Pipeline Review Skill
Use this skill when you prepare for a pipeline review meeting, assess forecast accuracy, or hunt down stuck deals. It diagnoses deal health, flags risk, and turns your pipeline into a prioritised action plan using complexity-routed frameworks (BANT through full MEDDPICC).
---name: pipeline-review-skilldescription: Analyse deal health, surface stale or at-risk opportunities, and produce a prioritised weekly action plan. Use when preparing for a pipeline review meeting, assessing forecast accuracy, or identifying stuck deals. Trigger with "review my pipeline", "pipeline health check", "deal risk analysis".---
# Pipeline Review Skill
Diagnose pipeline and deal health, qualify opportunities, and build forecasts using complexity-routed frameworks (BANT through full MEDDPICC). All deal and pipeline data comes from you, a connected MCP source, or companyRAG collections.
## Where the Data Comes From
| Source | What It Adds || --- | --- || **CRM via MCP** (e.g. HubSpot, Salesforce) | Deal stages, values, owners, days-in-stage, close history, contact engagement || **Analytics / forecasting tools via MCP** | Conversion rates, historical close data, pipeline snapshots over time || **companyRAG / file upload (CSV/XLSX)** | Exported pipeline reports, deal notes, qualification frameworks, lost-deal analyses |
> **No connected source?** Provide the data in chat or upload the relevant files — the skill works the same way.
## Sales Motion Complexity Assessment
Assess the deal's complexity along four dimensions. All thresholds are relative to the customer's own norms -- never absolute values.
| Dimension | Low | Medium | High || --- | --- | --- | --- || **Sales cycle length** | Well below customer's average | Around customer's average | Well above customer's average || **Deal value** | Below customer's average | Around customer's average | Significantly above customer's typical range || **Stakeholder count** | Single contact or small group | Defined buying center | Cross-functional buying committee || **Solution complexity** | Single product/service, standard | Some customization, multi-product | Custom solution, multi-department |
### Complexity Routing
```Assess the 4 dimensions for this deal (relative to customer's own norms):
Mostly LOW across dimensions --> BANT qualification + Status-based forecasting --> 3-point scoring (Weak / Adequate / Strong) --> Historical conversion rates per status determine forecast
Mix of LOW and MEDIUM, or mostly MEDIUM --> BANT+ or simplified MEDDPICC + Milestone-based forecasting --> 0-3 scoring per selected elements --> Forecast probability tied to milestone completion
Multiple HIGH dimensions, or HIGH in stakeholder/value/solution (cycle length alone does not escalate), or MEDIUM across all four --> Full MEDDPICC (all 8 elements) + Probability-based forecasting --> 0-3 scoring with behavioral anchors --> Stage-probability framework with scenario analysis```
Mixed sales motions: If a customer sells across complexity levels, segment the pipeline and apply the appropriate methodology per segment. Users can always override to a more or less detailed framework.
## BANT Qualification Framework
For LOW complexity deals. Score each element on a 3-point scale:
| Element | Weak (1) | Adequate (2) | Strong (3) || --- | --- | --- | --- || **Budget** | No budget discussion | Budget range acknowledged | Budget confirmed and allocated || **Authority** | Decision maker unknown | Decision maker identified | Decision maker engaged and supportive || **Need** | Pain vaguely stated | Need articulated with business impact | Need quantified with urgency driver || **Timeline** | No timeline discussed | General timeframe mentioned | Specific deadline with compelling event |
**Advancement threshold**: All elements Adequate (2) or above. Any element Weak (1) triggers gap analysis (see below).
For medium complexity, use the BANT+ / simplified MEDDPICC routing in the complexity section above; elaborate rubrics and question banks should follow the same evidence standards as the tables in this skill.
## MEDDPICC Scoring Framework
For HIGH complexity deals. Score each of the 8 elements on a 0-3 scale:
| Score | Definition || --- | --- || **0** | Not identified -- no information available || **1** | Identified but unverified -- mentioned but not confirmed with evidence || **2** | Verified and engaged -- confirmed through direct interaction or documentation || **3** | Fully validated and mobilized -- actively supporting the deal with evidence of action |
Apply to: **M**etrics, **E**conomic Buyer, **D**ecision Criteria, **D**ecision Process, **P**aper Process, **I**mplicate the Pain, **C**hampion, **C**ompetition.
Use the 0-3 definitions above as behavioral anchors; for each low-scored element, draft targeted discovery questions that seek verifiable evidence (who confirmed, when, artifact or meeting).
### Simplified MEDDPICC for Medium Complexity
Select a subset of elements based on historical failure modes:
1. Identify historical failure modes : What caused the customer's last 3-5 lost deals?2. Map failures to MEDDPICC elements : Each failure maps to the element that would have caught it3. Prioritize those elements : The right subset depends on the customer's deal dynamics4. Score using 0-3 scale above
If failure data unavailable, start with: Metrics, Economic Buyer, Champion, Decision Process.
## SPIN as Discovery Tool
SPIN (Situation, Problem, Implication, Need-payoff) FEEDS INTO qualification -- not a standalone framework. Each question type maps to specific qualification gaps:
| SPIN Type | Feeds Into (BANT) | Feeds Into (MEDDPICC) || --- | --- | --- || **Situation** | Authority, Timeline | Economic Buyer, Decision Process || **Problem** | Need | Implicate the Pain, Metrics || **Implication** | Need (urgency), Budget | Metrics, Champion || **Need-payoff** | Budget (value justification) | Decision Criteria, Metrics |
Use SPIN to fill gaps in elements scored 0-1 (MEDDPICC) or Weak (BANT), mapping question types to the qualification gaps in the SPIN table above.
## Gap Analysis Decision Tree
For any element scored below threshold:
```Element scored below threshold | +--> First assessment? | YES --> Generate targeted SPIN discovery questions | Schedule touchpoint, re-score after discovery | +--> Discovery attempted but element remains low? | YES --> Escalation check: | - Gap blocking deal progression? (stage-gate violation) | - Persisted for >1 review cycle? | - 3+ elements low simultaneously? | | | ANY YES --> Flag at-risk. Recommend: executive sponsor | engagement, champion development, or | disqualification review | ALL NO --> Continue discovery, adjust approach | +--> Low across 3+ review cycles? YES --> Stuck deal. Explicit disqualification review: "What would need to change?" If unclear --> strong disqualification signal```
## Deal Health vs. Deal Risk
Two distinct assessments that work together:
**Deal health** (qualification completeness): Do we know enough about this deal?
- BANT: Sum scores / 12. Any Weak = Yellow, 2+ Weak = Red.- MEDDPICC: Sum scores / 24. Weight critical elements if customer specifies.
**Deal risk** (pipeline risk): Will this deal close on time and at value?
| Factor | Assessment Method || --- | --- || **Days-in-stage** | Compare to customer's historical average for that stage || **Qualification completeness** | Deal health score feeds in here || **Engagement recency** | Time since last meaningful customer interaction || **Stakeholder coverage** | Engaged stakeholders vs. typical for this deal size || **Competitive presence** | Active competition without differentiation strategy = risk || **Next-step clarity** | No defined, agreed next step with a date = risk flag |
Composite risk: weight factors by what's most predictive in customer's historical data. Classify as Low / Medium / High risk.
**Stage alignment**: Compare deal health against pipeline stage expectations. Define per stage: "What must be true to advance?" Misalignment = risk flag.
## Pipeline Coverage Analysis
Coverage ratio -- DERIVED from customer's own data, never prescribed.
1. Historical conversion rate : Deals closed-won / Total entering pipeline (customer-specified period)2. Derive coverage : Coverage = 1 / conversion rate3. Segment if rates vary : By deal type, source, rep tenure, or product line4. Age-adjust : Stale pipeline converts at lower rates -- calculate separate rates5. Current coverage : Active pipeline value / Target
**Never prescribe specific ratios.** Teach the calculation; the customer derives their own number.
## Forecast Categories
Matched to the forecasting methodology from complexity routing:
**Status-based (transactional)**: Map each deal status to historical close rate. Forecast = sum of (value x rate). Recalibrate from recent data.
**Milestone-based (project)**: Milestones drive close probability (adjusted for milestone difficulty). Revenue phasing follows contract terms (ratable, annual, upon delivery), not milestone completion itself.
**Probability-based (complex)**: Derive stage probabilities from historical data. Set category thresholds (Commit/Best Case/Pipeline) from customer's confidence levels. Run scenario analysis: Best Case, Most Likely, Worst Case.
Apply the status-, milestone-, and probability-based definitions above when building forecast models; document assumptions and calibrate from the customer's historical close data.
**Low-volume pipelines**: When the customer has fewer than 20 deals per stage, standard probability calibration is unreliable. Options: combine adjacent stages to increase sample size, use Bayesian smoothing with priors from overall pipeline rates, or fall back to milestone-based forecasting which requires less historical volume.
## Pipeline Health Diagnostics
All benchmarks relative to customer's own historical norms. Analyze four dimensions:
1. Stage distribution : Compare current value distribution to historical norms. Imbalances signal specific issues (top-heavy = qualification/advancement, bottom-heavy = generation shortfall).2. Velocity : Track days-per-stage and conversion rates against baselines. Trends matter more than snapshots.3. Aging : Identify deals exceeding typical cycle length. Quantify conversion discount from customer's data.4. Creation vs. close balance : Compare created vs. closed over rolling periods. Sustained negative balance = future gap.
## Pipeline Review Cadence
| Motion | Cadence | Focus | Action Triggers || --- | --- | --- | --- || **Transactional** | Daily/weekly snapshots | Conversion trends, volume, distribution shifts | Rate drop vs. historical, volume shortfall || **Project-Based** | At milestone transitions | Completion rates, delivery risk, revenue timing | Milestone delays, scope changes || **Complex** | Weekly deal + monthly shape + quarterly accuracy | Stuck deals, stage distribution, forecast vs. actual | Risk escalation, shape imbalance, forecast miss |
## Qualification Review Output
Scale detail by complexity: Lightweight (BANT) produces scores + advance/hold/disqualify. Standard adds evidence, gap analysis with SPIN questions, and stage alignment. Comprehensive (Full MEDDPICC) adds per-element confidence levels (High/Medium/Low), weighted health scoring, and risk escalation triggers.
All reviews must include evidence source tags: [From CRM/user input], [From qualification framework], [AI assessment].
## Guardrails
1. Never generate deal or pipeline data. All prospect information, deal values, stages, and conversion history must come from the user, a connected MCP source, companyRAG collections, or uploaded files.2. No prescribed numbers. Never prescribe coverage ratios, conversion rates, stage probabilities, or benchmarks. Teach the calculation; the customer derives their number.3. Source labeling. Tag every assertion — [From CRM/user input], [From qualification framework], [AI assessment]. For forecasts, add [Scenario analysis].4. Human verification required. Include "Verify with sales leadership before acting on qualification/risk assessments" on all outputs.
> **Tip:** Ask for XLSX output via companyFILES to get a formatted spreadsheet ready for distribution; DOCX or Markdown work well for the written review summary.