From octave
Analyzes won/lost deals, conversation patterns, findings, and segments to refine ICP definitions, personas, and targeting criteria. Use /octave:icp-refine for win/loss insights.
npx claudepluginhub octavehq/lfgtm --plugin octaveThis skill uses the workspace's default tool permissions.
Analyze deal outcomes, conversation patterns, and qualification scores to refine your ICP definitions. Compares what your library says your ideal customer looks like against what actually wins — then recommends updates.
Discovers and defines Ideal Customer Profiles using firmographic criteria, buyer personas, scoring matrices, anti-ICP signals, and validation methodology. Useful for targeting, sales, and product strategy.
Surfaces findings, trends, and patterns from sales calls, emails, and deals including objections, pain points, questions, competitors, and value props. Filters by type, period, persona, segment, company.
Analyzes companies to identify Ideal Customer Profiles (ICPs): buyer personas, verticals, and tiers using website research, market context, and tiered framework.
Share bugs, ideas, or general feedback.
Analyze deal outcomes, conversation patterns, and qualification scores to refine your ICP definitions. Compares what your library says your ideal customer looks like against what actually wins — then recommends updates.
/octave:icp-refine [--period <days>] [--segment <name>] [--focus wins|losses|both]
/octave:icp-refine # Full ICP analysis (last 180 days)
/octave:icp-refine --period 90 # Last quarter
/octave:icp-refine --segment "Enterprise" # Specific segment
/octave:icp-refine --focus wins # Only analyze what's working
/octave:icp-refine --focus losses # Only analyze what's not working
When the user runs /octave:icp-refine:
If no options specified, use defaults and confirm:
I'll analyze your deal data to refine your ICP.
Period: Last 180 days (change with --period)
Segments: All (change with --segment)
Focus: Wins and losses
Starting analysis...
# Get current segments (this IS the ICP definition)
list_all_entities({ entityType: "segment" })
# Get full segment details
get_entity({ oId: "<segment_oId>" }) // for each segment
# Get current personas
list_all_entities({ entityType: "persona" })
get_entity({ oId: "<persona_oId>" }) // for key personas
# Get products/services (what we're selling)
list_all_entities({ entityType: "product" })
list_all_entities({ entityType: "service" })
# Get won deals
list_events({
startDate: "<period start>",
filters: {
eventTypes: ["DEAL_WON"]
}
})
# Get lost deals
list_events({
startDate: "<period start>",
filters: {
eventTypes: ["DEAL_LOST"]
}
})
# Get findings from won deals
list_findings({
query: "why we won success factors decision criteria champion",
startDate: "<period start>",
eventFilters: {
outcomeFilters: ["WON"]
}
})
# Get findings from lost deals
list_findings({
query: "why we lost objections blockers competition pricing",
startDate: "<period start>",
eventFilters: {
outcomeFilters: ["LOST"]
}
})
# Get positive conversation signals
list_findings({
query: "excited interested positive resonated value",
startDate: "<period start>",
eventFilters: {
sentiments: ["POSITIVE"]
}
})
# Get negative signals
list_findings({
query: "concerned hesitant not a fit wrong timing",
startDate: "<period start>",
eventFilters: {
sentiments: ["NEGATIVE"]
}
})
For each won deal, extract:
For each lost deal, extract:
See refinement-report-template.md for the full ICP refinement report template.
# Update segment
update_entity({
entityType: "segment",
oId: "<segment_oId>",
instructions: "<specific updates based on findings>"
})
# Update persona
update_entity({
entityType: "persona",
oId: "<persona_oId>",
instructions: "<specific updates>"
})
# Update playbook value props
update_value_props({
playbookOId: "<playbook_oId>",
updates: [{ oId: "<vp_oId>", details: "<updated details>" }],
reasoning: "Updated based on ICP refinement analysis: [evidence]"
})
# Create new persona if recommended
create_entity({
entityType: "persona",
name: "<new persona name>",
instructions: "<details from deal analysis>"
})
What would you like to do next?
1. Deep dive on a specific finding
2. Analyze a specific segment or persona
3. Compare current quarter vs. previous
4. Update a specific library entity
5. Generate updated enablement materials
6. Export the full report
7. Done
list_all_entities - Segments, personas, productsget_entity - Full entity details for ICP definitionlist_events - Won/lost dealslist_findings - Conversation insights, objections, signalsget_event_detail - Deep dive into specific dealsupdate_entity - Update segments, personasupdate_value_props - Update playbook value propscreate_entity - New personas or segmentssearch_knowledge_base - Cross-reference patternsNo Deal Data:
No deal outcomes found in the last [N] days.
ICP refinement requires win/loss data. Options:
- Extend the time period (try --period 365)
- Review conversation data instead (calls/emails without deal outcomes)
- Do a manual ICP review using your library definitions
Insufficient Data:
Found only [N] deals. Statistical patterns may not be reliable.
I'll highlight patterns but flag low-confidence findings. Consider extending the period or combining with qualitative analysis.
No Segments Defined:
No segments found in your library.
I can still analyze deal patterns, but there's nothing to compare against. Consider creating segments first:
/octave:library create segmentOr I'll suggest segment definitions based on the deal data.
/octave:wins-losses - Deeper win/loss analysis (complements ICP refinement)/octave:insights - Field intelligence trends/octave:prospector - Use refined ICP to find new prospects/octave:audit - Check library health after updates/octave:library - Manually update entities