[{"data":1,"prerenderedAt":177},["ShallowReactive",2],{"/en/workflows/branch-signal-decision-guardrails":3},{"id":4,"slug":5,"locale":6,"translationGroupId":7,"localeSwitchApproved":8,"title":9,"description":10,"documentationMarkdown":11,"workflowJson":12,"category":156,"tags":157,"integrations":161,"difficulty":164,"author":165,"verified":33,"featured":33,"date":166,"modified":166,"icon":7,"imageSrc":7,"path":167,"alternates":168,"seo":169},"0e1756bf-3675-4f87-9fe5-0e4e80adc69e","branch-signal-decision-guardrails","en",null,true,"Branch Signal Decision Guardrails","A chat-based decision support workflow that helps leaders pressure-test branch metrics, spot dirty signals early, and know when to trust automation vs. human judgment—before a confident meeting turns confidently wrong.","## How it works\nThis workflow turns “we have numbers” into “we have decision‑worthy evidence.” It starts with a knowledge-base answer (so you’re not reinventing guidance every conversation), then offers a few sharp, decision-shaped routes: sanity-checking a metric, comparing branches without fooling yourself, and deciding where automation is safe vs. where human judgment is still doing the real work.\n\nIt’s built for the moment right before decisions harden: the pre-meeting, the weekly review, the “why is Branch 12 down?” thread. The goal is not perfect data—it’s catching polished noise early, keeping messy truth intact, and getting to a call you can defend.\n\n## Key features\n- Knowledge base policy up front to deliver consistent, reusable guidance before routing\n- Interactive menu that routes leaders into common decision situations (sanity check, branch comparison, automation vs judgment)\n- Practical “red flag” checklists that help spot dirty signal before it derails a meeting\n- Clear guardrails on which branch numbers are usually trustworthy vs. cosmetically clean\n- Optional human handoff path when the situation needs investigation or authority\n\n## Step-by-step\n1. **Trigger:** A user starts the conversation in Calypso (e.g., WhatsApp chat entry point).\n2. **Knowledge base guidance:** Calypso applies the **Knowledge Base Policy** to answer naturally in plain language, using your existing signal/metric guidance.\n3. **Choose the decision situation:** The workflow sends an interactive menu with four options:\n   1) **Run a signal sanity check**\n   2) **Compare branches safely**\n   3) **Automation vs. human judgment**\n   4) **Talk to a person**\n4. **Route based on selection:**\n   - If **Run a signal sanity check** → user receives a fast checklist to detect “clean-looking, wrong” signals.\n   - If **Compare branches safely** → user receives a practical comparison framework (what usually gets misread first).\n   - If **Automation vs. human judgment** → user receives rules of thumb for when automation is safe, and when leaders must stay in the loop.\n   - If **Talk to a person** → the workflow hands off to routing with a clear handoff message.\n\n## Setup requirements\n- **Calypso Knowledge Base:** Add or confirm articles that cover branch metric definitions, common failure modes (attribution, conversation tagging, outages), and decision guidelines. No external credentials required.\n- **Channel support for interactive messages:** Use a Calypso channel that supports interactive buttons (e.g., **WhatsApp**).\n- **Optional routing:** If you want the “Talk to a person” path to go somewhere specific, ensure a routing department exists and update the department fields in the fallback node (otherwise it can remain a general handoff).",{"id":13,"teamId":14,"name":9,"version":15,"workflowVersion":16,"nodes":17,"connections":132,"routingEnabled":8,"active":33},"wf_branch_signal_decision_guardrails_v1","calypso-public-library","1.0.0",1,[18,34,40,52,77,87,92,98,104,111,117,123],{"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],100,80,{"name":9,"description":26,"tags":27,"triggerType":31},"Decision support for branch metrics and messy signals: sanity checks, safer comparisons, automation vs judgment, and optional handoff.",[28,29,30],"signal-hygiene","branch-metrics","decision-systems","input","policy",false,{"id":35,"name":36,"type":31,"typeVersion":16,"position":37,"parameters":39,"category":31,"deletable":33,"connectable":8},"node_input","Incoming message",[23,38],200,{},{"id":41,"name":42,"type":43,"typeVersion":16,"position":44,"parameters":46,"category":51,"deletable":8,"connectable":8},"node_kb_policy","Knowledge base guidance","knowledge-base-policy",[45,38],320,{"enabled":8,"fallbackToRouting":8,"sticky":33,"stickyMode":47,"activationOpener":48,"personalization":50},"default",{"enabled":8,"instruction":49},"Use a practical, plain-spoken tone. Focus on decision readiness: what to trust, what to verify, and what commonly breaks first in branch metrics, conversations, and attribution. Avoid academic language; use crisp checklists and judgment calls.",{"useContactName":8},"response",{"id":53,"name":54,"type":55,"typeVersion":16,"position":56,"parameters":58,"category":51,"deletable":8,"connectable":8},"node_menu","Choose what you need","interactive-message",[57,38],560,{"messageType":59,"headerText":60,"bodyText":61,"footerText":62,"sectionTitle":63,"buttons":64,"ctaDisplayText":63,"ctaUrl":63},"button","Decision guardrails for messy signals","Pick the situation. I’ll keep it practical and meeting-proof.","Tip: A metric can be precise and still be wrong.","",[65,68,71,74],{"id":66,"title":67},"sanity_check","Signal sanity check",{"id":69,"title":70},"compare_branches","Compare branches",{"id":72,"title":73},"automation_vs_judgment","Auto vs judgment",{"id":75,"title":76},"handoff","Talk to a person",{"id":78,"name":79,"type":80,"typeVersion":16,"position":81,"parameters":84,"category":86,"deletable":8,"connectable":8},"node_if_sanity","If: sanity check","if",[82,83],800,120,{"buttonId":66,"operator":85},"equals","routing",{"id":88,"name":89,"type":80,"typeVersion":16,"position":90,"parameters":91,"category":86,"deletable":8,"connectable":8},"node_if_compare","If: compare branches",[82,38],{"buttonId":69,"operator":85},{"id":93,"name":94,"type":80,"typeVersion":16,"position":95,"parameters":97,"category":86,"deletable":8,"connectable":8},"node_if_auto","If: automation vs judgment",[82,96],280,{"buttonId":72,"operator":85},{"id":99,"name":100,"type":80,"typeVersion":16,"position":101,"parameters":103,"category":86,"deletable":8,"connectable":8},"node_if_handoff","If: handoff",[82,102],360,{"buttonId":75,"operator":85},{"id":105,"name":67,"type":106,"typeVersion":16,"position":107,"parameters":109,"category":51,"deletable":8,"connectable":8},"node_text_sanity","text-message",[108,83],1040,{"text":110},"Signal sanity check (fast, not fussy):\n\n1) **Definition check:** What exactly is being counted—and what’s excluded? If two people define it differently, it’s not decision-ready.\n2) **Denominator check:** Is the rate driven by volume changes (fewer customers/leads/calls) rather than true performance?\n3) **Freshness check:** Any delays, backfills, or system changes in the last 7–14 days?\n4) **Coverage check:** Are some branches missing events, tagging, or conversation capture? Missing data often looks like “great performance.”\n5) **Incentive check:** If someone could ‘game’ this number, assume someone eventually will.\n\nIf even one of these is shaky, don’t discard the metric—**downgrade it**: use it for questions, not conclusions.",{"id":112,"name":113,"type":106,"typeVersion":16,"position":114,"parameters":115,"category":51,"deletable":8,"connectable":8},"node_text_compare","Compare branches safely",[108,38],{"text":116},"Comparing branches without fooling yourself:\n\n- **Normalize first:** Compare rates (and confidence) before totals. Totals mostly measure branch size.\n- **Check mix shifts:** A branch can ‘improve’ simply by getting easier cases/leads.\n- **Look for instrumentation differences:** Different tagging discipline, different conversation capture, different staff workflows = fake gaps.\n- **Use “two-proof” decisions:** Don’t act on a single source. Pair **one operational signal** (e.g., conversations) with **one outcome signal** (e.g., conversions) before changing policy.\n- **Beware tiny sample heroics:** If the sample is small, treat it as a story prompt, not evidence.\n\nThe thing teams misread most: **a clean ranking**. Rankings feel decisive even when the measurement isn’t comparable.",{"id":118,"name":119,"type":106,"typeVersion":16,"position":120,"parameters":121,"category":51,"deletable":8,"connectable":8},"node_text_auto","Automation vs judgment",[108,96],{"text":122},"When to trust automation (and when to keep a human in the loop):\n\n**Trust automation when:**\n- The input is consistent (same definitions, stable capture).\n- Errors are obvious and reversible.\n- The decision is low-stakes or bounded (limits, guardrails, approvals).\n\n**Require human judgment when:**\n- The metric can be gamed, or incentives are strong.\n- Small changes trigger big consequences (staffing cuts, branch reorgs, customer policy shifts).\n- You’re mixing signals (conversations + attribution + branch events) and one source can dominate by accident.\n\nRule of thumb: automation is great at **repeatable sorting**. Humans are still needed for **meaning, exceptions, and consequences**.",{"id":124,"name":125,"type":126,"typeVersion":16,"position":127,"parameters":128,"category":131,"deletable":8,"connectable":8},"node_fallback_handoff","Handoff to a person","fallback",[108,102],{"handoffMessage":129,"departmentId":63,"departmentName":130},"Got it—this needs a human look. Please share: (1) which branch, (2) which metric/signal, (3) time range, and (4) what decision you’re trying to make. I’m handing this over now.","Operations Support","terminal",[133,137,139,141,143,145,147,150,152,154],{"id":134,"source":35,"target":41,"sourceHandle":135,"targetHandle":136,"type":47},"conn_input_to_kb","out","in",{"id":138,"source":41,"target":53,"sourceHandle":135,"targetHandle":136,"type":47},"conn_kb_to_menu",{"id":140,"source":53,"target":78,"sourceHandle":135,"targetHandle":136,"type":47},"conn_menu_to_if_sanity",{"id":142,"source":53,"target":88,"sourceHandle":135,"targetHandle":136,"type":47},"conn_menu_to_if_compare",{"id":144,"source":53,"target":93,"sourceHandle":135,"targetHandle":136,"type":47},"conn_menu_to_if_auto",{"id":146,"source":53,"target":99,"sourceHandle":135,"targetHandle":136,"type":47},"conn_menu_to_if_handoff",{"id":148,"source":78,"target":105,"sourceHandle":149,"targetHandle":136,"type":47},"conn_if_sanity_to_text","true",{"id":151,"source":88,"target":112,"sourceHandle":149,"targetHandle":136,"type":47},"conn_if_compare_to_text",{"id":153,"source":93,"target":118,"sourceHandle":149,"targetHandle":136,"type":47},"conn_if_auto_to_text",{"id":155,"source":99,"target":124,"sourceHandle":149,"targetHandle":136,"type":47},"conn_if_handoff_to_fallback","automation",[28,29,30,158,159,160],"data-trust","triage","leadership",[162,163],"WhatsApp","Calypso Knowledge Base","intermediate","Calypso","2026-05-13T11:02:51.967Z","/en/workflows/branch-signal-decision-guardrails",{"en":167},{"title":9,"description":170,"ogDescription":171,"twitterDescription":172,"canonicalPath":167,"robots":173,"schemaType":174,"alternates":175},"Route leaders through signal sanity checks, branch comparisons, and automation vs judgment prompts—so decisions rely on evidence, not polished noise.","A practical decision workflow for branch metrics: spot dirty signal early, compare branches without self deception, and know when automation is safe vs. when humans must decide.","Turn messy branch signals into decision worthy evidence with sanity checks, safer comparisons, and clear automation vs. judgment guardrails.","index,follow","HowTo",[176],{"hreflang":6,"href":167},1780761214383]