[{"data":1,"prerenderedAt":228},["ShallowReactive",2],{"/en/workflows/branch-metrics-conversation-signal-triage":3},{"id":4,"slug":5,"locale":6,"translationGroupId":7,"localeSwitchApproved":8,"title":9,"description":10,"documentationMarkdown":11,"workflowJson":12,"category":209,"tags":210,"integrations":213,"difficulty":215,"author":214,"verified":33,"featured":33,"date":216,"modified":216,"icon":7,"imageSrc":7,"path":217,"alternates":218,"seo":219},"f619555b-fd18-40a3-87f5-13a710e6dc93","branch-metrics-conversation-signal-triage","en",null,true,"Branch Metrics & Conversation Signal Triage","A guided triage menu that helps operators sanity-check branch numbers, spot dirty signals, and choose when automation is safe vs when human judgment must step in.","## How it works\nThis workflow turns “we have numbers” into “we have decision‑ready evidence.” It gives operators a single menu that routes to practical guardrails for branch metrics, conversation signals, attribution, and automation.\n\nIt’s designed for the moment right before a confident meeting goes off the rails: when the dashboard looks clean, the story sounds plausible, and the mistake is about to become policy. The flow helps you pressure-test what deserves trust, what’s polished noise, and where human judgment still beats automated certainty.\n\n## Key features\n- Uses a knowledge base policy first so replies stay consistent with your internal definitions and data sources.\n- Interactive menu routes users to decision-shaped guidance (not abstract “best practices”).\n- Dedicated paths for branch number trust checks, dirty-signal detection, automation vs judgment, cross-branch comparisons, and signal culture.\n- Built-in “talk to a human” handoff for ambiguous cases and high-stakes decisions.\n\n## Step-by-step\n1. **Trigger:** A user starts the workflow.\n2. **Knowledge base policy:** The assistant answers using your approved internal guidance first (definitions, source-of-truth notes, metric caveats).\n3. **Menu:** The user chooses what they need help with:\n   - **Which branch numbers deserve trust?**\n   - **How to spot dirty signal early**\n   - **Automation vs human judgment**\n   - **Comparing branches & attribution traps**\n   - **Build a signal culture (decisions, not slides)**\n   - **Talk to a human**\n4. **Routing (IF checks):** The workflow matches the selected button and sends the corresponding guidance message.\n5. **Handoff (optional):** If the user selects **Talk to a human** (or if the selection can’t be matched), the workflow hands off to **Insights Support** with a clear context-setting message.\n\n## Setup requirements\n- **Calypso** with interactive messages enabled in your messaging channel.\n- (Optional but recommended) A maintained internal knowledge base covering branch metric definitions, attribution rules, and source-of-truth tables.\n- No additional credentials are required for this template as-is.",{"id":13,"teamId":14,"name":9,"version":15,"workflowVersion":16,"nodes":17,"connections":172,"routingEnabled":8,"active":33},"wf_branch_signal_triage_001","calypso-public-library","1.0.0",1,[18,34,41,52,83,93,101,107,113,119,125,131,137,143,149,155,161],{"id":19,"name":20,"type":21,"typeVersion":16,"position":22,"parameters":25,"category":32,"deletable":33,"connectable":33},"node_flow_configs","Workflow settings","flow-configs",[23,24],-420,-40,{"name":9,"description":26,"tags":27,"triggerType":31},"Guided triage for branch metrics and conversation signals: trust checks, dirty-signal detection, automation vs judgment, and attribution pitfalls.",[28,29,30],"signal-quality","branch-metrics","decision-systems","input","policy",false,{"id":35,"name":36,"type":31,"typeVersion":16,"position":37,"parameters":40,"category":31,"deletable":33,"connectable":8},"node_input","Start",[38,39],-260,120,{},{"id":42,"name":43,"type":44,"typeVersion":16,"position":45,"parameters":46,"category":51,"deletable":8,"connectable":8},"node_kb_policy","Knowledge base guardrails","knowledge-base-policy",[24,39],{"enabled":8,"fallbackToRouting":8,"sticky":33,"stickyMode":47,"activationOpener":48,"personalization":50},"default",{"enabled":8,"instruction":49},"Use our internal guidance on branch metrics, attribution, and signal definitions. When the user asks for decision advice, prioritize practical checks: source-of-truth, denominator sanity, lag windows, selection bias, and instrumentation changes. Keep it plainspoken and decision-shaped.",{"useContactName":8},"response",{"id":53,"name":54,"type":55,"typeVersion":16,"position":56,"parameters":58,"category":51,"deletable":8,"connectable":8},"node_menu","Choose your triage topic","interactive-message",[57,39],220,{"messageType":59,"headerText":60,"bodyText":61,"footerText":62,"sectionTitle":63,"buttons":64,"ctaDisplayText":63,"ctaUrl":63},"button","Signal triage menu","Pick what you’re deciding. I’ll help you pressure-test the signal before it turns into a confident mistake.","Tip: Clean numbers? Check what changed in tracking.","",[65,68,71,74,77,80],{"id":66,"title":67},"audit_branch_numbers","Trust branch nums",{"id":69,"title":70},"spot_dirty_signal","Spot dirty signal",{"id":72,"title":73},"automation_vs_judgment","Auto vs judgment",{"id":75,"title":76},"compare_branches_attribution","Compare & attrib",{"id":78,"title":79},"build_signal_culture","Signal culture",{"id":81,"title":82},"talk_to_human","Talk to human",{"id":84,"name":85,"type":86,"typeVersion":16,"position":87,"parameters":90,"category":92,"deletable":8,"connectable":8},"node_if_branch","IF: audit branch numbers","if",[88,89],480,40,{"buttonId":66,"operator":91},"equals","routing",{"id":94,"name":95,"type":96,"typeVersion":16,"position":97,"parameters":99,"category":51,"deletable":8,"connectable":8},"node_msg_branch","Branch numbers trust check","text-message",[98,89],760,{"text":100},"Branch numbers that deserve trust usually have three boring properties: (1) a clear source-of-truth (one system owns the definition), (2) a stable denominator (you can explain who/what was eligible), and (3) consistent collection (no silent changes in tracking, tags, or staffing).\n\nQuick checks before you repeat a number in a meeting:\n- **Denominator sanity:** “Out of how many opportunities?” If nobody can answer in one sentence, it’s not decision-ready.\n- **Lag & cutoff:** Is this counting the same time window across branches? Late postings create fake ‘winners.’\n- **Instrumentation change:** Any new POS flow, CRM field, call routing tweak, or campaign tagging change recently? That’s where ‘performance’ comes from most often.\n- **Small-sample gloss:** Averages hide chaos. Ask for median + count.\n\nPolished noise usually comes from proxies (e.g., ‘intent’ tags), partial coverage, or anything that depends on perfect human entry at a busy branch.",{"id":102,"name":103,"type":86,"typeVersion":16,"position":104,"parameters":106,"category":92,"deletable":8,"connectable":8},"node_if_dirty","IF: spot dirty signal",[88,105],140,{"buttonId":69,"operator":91},{"id":108,"name":109,"type":96,"typeVersion":16,"position":110,"parameters":111,"category":51,"deletable":8,"connectable":8},"node_msg_dirty","Dirty signal early warnings",[98,105],{"text":112},"Dirty signal rarely announces itself. It shows up as *plausible* consistency… right until you make a big call.\n\nEarly warning signs (the ‘smells’):\n- **Step-changes with no operational story:** If the chart jumps but nobody changed staffing, hours, routing, or offers, assume measurement changed.\n- **Too-perfect uniformity:** If every branch improves the same amount, that’s often a template, not reality.\n- **Missingness that’s not random:** Who doesn’t get recorded? Busy times, complex cases, certain teams—exactly where the truth lives.\n- **Label drift:** Conversation tags mean different things across branches. If definitions aren’t enforced, trends are theater.\n\nPractical move: pick 10 real records (calls/chats/leads) and audit them end-to-end. Dirty signal hates small, concrete samples.",{"id":114,"name":115,"type":86,"typeVersion":16,"position":116,"parameters":118,"category":92,"deletable":8,"connectable":8},"node_if_auto","IF: automation vs judgment",[88,117],240,{"buttonId":72,"operator":91},{"id":120,"name":121,"type":96,"typeVersion":16,"position":122,"parameters":123,"category":51,"deletable":8,"connectable":8},"node_msg_auto","When to trust automation",[98,117],{"text":124},"Trust automation when the world is *stable* and the cost of being wrong is *bounded*.\n\nGood automation territory:\n- Repetitive classification with tight definitions (e.g., clear eligibility rules)\n- Alerts on obvious breakages (tracking stops, volume drops to zero, impossible values)\n- Consistency enforcement (required fields, dedupe rules)\n\nHuman judgment still required when:\n- **Incentives are involved:** People will unconsciously (or consciously) game definitions.\n- **Edge cases matter:** A few ‘weird’ cases drive profit, risk, or reputation.\n- **The metric is a proxy:** Proxies can be useful, but they’re fragile—humans must sanity-check them against reality.\n\nRule of thumb: automate the *detection* of anomalies; use humans for the *interpretation* and the decision.",{"id":126,"name":127,"type":86,"typeVersion":16,"position":128,"parameters":130,"category":92,"deletable":8,"connectable":8},"node_if_compare","IF: compare branches & attribution",[88,129],340,{"buttonId":75,"operator":91},{"id":132,"name":133,"type":96,"typeVersion":16,"position":134,"parameters":135,"category":51,"deletable":8,"connectable":8},"node_msg_compare","Branch comparison traps",[98,129],{"text":136},"Branch comparisons go wrong in the same predictable ways:\n\nWhat teams misread first:\n- **Different mix, same metric:** A branch with easier demand looks ‘better’ unless you control for lead type, product mix, and seasonality.\n- **Capacity constraints:** A high-performing branch may simply be declining more (or recording less) because they’re overloaded.\n- **Attribution double-counting:** One customer touches multiple channels; the neatest report often assigns credit too confidently.\n- **Survivorship bias:** You’re comparing the cases that got logged, not the cases that happened.\n\nSafer comparison moves:\n- Compare **within the same segment** (same lead type + same time window).\n- Look at **rates + counts** together.\n- Treat attribution as a *hypothesis generator*, not a scoreboard.",{"id":138,"name":139,"type":86,"typeVersion":16,"position":140,"parameters":142,"category":92,"deletable":8,"connectable":8},"node_if_culture","IF: build signal culture",[88,141],440,{"buttonId":78,"operator":91},{"id":144,"name":145,"type":96,"typeVersion":16,"position":146,"parameters":147,"category":51,"deletable":8,"connectable":8},"node_msg_culture","Build a signal culture",[98,141],{"text":148},"A strong signal culture doesn’t create more dashboards—it creates faster, calmer decisions.\n\nHabits that actually work:\n- **One owner per metric:** Someone is accountable for definition changes, not just refreshes.\n- **‘What would change my mind?’ in every review:** If the answer is ‘nothing,’ you’re doing theatre.\n- **Decision logs:** Write down the decision, the metric used, and the expected result. This is how you detect bad signals later.\n- **Small audits as a ritual:** 10-record spot checks beat 100-slide decks.\n\nWitty but true: if your best metric can’t survive one skeptical question, it wasn’t a metric—it was a mood.",{"id":150,"name":151,"type":86,"typeVersion":16,"position":152,"parameters":154,"category":92,"deletable":8,"connectable":8},"node_if_human","IF: talk to a human",[88,153],540,{"buttonId":81,"operator":91},{"id":156,"name":157,"type":96,"typeVersion":16,"position":158,"parameters":159,"category":51,"deletable":8,"connectable":8},"node_msg_unknown","Unrecognized selection",[98,153],{"text":160},"I didn’t catch a menu selection. If this is a high-stakes decision—or the data feels ‘too clean’—I can hand you to Insights Support to sanity-check sources and definitions.",{"id":162,"name":163,"type":164,"typeVersion":16,"position":165,"parameters":167,"category":171,"deletable":8,"connectable":8},"node_fallback","Handoff to Insights Support","fallback",[166,153],1020,{"handoffMessage":168,"departmentId":169,"departmentName":170},"I’m handing you to Insights Support. Please share: the branch(es), the metric(s), the time window, and what decision you’re about to make—plus any recent process or tracking changes.","insights-support","Insights Support","terminal",[173,177,179,181,184,187,189,191,193,195,197,199,201,203,205,207],{"id":174,"source":35,"target":42,"sourceHandle":175,"targetHandle":175,"type":176},"c1","main","smoothstep",{"id":178,"source":42,"target":53,"sourceHandle":175,"targetHandle":175,"type":176},"c2",{"id":180,"source":53,"target":84,"sourceHandle":175,"targetHandle":175,"type":176},"c3",{"id":182,"source":84,"target":94,"sourceHandle":183,"targetHandle":175,"type":176},"c4_true","true",{"id":185,"source":84,"target":102,"sourceHandle":186,"targetHandle":175,"type":176},"c4_false","false",{"id":188,"source":102,"target":108,"sourceHandle":183,"targetHandle":175,"type":176},"c5_true",{"id":190,"source":102,"target":114,"sourceHandle":186,"targetHandle":175,"type":176},"c5_false",{"id":192,"source":114,"target":120,"sourceHandle":183,"targetHandle":175,"type":176},"c6_true",{"id":194,"source":114,"target":126,"sourceHandle":186,"targetHandle":175,"type":176},"c6_false",{"id":196,"source":126,"target":132,"sourceHandle":183,"targetHandle":175,"type":176},"c7_true",{"id":198,"source":126,"target":138,"sourceHandle":186,"targetHandle":175,"type":176},"c7_false",{"id":200,"source":138,"target":144,"sourceHandle":183,"targetHandle":175,"type":176},"c8_true",{"id":202,"source":138,"target":150,"sourceHandle":186,"targetHandle":175,"type":176},"c8_false",{"id":204,"source":150,"target":162,"sourceHandle":183,"targetHandle":175,"type":176},"c9_true",{"id":206,"source":150,"target":156,"sourceHandle":186,"targetHandle":175,"type":176},"c9_false",{"id":208,"source":156,"target":162,"sourceHandle":175,"targetHandle":175,"type":176},"c10","automation",[28,29,30,211,212],"attribution","automation-guardrails",[214],"Calypso","intermediate","2026-07-07T11:05:22.215Z","/en/workflows/branch-metrics-conversation-signal-triage",{"en":217},{"title":220,"description":221,"ogDescription":222,"twitterDescription":223,"canonicalPath":217,"robots":224,"schemaType":225,"alternates":226},"Branch Metrics & Signal Triage for Decisions","Route operators through branch metric trust checks, dirty signal spotting, and automation vs judgment—with a human handoff when needed.","A practical decision triage menu for branch metrics and conversation signals: spot dirty data early, avoid attribution traps, and know when humans must step in.","Decision ready signal triage for branch metrics and conversations: identify polished noise, catch dirty signal early, and choose automation vs judgment.","index,follow","HowTo",[227],{"hreflang":6,"href":217},1785947667554]