[{"data":1,"prerenderedAt":276},["ShallowReactive",2],{"/en/workflows/signal-decision-safety-checks":3},{"id":4,"slug":5,"locale":6,"translationGroupId":7,"localeSwitchApproved":8,"title":9,"description":10,"documentationMarkdown":11,"workflowJson":12,"category":256,"tags":257,"integrations":260,"difficulty":263,"author":264,"verified":33,"featured":33,"date":265,"modified":265,"icon":7,"imageSrc":7,"path":266,"alternates":267,"seo":268},"34e6c173-e7ba-41a9-9e5c-8a71bf398b8f","signal-decision-safety-checks","en",null,true,"Signal Decision Safety Checks","A guided decision coach that helps teams separate trustworthy branch signals from polished noise, spot dirty data before meetings, and choose when automation needs human judgment.","## How it works\nThis workflow turns messy branch-level metrics, conversations, and event “facts” into decision-ready guidance—without pretending your data is cleaner than it is. It starts by letting the Knowledge Base answer specific questions, then offers a menu of practical safety checks that help leaders avoid confident wrong decisions.\n\nUse it when a team is about to compare branches, defend a dashboard number, or automate a call-and-response process. The outcome isn’t more slides—it’s clearer judgment about what to trust, what to verify, and when to pull a human into the loop.\n\n## Key features\n- Knowledge Base-first answers for common signal and measurement questions, with routing fallback to guided triage\n- A decision-shaped menu focused on branch numbers, dirty signals, automation vs. judgment, comparisons, and attribution\n- Practical “reality checks” delivered as concise message playbooks (built for use right before a meeting)\n- Built-in loop to run multiple checks in one conversation without restarting\n- Optional handoff to an Insights team when the situation needs human review\n\n## Step-by-step\n1. **Trigger:** A user starts the workflow from Calypso Conversations.\n2. **Knowledge Base assist:** The workflow first attempts to answer using your Knowledge Base (helpful for “what does this metric mean?” and “how is this tracked?”).\n3. **Choose a safety check:** The user selects one option from the interactive menu:\n   - *Which branch numbers deserve trust?*\n   - *Spot dirty signal before the meeting*\n   - *Trust automation vs. human judgment*\n   - *Comparing branches & attribution traps*\n   - *Build a signal culture (less slide, more decision)*\n4. **Get the playbook:** The workflow sends a tailored, operator-friendly checklist for the chosen situation.\n5. **Next step:** The user chooses whether to run another check or request an Insights handoff.\n6. **Escalation (optional):** If requested, the workflow routes the conversation to the **Insights** department with a clear handoff message.\n\n## Setup requirements\n- **Calypso Conversations** enabled for your workspace\n- **Calypso Knowledge Base** connected and populated (recommended). If you don’t have KB content yet, the guided safety checks still work.\n- Optional: an **Insights** (or equivalent) department configured for handoff routing. No external credentials are required.",{"id":13,"teamId":14,"name":9,"version":15,"workflowVersion":16,"nodes":17,"connections":204,"routingEnabled":8,"active":33},"wf_signal_decision_safety_checks","calypso-public-library","1.0.0",1,[18,34,40,52,80,90,96,102,108,114,122,128,134,140,146,153,168,174,180,187,197],{"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],60,40,{"name":9,"description":26,"tags":27,"triggerType":31},"Guided decision coach for trustworthy branch signals, dirty-signal spotting, and automation guardrails.",[28,29,30],"signal-quality","decision-systems","branch-metrics","input","policy",false,{"id":35,"name":36,"type":31,"typeVersion":16,"position":37,"parameters":39,"category":31,"deletable":33,"connectable":8},"node_input","Conversation started",[23,38],160,{},{"id":41,"name":42,"type":43,"typeVersion":16,"position":44,"parameters":46,"category":51,"deletable":8,"connectable":8},"node_kb_policy","Knowledge Base assist","knowledge-base-policy",[45,38],300,{"enabled":8,"fallbackToRouting":8,"sticky":33,"stickyMode":47,"activationOpener":48,"personalization":50},"default",{"enabled":8,"instruction":49},"If the user asks what a metric means, how it’s computed, or what a dashboard number represents, answer from the Knowledge Base first. If they’re heading into a decision, route them into the safety-check menu.",{"useContactName":8},"response",{"id":53,"name":54,"type":55,"typeVersion":16,"position":56,"parameters":58,"category":51,"deletable":8,"connectable":8},"node_menu_main","Safety check menu","interactive-message",[57,38],560,{"messageType":59,"headerText":9,"bodyText":60,"footerText":61,"sectionTitle":62,"buttons":63,"ctaDisplayText":79,"ctaUrl":79},"list","Pick what you’re trying to decide. I’ll help you separate trustworthy signals from polished noise—before a confident meeting goes off the rails.","Tip: If a number looks too clean, it usually is.","Choose a safety check",[64,67,70,73,76],{"id":65,"title":66},"check_branch_numbers","Trust branch numbers",{"id":68,"title":69},"spot_dirty_signal","Spot dirty signal",{"id":71,"title":72},"automation_vs_judgment","Automation vs judgment",{"id":74,"title":75},"compare_branches_attribution","Compare & attribution",{"id":77,"title":78},"build_signal_culture","Signal culture","",{"id":81,"name":82,"type":83,"typeVersion":16,"position":84,"parameters":87,"category":89,"deletable":8,"connectable":8},"node_if_branch_numbers","If: Branch numbers","if",[85,86],820,120,{"buttonId":65,"operator":88},"equals","routing",{"id":91,"name":92,"type":83,"typeVersion":16,"position":93,"parameters":95,"category":89,"deletable":8,"connectable":8},"node_if_dirty_signal","If: Dirty signal",[85,94],220,{"buttonId":68,"operator":88},{"id":97,"name":98,"type":83,"typeVersion":16,"position":99,"parameters":101,"category":89,"deletable":8,"connectable":8},"node_if_automation","If: Automation vs judgment",[85,100],320,{"buttonId":71,"operator":88},{"id":103,"name":104,"type":83,"typeVersion":16,"position":105,"parameters":107,"category":89,"deletable":8,"connectable":8},"node_if_compare","If: Compare & attribution",[85,106],420,{"buttonId":74,"operator":88},{"id":109,"name":110,"type":83,"typeVersion":16,"position":111,"parameters":113,"category":89,"deletable":8,"connectable":8},"node_if_culture","If: Signal culture",[85,112],520,{"buttonId":77,"operator":88},{"id":115,"name":116,"type":117,"typeVersion":16,"position":118,"parameters":120,"category":51,"deletable":8,"connectable":8},"node_msg_branch_numbers","Playbook: Trustworthy branch numbers","text-message",[119,86],1090,{"text":121},"Branch numbers that deserve trust usually have three things: (1) a stable definition, (2) a believable path from reality → system, and (3) a known failure mode.\n\nQuick checks before you defend the number:\n- Definition: If two smart people describe it differently, it’s not decision-grade yet.\n- Coverage: Ask “what isn’t counted?” Missingness often hides in the ‘other’ bucket.\n- Lag: If it updates instantly but depends on humans, be suspicious.\n- Incentives: If a branch is rewarded on it, expect ‘creative compliance’.\n- Edge cases: Returns, reversals, rebookings—do they get quietly reclassified?\n\nRule of thumb: A metric that never embarrasses you is probably lying by omission.",{"id":123,"name":124,"type":117,"typeVersion":16,"position":125,"parameters":126,"category":51,"deletable":8,"connectable":8},"node_msg_dirty_signal","Playbook: Spot dirty signal",[119,94],{"text":127},"Dirty signal rarely looks dirty. It looks *confident*. Here’s how to catch it before the meeting builds a castle on sand:\n\nRed flags (fast to check):\n- Sudden step-change exactly at month/quarter boundaries\n- Perfectly smooth lines in a world that’s messy (real behavior is lumpy)\n- Big improvements with no operational story (no staffing, pricing, inventory, policy change)\n- One branch is ‘special’ every week (either a hero… or a tracking glitch)\n- Attribution that credits the last touch for everything (convenient, not true)\n\nMeeting-saving move: Ask “What would we expect to see if this were wrong?” If nobody can answer, you’re about to ship a confident mistake.",{"id":129,"name":130,"type":117,"typeVersion":16,"position":131,"parameters":132,"category":51,"deletable":8,"connectable":8},"node_msg_automation","Playbook: Automation vs human judgment",[119,100],{"text":133},"Automation is great at repetition and terrible at ambiguity wearing a nametag.\n\nTrust automation when:\n- The input is constrained (clear fields, consistent events)\n- The cost of a wrong call is low and reversible\n- You can monitor drift (simple checks, not a quarterly archaeology dig)\n\nKeep humans in the loop when:\n- Incentives exist to game the input\n- Edge cases carry real cost (compliance, customer harm, revenue recognition)\n- The signal depends on interpretation (conversation notes, reasons, dispositions)\n\nPractical guardrail: automate the *default*, and route the weird stuff to a person. Most systems fail by treating weird as normal until the dashboard looks ‘amazing’.",{"id":135,"name":136,"type":117,"typeVersion":16,"position":137,"parameters":138,"category":51,"deletable":8,"connectable":8},"node_msg_compare","Playbook: Comparing branches & attribution traps",[119,106],{"text":139},"Branch comparisons go wrong in predictable ways: you compare outcomes but ignore context, or you compare context but pretend it’s an excuse.\n\nCommon misreads:\n- Mix shift: One branch serves a different customer mix; averages hide it.\n- Volume effects: Small branches swing wildly; big ones look ‘stable’.\n- Operational capacity: Short staffing looks like demand drop—until it isn’t.\n- Attribution mirage: If the model gives 100% credit to the last visible touch, it’s storytelling, not measurement.\n\nTwo questions that save you:\n1) “What changed in inputs (leads, footfall, inventory, staffing) before we celebrate outputs?”\n2) “If we swapped processes between Branch A and B, would we expect the same result?” If not, you’re not comparing performance—you’re comparing environments.",{"id":141,"name":142,"type":117,"typeVersion":16,"position":143,"parameters":144,"category":51,"deletable":8,"connectable":8},"node_msg_culture","Playbook: Build a signal culture",[119,112],{"text":145},"A good signal culture produces decisions, not deck art.\n\nWhat actually works:\n- One owner per metric definition (not per dashboard tile)\n- ‘Known failure modes’ documented: how this metric breaks in the real world\n- A short pre-meeting ritual: definition → coverage → incentives → edge cases\n- Reward people for finding measurement flaws early (it’s cheaper than being wrong loudly)\n\nWitty-but-true rule: If people only ask about data quality after the decision, you don’t have a signal culture—you have a slide culture.",{"id":147,"name":148,"type":117,"typeVersion":16,"position":149,"parameters":151,"category":51,"deletable":8,"connectable":8},"node_msg_no_match","No match helper",[119,150],640,{"text":152},"I didn’t catch a menu selection. Pick one of the safety checks above—better a quick check now than a confident cleanup later.",{"id":154,"name":155,"type":55,"typeVersion":16,"position":156,"parameters":158,"category":51,"deletable":8,"connectable":8},"node_menu_next","Next step",[157,100],1350,{"messageType":159,"headerText":155,"bodyText":160,"footerText":79,"sectionTitle":79,"buttons":161,"ctaDisplayText":79,"ctaUrl":79},"button","Want to run another safety check, or should we route this to Insights for a human look?",[162,165],{"id":163,"title":164},"run_another_check","Run another check",{"id":166,"title":167},"handoff_to_insights","Insights review",{"id":169,"name":170,"type":83,"typeVersion":16,"position":171,"parameters":173,"category":89,"deletable":8,"connectable":8},"node_if_run_another","If: Run another",[172,45],1580,{"buttonId":163,"operator":88},{"id":175,"name":176,"type":83,"typeVersion":16,"position":177,"parameters":179,"category":89,"deletable":8,"connectable":8},"node_if_handoff","If: Handoff to Insights",[172,178],380,{"buttonId":166,"operator":88},{"id":181,"name":182,"type":117,"typeVersion":16,"position":183,"parameters":185,"category":51,"deletable":8,"connectable":8},"node_msg_restart","Run another: restart note",[184,45],1820,{"text":186},"To run another safety check, start the workflow again and pick a new option from the menu. (Yes, it’s one extra click. Still cheaper than one confident wrong decision.)",{"id":188,"name":189,"type":190,"typeVersion":16,"position":191,"parameters":192,"category":196,"deletable":8,"connectable":8},"node_fallback_insights","Route to Insights","fallback",[184,178],{"handoffMessage":193,"departmentId":194,"departmentName":195},"Routing to Insights for human review. Please share: (1) which branches/periods, (2) the exact metric names, (3) what decision you’re about to make, and (4) what would change your mind.","insights","Insights","terminal",{"id":198,"name":199,"type":117,"typeVersion":16,"position":200,"parameters":202,"category":51,"deletable":8,"connectable":8},"node_msg_next_no_match","Next step: No match",[184,201],480,{"text":203},"No problem—choose one of the buttons to continue. If it’s high-stakes or messy, “Insights review” is the grown-up move.",[205,210,212,214,217,220,222,224,226,228,230,232,234,236,238,240,242,244,246,248,250,252,254],{"id":206,"source":35,"target":41,"sourceHandle":207,"targetHandle":208,"type":209},"conn_input_to_kb","out","in","edge",{"id":211,"source":41,"target":53,"sourceHandle":207,"targetHandle":208,"type":209},"conn_kb_to_menu",{"id":213,"source":53,"target":81,"sourceHandle":207,"targetHandle":208,"type":209},"conn_menu_to_if1",{"id":215,"source":81,"target":115,"sourceHandle":216,"targetHandle":208,"type":209},"conn_if1_true_to_msg1","true",{"id":218,"source":81,"target":91,"sourceHandle":219,"targetHandle":208,"type":209},"conn_if1_false_to_if2","false",{"id":221,"source":91,"target":123,"sourceHandle":216,"targetHandle":208,"type":209},"conn_if2_true_to_msg2",{"id":223,"source":91,"target":97,"sourceHandle":219,"targetHandle":208,"type":209},"conn_if2_false_to_if3",{"id":225,"source":97,"target":129,"sourceHandle":216,"targetHandle":208,"type":209},"conn_if3_true_to_msg3",{"id":227,"source":97,"target":103,"sourceHandle":219,"targetHandle":208,"type":209},"conn_if3_false_to_if4",{"id":229,"source":103,"target":135,"sourceHandle":216,"targetHandle":208,"type":209},"conn_if4_true_to_msg4",{"id":231,"source":103,"target":109,"sourceHandle":219,"targetHandle":208,"type":209},"conn_if4_false_to_if5",{"id":233,"source":109,"target":141,"sourceHandle":216,"targetHandle":208,"type":209},"conn_if5_true_to_msg5",{"id":235,"source":109,"target":147,"sourceHandle":219,"targetHandle":208,"type":209},"conn_if5_false_to_nomatch",{"id":237,"source":115,"target":154,"sourceHandle":207,"targetHandle":208,"type":209},"conn_msg1_to_next",{"id":239,"source":123,"target":154,"sourceHandle":207,"targetHandle":208,"type":209},"conn_msg2_to_next",{"id":241,"source":129,"target":154,"sourceHandle":207,"targetHandle":208,"type":209},"conn_msg3_to_next",{"id":243,"source":135,"target":154,"sourceHandle":207,"targetHandle":208,"type":209},"conn_msg4_to_next",{"id":245,"source":141,"target":154,"sourceHandle":207,"targetHandle":208,"type":209},"conn_msg5_to_next",{"id":247,"source":154,"target":169,"sourceHandle":207,"targetHandle":208,"type":209},"conn_next_to_if_run",{"id":249,"source":169,"target":181,"sourceHandle":216,"targetHandle":208,"type":209},"conn_if_run_true_to_restart_msg",{"id":251,"source":169,"target":175,"sourceHandle":219,"targetHandle":208,"type":209},"conn_if_run_false_to_if_handoff",{"id":253,"source":175,"target":188,"sourceHandle":216,"targetHandle":208,"type":209},"conn_if_handoff_true_to_fallback",{"id":255,"source":175,"target":198,"sourceHandle":219,"targetHandle":208,"type":209},"conn_if_handoff_false_to_next_nomatch","automation",[28,29,30,258,259],"data-hygiene","automation-guardrails",[261,262],"Calypso Conversations","Calypso Knowledge Base","intermediate","Calypso","2026-05-08T11:04:28.138Z","/en/workflows/signal-decision-safety-checks",{"en":266},{"title":9,"description":269,"ogDescription":270,"twitterDescription":271,"canonicalPath":266,"robots":272,"schemaType":273,"alternates":274},"Coach teams to trust the right branch signals, spot dirty data early, and choose when automation needs human judgment—via guided chat prompts.","A practical chat workflow for signal hygiene: identify trustworthy branch numbers, catch dirty signals before meetings, and escalate to Insights when needed.","Turn messy branch signals into decision ready judgment: trust checks, dirty signal spotting, automation guardrails, and optional Insights handoff.","index,follow","HowTo",[275],{"hreflang":6,"href":266},1778614430324]