and rep-level forecast accuracy history. Rolling up pipeline at face value produces noise.

**4. Conflating ACV, ARR, and TCV.**

These diverge significantly for multi-year contracts, ramp deals, and contracts with implementation fees. Using them interchangeably corrupts pipeline math. Clarify which metric the company uses for pipeline weighting and quota attainment before running any pipeline calculation. See glossary.md.

**5. Leaving stale deals open.**

A stale deal (no stage advancement or buyer engagement in >30 days) is not necessarily lost — but it needs action. A dead deal is one where the buyer has disengaged, the champion has left, or the budget has been reallocated. CRMs accumulate stale deals that inflate coverage ratios and distort velocity metrics. Pipeline hygiene means closing them. Leaving them open to protect coverage optics harms every downstream analysis.

**6. ICP vs. persona confusion.**

An ICP (Ideal Customer Profile) is a firmographic description of the accounts most likely to buy, expand, and retain — company size, industry, revenue range, tech stack, growth stage. A persona is a role-level description of the buyer or user within that account — title, responsibilities, pain points, buying criteria. These are different filters operating at different levels. Qualifying against ICP criteria tells you whether the account is worth pursuing. Qualifying against persona criteria tells you whether you're talking to the right people inside it. Confusing them produces pipelines full of wrong-company or wrong-contact records.

## MEDDIC / MEDDPICC

The dominant qualification framework in enterprise B2B. Agents should know each element and what evidence it requires — not just the acronym.

**M — Metrics.** The quantifiable business outcome the buyer needs. Not "they want better efficiency" but "they need to reduce processing time by 40% to meet their Q3 SLA." Without metrics, there is no ROI case and no urgency.

**E — Economic Buyer.** The person with authority to approve the budget. Not the champion, not the project sponsor — the person who can say yes and make the money move. If the seller has never spoken to the Economic Buyer, the deal is unqualified regardless of how many meetings have happened.

**D — Decision Criteria.** The explicit criteria the buyer will use to evaluate and select a solution — technical fit, security, integrations, pricing model, support terms. If the seller doesn't know the criteria, they're guessing at what to present.

**D — Decision Process.** The steps and timeline the buyer will follow to reach a decision: who is involved, what approvals are required, what procurement process applies. Without this, close dates are fiction.

**I — Identify Pain.** The specific business problem the buyer is motivated to solve, with urgency sufficient to act. Tied to Metrics but distinct — pain is the driver; metrics are how you measure relief.

**C — Champion.** The internal advocate who wants you to win and will sell on your behalf when you're not in the room. A champion has credibility with the Economic Buyer and a personal stake in the outcome. Distinguishing a champion from a coach (friendly, no influence) or a sponsor (supportive, passive) is a qualification judgment that cannot be automated.

**MEDDPICC adds:**

**P — Paper Process.** The legal, procurement, and contracting steps required to close. Every enterprise deal has a paper process that adds weeks to the timeline. Flag any deal where paper process is undiscovered and close date is imminent.

**C (second C) — Competition.** Who the buyer is also evaluating, what the comparative strengths and weaknesses are, and what the buyer has said about alternatives. "No competition" is almost never true in a real evaluation.

A deal is fully MEDDIC-qualified when every element has documented evidence. Partial completion is expected at early stages; gaps at late stages are risk indicators that belong in the forecast call, not hidden.

## What AI is changing

**Activity capture automation.** AI tools now log calls, emails, and meeting notes automatically into CRM fields — reducing the rep data-entry burden that is the primary source of CRM data rot. The pipeline health improvements this enables are real but lag-dependent: the CRM only reflects what was captured, and automatic capture has its own error modes (misattributed contacts, missed context, incomplete call summaries).

**Deal scoring and risk flagging.** ML-based deal scores — trained on historical win/loss patterns — surface late-stage risk that managers miss in manual review: declining engagement, champion turnover, stalled time-in-stage. These scores are probabilistic signals, not decisions. Agents that treat a high deal score as a reason to promote a deal to Commit are misusing the tool.

**Conversation intelligence.** Platforms like Gong and Chorus analyze call and meeting recordings to identify MEDDIC gap evidence, competitor mentions, pricing objections, and engagement patterns. This creates a more complete qualification record than what reps self-report. The limitation: it requires calls to be recorded and processed, and interpretation of conversational evidence still requires judgment.

**Forecast AI.** AI-assisted forecast tools (Clari, Salesforce Einstein, HubSpot AI) produce bottom-up forecast estimates that blend historical accuracy, stage data, engagement signals, and rep behavior. These outperform naive stage-based rollups — but they require sufficient historical data to calibrate, and they inherit whatever bias exists in the underlying pipeline data. A pipeline with stage inflation produces an AI forecast that still misses.

**What stays human.** Relationship judgment — whether this champion will actually go to bat, whether the Economic Buyer meeting felt like a green light or a courtesy — does not reduce to logged signals. Strategic account decisions, negotiation context, and the call about whether to walk away from a deal: these remain outside what AI forecast tools can replace. The CRM captures what happened; it does not capture what it meant.

Fetch sources.md for current platform docs and methodology guides before advising on specific tools.
<!--fold:6e1d87@file path="glossary.md" mode="644"-->
# Sales Pipeline Glossary

Precise definitions for terms that are frequently misused or conflated in AI-generated pipeline and CRM content.

---

**MQL (Marketing Qualified Lead)**
A lead that marketing has scored and passed to sales based on behavioral or demographic signals — page visits, content downloads, webinar attendance, firmographic fit. An MQL is an expression of marketing's confidence that a prospect is worth a sales conversation. It is not a sales qualification. It is an invitation to qualify.

**SQL (Sales Qualified Lead)**
A lead that a sales rep (typically an SDR or BDR) has engaged with and confirmed meets minimum criteria for entering the pipeline — typically some combination of budget existence, authority, need, and timeline. An SQL becomes an opportunity in the CRM. The MQL-to-SQL conversion rate is one of the primary handoff metrics between marketing and sales.

**SAL (Sales Accepted Lead)**
In organizations that run a three-tier qualification process, the SAL sits between MQL and SQL: marketing has passed it, and sales has accepted it for follow-up but not yet completed qualification. Used to measure whether marketing and sales agree on lead quality before full qualification effort begins. Not universal — many organizations go MQL → SQL directly.

**ICP (Ideal Customer Profile)**
A firmographic description of the accounts most likely to buy, retain, and expand — company size, industry, revenue range, tech stack, growth stage, organizational characteristics. The ICP is an account-level filter, not a contact-level filter. It answers: "Is this account worth pursuing?" Confusing ICP with persona (a role-level description) produces pipelines full of right-company/wrong-contact or wrong-company/right-contact records.

**Opportunity**
A qualified deal in the pipeline — a specific prospect, with an identified need, estimated value, and expected close date, that has cleared SQL criteria. Opportunities are the unit of work in pipeline management and the unit of measure in forecast rollups. Each opportunity has a stage, amount, close date, and owner.

**Pipeline**
The aggregate of all open opportunities within a defined time window (typically the current or next fiscal quarter). Pipeline is measured in total value, not probability-weighted value. "$2M in pipeline" means $2M in open opportunities, not $2M expected to close.

**Weighted Pipeline**
Open opportunity value multiplied by the probability of close assigned to each stage. A $100K deal at 50% probability contributes $50K to weighted pipeline. Weighted pipeline is an input to forecast modeling, not the forecast itself. Stage-level probabilities are historical averages, not deal-specific predictions. Treat weighted pipeline as a baseline, not a fact.

**Pipeline Coverage Ratio**
Total pipeline value divided by the revenue target for the period. If quota is $1M and open pipeline is $4M, coverage is 4x. Coverage is a leading indicator: too low predicts a miss; too high may indicate stage inflation or bloated pipeline with low-quality deals. The target coverage multiple is derived from historical win rate and average sales cycle — it is not a fixed number. Coverage below 2x is a structural problem, not a rep problem.

**Pipeline Velocity**
How fast revenue moves through the pipeline. Formula: (Number of Opportunities × Average Deal Size × Win Rate) ÷ Average Sales Cycle Length in days. Measured in dollars per day. Velocity is the compounding metric — it captures the four primary inputs simultaneously and makes it possible to model the revenue impact of improving any one of them. The most informative pipeline health metric that most CRM dashboards do not surface by default.

**ACV (Annual Contract Value)**
The average annual revenue from a contract, normalized to 12 months. For a 3-year contract worth $300K total, ACV is $100K/year. ACV is the standard metric for pipeline weighting and quota attainment when deal sizes and contract lengths vary. Relevant whenever contracts are multi-year, ramped, or include uneven payment schedules.

**TCV (Total Contract Value)**
The total revenue committed over the full contract term, including all years and any one-time fees. For a 3-year deal at $100K/year plus a $10K implementation fee, TCV is $310K. TCV reflects total deal economics but misleads quota conversations when contract lengths differ across deals.

**ARR (Annual Recurring Revenue)**
The annualized value of all active recurring contracts at a point in time — a snapshot of subscription revenue run rate. ARR is a company-level metric, not an opportunity-level metric. ACV is what a new deal contributes to ARR when it closes. Agents frequently conflate these: an opportunity has an ACV; the business has ARR.

**Win Rate**
The percentage of closed opportunities that were won: Won ÷ (Won + Lost). Excludes open opportunities. Measured at the stage where the final competitive decision occurs. A win rate calculated by including stale, abandoned, or disqualified opportunities is meaningless — those records should be closed out before the metric is calculated.

**Close Rate**
Sometimes used interchangeably with win rate. In some organizations, "close rate" specifically means the percentage of pipeline that closes within the forecast period regardless of win/loss — i.e., it measures pipeline-to-close conversion, not win/loss ratio. When the term appears in context, clarify which definition applies before using it in analysis.

**Stage Conversion Rate**
The percentage of opportunities that advance from one pipeline stage to the next. Conversion rates by stage reveal where deals stall or fall out of the pipeline. A 90% conversion from Discovery to Proposal with a 20% conversion from Proposal to Commit identifies the proposal stage as the primary drop-off point — the input for coaching and process intervention. Not the same as win rate.

**Forecast Category**
A seller-applied classification representing confidence level, set independently of pipeline stage. Common categories: Pipeline (possible), Best Case (likely if everything goes right), Commit (seller is confident this will close in the period), Closed (won or lost). Forecast categories roll up to produce the sales forecast. A deal's stage and its forecast category should cohere but are set independently — a rep can place a late-stage deal in "Pipeline" if the Economic Buyer hasn't been engaged.

**MEDDIC**
A B2B enterprise qualification framework: Metrics, Economic Buyer, Decision Criteria, Decision Process, Identify Pain, Champion. Each element requires documented evidence, not just a checkbox. MEDDIC is a completeness checklist, not a scoring system — partial completion at early stages is expected; gaps at late stages are forecast risk. See README.md for element definitions.

**MEDDPICC**
Extension of MEDDIC that adds Paper Process (the legal and procurement timeline required to contract) and Competition (who else the buyer is evaluating). Common in enterprise sales organizations where procurement lag is the primary source of late-stage slippage and where competitive positioning must be actively managed.

**Champion**
An internal advocate at the prospect organization who wants you to win and will actively sell on your behalf to the Economic Buyer and other stakeholders. A champion has credibility, internal influence, and a personal stake in the outcome. Distinguish from: a **coach** (someone friendly but without meaningful influence over the decision) and a **sponsor** (someone supportive but passive — will not go to bat). Deals without an identified champion are high-risk regardless of stage, amount, or time in process.

**Economic Buyer**
The person at the prospect organization with authority to approve the budget and commit to the purchase. Not the champion, not the project lead, not the procurement contact. MEDDIC qualification requires at least one direct interaction with the Economic Buyer before a deal belongs in forecast category Commit.

**Deal Score**
A ML-derived probability estimate — produced by tools like Salesforce Einstein, Clari, or Gong — that reflects the likelihood a deal will close based on historical patterns, engagement signals, and pipeline data. Deal scores are probabilistic signals, not decisions. A high deal score does not justify moving a deal to Commit; a low deal score is a prompt to investigate, not to close.

**Days Sales Outstanding (DSO)**
Average number of days between invoice issuance and payment receipt. A revenue operations and cash-flow metric rather than a pipeline metric, but relevant for evaluating contract terms and customer health. High DSO indicates collections friction or contract payment schedules misaligned with how the customer actually operates.

**Churn / Logo Churn / Revenue Churn**
Churn measures lost customers or revenue. **Logo churn** (customer churn) counts the number of customers who did not renew or cancelled. **Gross revenue churn** measures the ARR lost from non-renewals and downgrades, before accounting for expansion. **Net Revenue Retention (NRR)** subtracts gross churn from expansion revenue — NRR above 100% means the existing customer base grows even without new logos. Agents frequently conflate gross and net churn; always clarify which is being measured before interpreting retention metrics.

**Pipeline Hygiene**
The ongoing discipline of keeping pipeline data accurate and actionable: closing stale and dead deals, correcting close dates that have slipped without update, ensuring stage criteria are met before deals advance, and enforcing required field completeness. Hygiene is not administrative overhead — it is the prerequisite for every pipeline metric and forecast to mean anything. A pipeline reviewed for hygiene monthly is a different analytical object than one that hasn't been touched in a quarter.
<!--fold:6e1d87@file path="sources.md" mode="644"-->
# sources

Fetch these at task time. Ordered by relevance. Platform docs change frequently — always fetch, don't rely on training data.

## Pipeline fundamentals

1. Salesforce opportunity management — stages, tracking, forecasting setup:
   https://help.salesforce.com/s/articleView?id=sf.opportunities.htm

2. Salesforce forecast categories — how categories map to stages, rollup mechanics:
   https://help.salesforce.com/s/articleView?id=sf.forecasts3_overview.htm

3. HubSpot deal pipeline setup — stage definitions, lifecycle stages, pipeline configuration:
   https://knowledge.hubspot.com/crm-deals/set-up-and-customize-your-deal-pipelines-and-deal-stages

4. HubSpot sales analytics — forecast categories, weighted pipeline, revenue targets:
   https://knowledge.hubspot.com/sales/use-the-sales-analytics-tools

## Qualification frameworks

5. MEDDIC — framework overview, element definitions, application to discovery:
   https://www.meddic.com/meddic

6. MEDDPICC — adds Paper Process and Competition to MEDDIC:
   https://www.meddic.com/meddpicc

7. MEDDICC by Andy Whyte — the definitive practitioner guide:
   https://meddicc.com/

## Pipeline metrics

8. Pipeline velocity — formula, inputs, how to use it for capacity planning:
   https://www.saleshacker.com/pipeline-velocity/

9. Pipeline coverage ratio — target-setting, interpretation, gap analysis:
   https://www.outreach.io/blog/pipeline-coverage

10. Win rate vs. close rate — definitions, calculation, how each surfaces in pipeline review:
    https://www.salesforce.com/blog/how-to-improve-your-sales-win-rate/

## AI in CRM

11. Salesforce Einstein Activity Capture — automatic activity logging, what gets captured and what doesn't:
    https://help.salesforce.com/s/articleView?id=sf.einstein_sales_aac_overview.htm

12. HubSpot AI — deal insights, automatic enrichment, forecasting AI:
    https://www.hubspot.com/products/artificial-intelligence

13. Gong — conversation intelligence, deal risk scoring, MEDDIC gap analysis from call data:
    https://www.gong.io/product/gong-intelligence/

14. Clari — pipeline inspection, AI-based forecast management, rep accuracy tracking:
    https://www.clari.com/platform/

## Data quality and CRM hygiene

15. Salesforce data quality — duplicate management, validation rules, field completeness:
    https://help.salesforce.com/s/articleView?id=sf.data_quality.htm

16. HubSpot data quality tools — duplicate merging, property validation, data health dashboard:
    https://knowledge.hubspot.com/contacts/data-quality-overview
<!--fold:6e1d87@end-->
PORTDOWN_9F36F4F7

# ── post ──
MARKER=$(awk '/^---$/ { f++; if (f==2) exit; next } f==1 && /^marker:[[:space:]]/ { sub(/^marker:[[:space:]]+/, ""); print; exit }' "$DEST")
[ -z "$MARKER" ] && { echo "seed: archive has no marker — corrupt" >&2; exit 1; }
awk -v m="$MARKER" -v outdir="$TARGET" '
  BEGIN {
    # Match <!--fold:<m>@file path="X"--> with an optional mode attr after
    # the path (fold emits  mode="644"  on executables).
    file_re = "^<!--fold:" m "@file path=\"([^\"]+)\"( mode=\"[0-9]+\")?-->$"
    end_re  = "^<!--fold:" m "@end-->$"
  }
  $0 ~ end_re { if (current) close(current); exit }
  $0 ~ file_re {
    if (current) close(current)
    line = $0
    sub(/^<!--fold:[^@]+@file path="/, "", line); sub(/".*$/, "", line)
    current = outdir "/" line
    dir = current; sub(/\/[^\/]*$/, "", dir)
    if (dir != current) system("mkdir -p \"" dir "\"")
    printf "" > current
    next
  }
  current { print >> current }
' "$DEST"
SEED_EXTRACTED=$(find "$TARGET" -type f -not -path "$DEST" 2>/dev/null | wc -l)
if [ "$SEED_EXTRACTED" = "0" ]; then
  echo "seed: archive contained no files — refusing to delete the source" >&2
  echo "  archive preserved at: $DEST" >&2
  exit 1
fi
rm -f "$DEST"

echo "" >&2
echo "✓ seed unpacked → $TARGET ($SEED_EXTRACTED files)" >&2
find "$TARGET" -type f | sort | while IFS= read -r _sf; do
  echo "  ${_sf#${TARGET}/}" >&2
done
echo "" >&2
if [ -f "$TARGET/SKILL.md" ]; then
  echo "This seed contains a skill (SKILL.md). Install it in your agent's skills directory." >&2
  echo "" >&2
fi
echo "Install the seed skill if not already installed:" >&2
echo "  https://seed.show/skill" >&2
echo "" >&2
echo "Publisher prompt:" >&2
sed 's/^/  /' >&2 <<'__SEED_PROMPT_END_AC1F2B__'
Sales pipeline and CRM context for agents. Map of structural pipeline logic, stage discipline, MEDDIC qualification, forecast methodology, and what AI is changing — without tool-specific UI references. Fetch sources.md before advising on specific platforms.
__SEED_PROMPT_END_AC1F2B__
exit 0