[{"data":1,"prerenderedAt":290},["ShallowReactive",2],{"/en/workflows/branch-signal-sensemaking-coach":3},{"id":4,"slug":5,"locale":6,"translationGroupId":7,"localeSwitchApproved":8,"title":9,"description":10,"documentationMarkdown":11,"workflowJson":12,"category":269,"tags":270,"integrations":274,"difficulty":277,"author":278,"verified":33,"featured":33,"date":279,"modified":279,"icon":7,"imageSrc":7,"path":280,"alternates":281,"seo":282},"998fb52a-e217-4795-a2a0-979521b52067","branch-signal-sensemaking-coach","en",null,true,"Branch Signal Sensemaking Coach","An operator-friendly decision assistant that helps teams sort trustworthy branch signals from polished noise, spot dirty data before meetings, and choose when to rely on automation vs human judgment.","## How it works\nThis workflow turns “messy evidence” (branch numbers, conversation notes, attribution claims, and ad-hoc anecdotes) into decision-ready guidance—without pretending every dataset is clean just because it’s formatted.\n\nCustomers (or internal teams) choose what they’re trying to decide from a menu. Calypso first applies your Knowledge Base guidance (definitions, metric rules, and policy), then routes them to targeted coaching that helps them pressure-test the signal before they walk into a confident-but-wrong meeting.\n\n## Key features\n- Uses a Knowledge Base policy first, so responses stay consistent with your measurement definitions and business rules.\n- Menu-based routing with decision-shaped options (metric trust, dirty signal detection, automation vs judgment, branch comparisons, signal culture).\n- Practical checklists that call out the most common “looks fine until it isn’t” failure modes.\n- A single follow-up menu after each answer to either return to the main menu or escalate to a human.\n- Built-in handoff path to route ambiguous or high-stakes cases to your Analytics/Ops team.\n\n## Step-by-step\n1. **Trigger:** A user starts the flow (incoming message).\n2. **Knowledge base alignment:** Calypso applies your Knowledge Base policy to anchor definitions and tone.\n3. **Choose the decision:** The user selects what they need help with from the interactive menu.\n4. **Route to the right coaching path:** The workflow checks which button they clicked and sends the matching guidance.\n5. **Follow-up options:** After the guidance, the user can either return to the main menu or escalate for human review.\n6. **Escalation (optional):** If requested, the workflow hands off to the specified department with a clear handoff message.\n\n## Setup requirements\n- A connected messaging channel in Calypso (e.g., **WhatsApp** or **Calypso Inbox**) that supports interactive buttons.\n- A published **Knowledge Base** with your metric definitions, branch reporting rules, attribution conventions, and escalation guidelines.\n- No additional credentials are required to deploy this workflow.",{"id":13,"teamId":14,"name":9,"version":15,"workflowVersion":16,"nodes":17,"connections":213,"routingEnabled":8,"active":33},"wf_branch_signal_sensemaking_coach_v1","calypso-public-library","1.0.0",1,[18,34,41,52,83,93,98,104,110,116,122,130,136,142,148,154,170,177,183,190,200,207],{"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],-220,40,{"name":9,"description":26,"tags":27,"triggerType":31},"Guided coaching to help teams decide what branch signals deserve trust, spot dirty signal early, and choose when automation needs human judgment.",[28,29,30],"signal-quality","branch-metrics","decision-making","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],-40,200,{},{"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",[39,39],{"enabled":8,"fallbackToRouting":8,"sticky":8,"stickyMode":47,"activationOpener":48,"personalization":50},"default",{"enabled":8,"instruction":49},"You help teams turn messy branch signals into trustworthy decisions. Be practical, direct, and specific. Prefer simple checks over jargon. If uncertainty is high or stakes are high, recommend escalation.",{"useContactName":8},"response",{"id":53,"name":54,"type":55,"typeVersion":16,"position":56,"parameters":58,"category":51,"deletable":8,"connectable":8},"node_menu_main","Decision menu","interactive-message",[57,39],460,{"messageType":59,"headerText":60,"bodyText":61,"footerText":62,"sectionTitle":63,"buttons":64,"ctaDisplayText":63,"ctaUrl":63},"button","Signal Sensemaking","Pick what you’re trying to decide. I’ll help you pressure-test the signal before it pressure-tests you in a meeting.","Tip: when data looks unusually clean, ask what got excluded.","",[65,68,71,74,77,80],{"id":66,"title":67},"metric_trust","Trustworthy numbers",{"id":69,"title":70},"dirty_signal","Spot dirty signal",{"id":72,"title":73},"auto_vs_human","Auto vs judgment",{"id":75,"title":76},"compare_branches","Compare branches",{"id":78,"title":79},"signal_culture","Signal culture",{"id":81,"title":82},"talk_to_human","Escalate to a human",{"id":84,"name":85,"type":86,"typeVersion":16,"position":87,"parameters":90,"category":92,"deletable":8,"connectable":8},"node_if_metric_trust","If: metric trust","if",[88,89],720,120,{"buttonId":66,"operator":91},"equals","routing",{"id":94,"name":95,"type":86,"typeVersion":16,"position":96,"parameters":97,"category":92,"deletable":8,"connectable":8},"node_if_dirty_signal","If: dirty signal",[88,39],{"buttonId":69,"operator":91},{"id":99,"name":100,"type":86,"typeVersion":16,"position":101,"parameters":103,"category":92,"deletable":8,"connectable":8},"node_if_auto_vs_human","If: automation vs human",[88,102],280,{"buttonId":72,"operator":91},{"id":105,"name":106,"type":86,"typeVersion":16,"position":107,"parameters":109,"category":92,"deletable":8,"connectable":8},"node_if_compare_branches","If: compare branches",[88,108],360,{"buttonId":75,"operator":91},{"id":111,"name":112,"type":86,"typeVersion":16,"position":113,"parameters":115,"category":92,"deletable":8,"connectable":8},"node_if_signal_culture","If: signal culture",[88,114],440,{"buttonId":78,"operator":91},{"id":117,"name":118,"type":86,"typeVersion":16,"position":119,"parameters":121,"category":92,"deletable":8,"connectable":8},"node_if_talk_to_human","If: talk to human",[88,120],520,{"buttonId":81,"operator":91},{"id":123,"name":124,"type":125,"typeVersion":16,"position":126,"parameters":128,"category":51,"deletable":8,"connectable":8},"node_text_metric_trust","Guidance: which numbers to trust","text-message",[127,89],980,{"text":129},"Here’s the fast way to tell which branch numbers deserve trust vs polished noise:\n\n1) **Lineage check (first):** Can you name the source system *and* the rule that produced the number? If it’s “a spreadsheet someone updates,” treat it as a claim, not a fact.\n2) **Denominator sanity:** If the denominator quietly changed (active customers, eligible leads, staffed hours), the trend is theatre.\n3) **Lag + revision:** Good operational metrics often get revised. If yours never changes, it might be ignoring late-arriving truth.\n4) **Selection pressure:** Ask: *who benefits if this looks good?* The most “reliable” metrics tend to be the ones nobody gets paid to game.\n5) **Cross-check:** One metric is a story. Two independent sources is evidence. Try one operational cross-check (e.g., staffing, queue time, inventory movement).\n\nIf you tell me the metric name + how it’s produced, I can flag the usual failure modes.",{"id":131,"name":132,"type":125,"typeVersion":16,"position":133,"parameters":134,"category":51,"deletable":8,"connectable":8},"node_text_dirty_signal","Guidance: spot dirty signal",[127,39],{"text":135},"Dirty signal usually shows up wearing a suit. Here are the tells *before* a confident meeting goes off the rails:\n\n- **Too-smooth trend lines:** Real life is lumpy. Perfect smoothness often means heavy filtering, backfilling, or aggregation hiding volatility.\n- **Metric moves, reality doesn’t:** If the KPI improved but customer complaints, refunds, or wait times didn’t budge—suspect definition changes or selective inclusion.\n- **Sudden step-changes on a Monday:** Often a tracking change, routing change, or a “new rule” deployed—not operational behavior.\n- **Outliers that vanish:** If exceptions get manually removed, you’re cleaning away the truth (and the risk).\n- **Missingness that correlates:** If data is missing *more* when a branch is busy, performance will look better exactly when it’s worse.\n\nQuick move: ask for a 10-row sample of raw records behind the chart. If nobody can produce it, don’t bet the quarter on it.",{"id":137,"name":138,"type":125,"typeVersion":16,"position":139,"parameters":140,"category":51,"deletable":8,"connectable":8},"node_text_auto_vs_human","Guidance: automation vs judgment",[127,102],{"text":141},"When to trust automation vs when to keep a human in the loop:\n\n**Automation is great when:**\n- The decision is **frequent, reversible, and low-regret** (e.g., routine follow-ups, standard routing).\n- Inputs are **stable and observable** (clear timestamps, consistent event capture).\n- You can define **failure bounds** (what “wrong” looks like and how you’ll detect it).\n\n**Humans should stay involved when:**\n- The decision is **high-stakes or one-way** (pricing overrides, staffing cuts, public commitments).\n- The signal can be **gamed** or is socially negotiated (self-reported outcomes, “reasons” in notes).\n- Edge cases carry real cost (VIP customers, compliance, safety).\n\nRule of thumb: automate the *default*, and design an explicit, easy escalation path for anything weird. If escalation feels like failure, people will hide weirdness until it becomes expensive.",{"id":143,"name":144,"type":125,"typeVersion":16,"position":145,"parameters":146,"category":51,"deletable":8,"connectable":8},"node_text_compare_branches","Guidance: compare branches",[127,108],{"text":147},"Branch comparisons are where smart teams accidentally reward the wrong behavior. Use these guardrails:\n\n1) **Compare like with like:** Segment by branch type, maturity, footfall, and staffing model. Ranking mixed populations is a confidence trick.\n2) **Normalize the constraint:** If one branch is capacity-limited (staffing, inventory, appointment slots), raw output metrics punish reality and reward luck.\n3) **Watch the proxy swap:** Teams “improve” a metric by changing what gets counted (e.g., closing tickets faster by closing them earlier).\n4) **Attribution humility:** Conversations, campaigns, and walk-ins blend. When someone claims single-cause attribution, ask what got excluded.\n5) **Look for cost of winning:** If a top branch has higher refunds, churn, or complaint volume later, you’re seeing borrowed performance.\n\nIf you share what you’re comparing (metric + timeframe), I can suggest the safest normalization and a fair peer group.",{"id":149,"name":150,"type":125,"typeVersion":16,"position":151,"parameters":152,"category":51,"deletable":8,"connectable":8},"node_text_signal_culture","Guidance: build signal culture",[127,114],{"text":153},"A strong signal culture helps decisions happen (not just slides). Start here:\n\n- **One-page metric contracts:** For key metrics, document definition, source, refresh cadence, and known blind spots. If it can’t fit on a page, it’s not decision-ready.\n- **Make uncertainty speak:** Require every KPI readout to include “what would change my mind?” and “what could be wrong here?”\n- **Celebrate corrections:** If teams get punished for revising numbers, they’ll stop revising numbers. That’s how bad data wins.\n- **Default to samples:** A chart without a few raw examples nearby invites story-telling. Samples keep people honest.\n- **Decision log:** Track the decision, the signal used, and what happened. Feedback is how you earn the right to trust metrics.\n\nThe goal isn’t perfect data. It’s *reliable decisions under imperfect data*.",{"id":155,"name":156,"type":55,"typeVersion":16,"position":157,"parameters":160,"category":51,"deletable":8,"connectable":8},"node_menu_followup","Follow-up options",[158,159],1240,260,{"messageType":59,"headerText":161,"bodyText":162,"footerText":163,"sectionTitle":63,"buttons":164,"ctaDisplayText":63,"ctaUrl":63},"Next step","Want another angle, or should we pull in a human for review?","High-stakes + high-uncertainty is a great time to escalate.",[165,168],{"id":166,"title":167},"back_to_menu","Start over",{"id":169,"title":82},"escalate_handoff",{"id":171,"name":172,"type":86,"typeVersion":16,"position":173,"parameters":176,"category":92,"deletable":8,"connectable":8},"node_if_back_to_menu","If: start over",[174,175],1500,220,{"buttonId":166,"operator":91},{"id":178,"name":179,"type":86,"typeVersion":16,"position":180,"parameters":182,"category":92,"deletable":8,"connectable":8},"node_if_escalate_handoff","If: escalate handoff",[174,181],300,{"buttonId":169,"operator":91},{"id":184,"name":185,"type":125,"typeVersion":16,"position":186,"parameters":188,"category":51,"deletable":8,"connectable":8},"node_text_restart","Start over instructions",[187,175],1760,{"text":189},"To start over, send a new message and I’ll show the decision menu again. (Keeping this simple prevents the flow from looping in circles.)",{"id":191,"name":192,"type":193,"typeVersion":16,"position":194,"parameters":195,"category":199,"deletable":8,"connectable":8},"node_fallback_handoff","Handoff to Analytics/Ops","fallback",[187,114],{"handoffMessage":196,"departmentId":197,"departmentName":198},"Handing this over to a human for review. Please share: (1) which branch(es), (2) the metric or claim, (3) timeframe, and (4) what decision you’re trying to make.","analytics-ops","Analytics & Ops","terminal",{"id":201,"name":202,"type":125,"typeVersion":16,"position":203,"parameters":205,"category":51,"deletable":8,"connectable":8},"node_text_unknown_main","Unknown selection (main)",[127,204],560,{"text":206},"I didn’t catch that selection. Please tap one of the buttons in the menu so I can route you to the right checklist.",{"id":208,"name":209,"type":125,"typeVersion":16,"position":210,"parameters":211,"category":51,"deletable":8,"connectable":8},"node_text_unknown_followup","Unknown selection (follow-up)",[187,181],{"text":212},"I didn’t catch that. Choose **Start over** or **Escalate to a human**.",[214,219,221,223,226,229,231,233,235,237,239,241,243,245,247,249,251,253,255,257,259,261,263,265,267],{"id":215,"source":35,"target":42,"sourceHandle":216,"targetHandle":217,"type":218},"edge_input_to_kb","out","in","edge",{"id":220,"source":42,"target":53,"sourceHandle":216,"targetHandle":217,"type":218},"edge_kb_to_menu",{"id":222,"source":53,"target":84,"sourceHandle":216,"targetHandle":217,"type":218},"edge_menu_to_if_metric",{"id":224,"source":84,"target":123,"sourceHandle":225,"targetHandle":217,"type":218},"edge_if_metric_true_to_text","true",{"id":227,"source":84,"target":94,"sourceHandle":228,"targetHandle":217,"type":218},"edge_if_metric_false_to_if_dirty","false",{"id":230,"source":94,"target":131,"sourceHandle":225,"targetHandle":217,"type":218},"edge_if_dirty_true_to_text",{"id":232,"source":94,"target":99,"sourceHandle":228,"targetHandle":217,"type":218},"edge_if_dirty_false_to_if_auto",{"id":234,"source":99,"target":137,"sourceHandle":225,"targetHandle":217,"type":218},"edge_if_auto_true_to_text",{"id":236,"source":99,"target":105,"sourceHandle":228,"targetHandle":217,"type":218},"edge_if_auto_false_to_if_compare",{"id":238,"source":105,"target":143,"sourceHandle":225,"targetHandle":217,"type":218},"edge_if_compare_true_to_text",{"id":240,"source":105,"target":111,"sourceHandle":228,"targetHandle":217,"type":218},"edge_if_compare_false_to_if_culture",{"id":242,"source":111,"target":149,"sourceHandle":225,"targetHandle":217,"type":218},"edge_if_culture_true_to_text",{"id":244,"source":111,"target":117,"sourceHandle":228,"targetHandle":217,"type":218},"edge_if_culture_false_to_if_handoff",{"id":246,"source":117,"target":191,"sourceHandle":225,"targetHandle":217,"type":218},"edge_if_handoff_true_to_fallback",{"id":248,"source":117,"target":201,"sourceHandle":228,"targetHandle":217,"type":218},"edge_if_handoff_false_to_unknown_main",{"id":250,"source":123,"target":155,"sourceHandle":216,"targetHandle":217,"type":218},"edge_metric_text_to_followup",{"id":252,"source":131,"target":155,"sourceHandle":216,"targetHandle":217,"type":218},"edge_dirty_text_to_followup",{"id":254,"source":137,"target":155,"sourceHandle":216,"targetHandle":217,"type":218},"edge_auto_text_to_followup",{"id":256,"source":143,"target":155,"sourceHandle":216,"targetHandle":217,"type":218},"edge_compare_text_to_followup",{"id":258,"source":149,"target":155,"sourceHandle":216,"targetHandle":217,"type":218},"edge_culture_text_to_followup",{"id":260,"source":155,"target":171,"sourceHandle":216,"targetHandle":217,"type":218},"edge_followup_to_if_back",{"id":262,"source":171,"target":184,"sourceHandle":225,"targetHandle":217,"type":218},"edge_if_back_true_to_restart",{"id":264,"source":171,"target":178,"sourceHandle":228,"targetHandle":217,"type":218},"edge_if_back_false_to_if_escalate",{"id":266,"source":178,"target":191,"sourceHandle":225,"targetHandle":217,"type":218},"edge_if_escalate_true_to_fallback",{"id":268,"source":178,"target":208,"sourceHandle":228,"targetHandle":217,"type":218},"edge_if_escalate_false_to_unknown_followup","automation",[28,29,30,271,272,273],"data-hygiene","attribution","ops-insights",[275,276],"WhatsApp","Calypso Inbox","intermediate","Calypso","2026-06-29T11:04:34.934Z","/en/workflows/branch-signal-sensemaking-coach",{"en":280},{"title":9,"description":283,"ogDescription":284,"twitterDescription":285,"canonicalPath":280,"robots":286,"schemaType":287,"alternates":288},"Help teams judge which branch signals to trust, spot dirty data early, and know when automation needs human judgment—via guided routing.","A guided decision coach for branch metrics and messy signals: detect polished noise, sanity check comparisons, and escalate high stakes calls when needed.","Turn messy branch signals into better decisions. Spot dirty data before meetings, pressure test comparisons, and know when automation needs humans.","index,follow","HowTo",[289],{"hreflang":6,"href":280},1785947667682]