[{"data":1,"prerenderedAt":237},["ShallowReactive",2],{"/en/workflows/signal-research-decision-coach":3},{"id":4,"slug":5,"locale":6,"translationGroupId":7,"localeSwitchApproved":8,"title":9,"description":10,"documentationMarkdown":11,"workflowJson":12,"category":216,"tags":217,"integrations":221,"difficulty":224,"author":225,"verified":33,"featured":33,"date":226,"modified":226,"icon":7,"imageSrc":7,"path":227,"alternates":228,"seo":229},"816c7c9d-6876-44e2-8f34-c7eb7ebcdf99","signal-research-decision-coach","en",null,true,"Signal & Research Decision Coach","A guided decision-support flow that helps teams separate trustworthy branch signals from polished noise, spot dirty data early, and decide when automation is safe vs when judgment must step in.","## How it works\nThis workflow acts like a calm, experienced advisor sitting in the meeting with you—before the team confidently drives off a cliff with “clean-looking” numbers. It uses your Knowledge Base first (for your internal definitions and policies), then offers a short menu of decision-shaped prompts to quickly diagnose whether a branch metric, conversation trend, or attribution story deserves trust.\n\nOperators use it to triage messy evidence without “sanitizing away” the truth: you’ll get practical checks for dirty signal, comparisons that usually mislead, and guidance on when automation can be trusted—and when it needs a human in the loop.\n\n## Key features\n- Knowledge Base-first responses for your internal metric definitions, thresholds, and branch policies\n- Button-based menu that routes to the exact decision situation (trust, hygiene, automation, comparisons, culture)\n- Fast “dirty signal” checks designed to catch problems *before* the confident meeting\n- Clear guidance on when to accept automated outputs vs when to require human review\n- One-tap handoff path for cases that should go to a human analyst\n\n## Step-by-step\n1. **Input (Trigger):** A teammate asks a question or requests a sanity check on a branch number, conversation trend, or attribution claim.\n2. **Knowledge Base Policy:** If the question maps to known internal definitions (e.g., “What counts as a qualified conversation?”), the workflow answers from your Knowledge Base. If not, it continues to guided routing.\n3. **Interactive Menu:** The user selects what they need:\n   1) *Which branch numbers deserve trust?*\n   2) *Spot dirty signal before the meeting*\n   3) *Automation vs human judgment*\n   4) *Messy evidence → usable insight*\n   5) *Comparing branches & attribution pitfalls*\n   6) *Build a signal culture (less slides, more decisions)*\n   7) *Talk to a human analyst*\n4. **Decision Guidance (Routed Text):** The workflow returns a concise playbook tailored to the selected situation.\n5. **Human Handoff (Optional):** If the user chooses the analyst option, the workflow sends a handoff message and routes to the specified department.\n\n## Setup requirements\n- **Calypso Knowledge Base:** Recommended. Add (or confirm) entries for metric definitions, branch comparison rules, attribution rules, and escalation policies.\n- **Credentials:** None required for this workflow as configured.\n- **Optional routing:** Configure the target department for analyst handoff (name/id) if you want the “Talk to a human analyst” path to route correctly.",{"id":13,"teamId":14,"name":9,"version":15,"workflowVersion":16,"nodes":17,"connections":179,"routingEnabled":8,"active":33},"wf_signal_research_decision_coach_v1","calypso-public-library","1.0.0",1,[18,34,41,52,86,95,103,109,115,121,127,133,139,145,151,157,163,169],{"id":19,"name":20,"type":21,"typeVersion":16,"position":22,"parameters":25,"category":32,"deletable":33,"connectable":33},"fc1","Workflow settings","flow-configs",[23,24],-320,-40,{"name":9,"description":26,"tags":27,"triggerType":31},"Guided routing for signal trust, dirty data checks, automation guardrails, and branch comparison pitfalls.",[28,29,30],"signal-quality","decision-making","branch-metrics","input","policy",false,{"id":35,"name":36,"type":31,"typeVersion":16,"position":37,"parameters":40,"category":31,"deletable":33,"connectable":8},"in1","User message",[38,39],-120,120,{},{"id":42,"name":43,"type":44,"typeVersion":16,"position":45,"parameters":46,"category":51,"deletable":8,"connectable":8},"kb1","Knowledge Base answer (if available)","knowledge-base-policy",[39,39],{"enabled":8,"fallbackToRouting":8,"sticky":8,"stickyMode":47,"activationOpener":48,"personalization":50},"ai_sticky_release",{"enabled":8,"instruction":49},"If the user asks about definitions, thresholds, or policies, answer from the Knowledge Base. Otherwise, guide them with the decision menu and keep it practical (no academic tone).",{"useContactName":33},"response",{"id":53,"name":54,"type":55,"typeVersion":16,"position":56,"parameters":58,"category":51,"deletable":8,"connectable":8},"menu1","Choose a decision check","interactive-message",[57,39],380,{"messageType":59,"headerText":60,"bodyText":61,"footerText":62,"sectionTitle":63,"buttons":64,"ctaDisplayText":63,"ctaUrl":63},"button","Decision Signal Coach","Pick what you need. If a number looks *too tidy*, it’s usually hiding a footnote.","Tip: If you can't explain the change, pause celebration.","",[65,68,71,74,77,80,83],{"id":66,"title":67},"trust_numbers","Trust numbers",{"id":69,"title":70},"dirty_signal","Dirty signal check",{"id":72,"title":73},"auto_vs_human","Auto vs human",{"id":75,"title":76},"messy_to_insight","Messy to insight",{"id":78,"title":79},"compare_attrib","Compare & credit",{"id":81,"title":82},"signal_culture","Signal culture",{"id":84,"title":85},"talk_analyst","Talk to analyst",{"id":87,"name":88,"type":89,"typeVersion":16,"position":90,"parameters":92,"category":94,"deletable":8,"connectable":8},"if_trust","If: trust_numbers","if",[91,24],640,{"buttonId":66,"operator":93},"equals","routing",{"id":96,"name":97,"type":98,"typeVersion":16,"position":99,"parameters":101,"category":51,"deletable":8,"connectable":8},"msg_trust","Trustworthy numbers checklist","text-message",[100,24],900,{"text":102},"Here’s what usually *deserves trust* (and what’s polished noise):\n\n1) **Definition stability:** Same numerator/denominator as last week? If the definition moved, your trend is fiction.\n2) **Coverage, not just accuracy:** A ‘perfect’ metric that covers 40% of cases is a confidence trap.\n3) **Lag & revision behavior:** If it revises after the fact, treat today’s version as a draft.\n4) **Branch comparability:** Different staffing, hours, offer mix, or lead sources? Then you’re comparing *systems*, not branches.\n5) **Outlier audit:** Ask: “Which 10 records could flip this result?” If nobody can answer, you’re running on vibes.\n\nA reliable number is boring: consistent definition, consistent collection, and predictable failure modes.",{"id":104,"name":105,"type":89,"typeVersion":16,"position":106,"parameters":108,"category":94,"deletable":8,"connectable":8},"if_dirty","If: dirty_signal",[91,107],60,{"buttonId":69,"operator":93},{"id":110,"name":111,"type":98,"typeVersion":16,"position":112,"parameters":113,"category":51,"deletable":8,"connectable":8},"msg_dirty","Dirty signal early warnings",[100,107],{"text":114},"Dirty data rarely looks dirty. It looks *confident*. Try these fast tripwires before the meeting:\n\n- **Too-smooth line:** If variance disappears overnight, you might be seeing a logging change, not performance.\n- **Impossible timing:** Conversions happening at odd hours or in perfect batches = automation artifacts or backfills.\n- **Sudden mix shift:** Same total volume, different channel/branch mix? That’s often routing, tagging, or attribution drift.\n- **Duplicate fingerprints:** Repeated phone/email/device patterns can inflate “unique” counts.\n- **One field doing all the work:** If one tag, one rep, or one branch explains the whole movement—inspect that input.\n\nRule of thumb: if the story is clean, demand to see the mess it came from.",{"id":116,"name":117,"type":89,"typeVersion":16,"position":118,"parameters":120,"category":94,"deletable":8,"connectable":8},"if_auto","If: auto_vs_human",[91,119],160,{"buttonId":72,"operator":93},{"id":122,"name":123,"type":98,"typeVersion":16,"position":124,"parameters":125,"category":51,"deletable":8,"connectable":8},"msg_auto","When to trust automation",[100,119],{"text":126},"Automation is great at repetition. It’s terrible at noticing when reality changed.\n\n**Trust automation when:**\n- The inputs are instrumented (you can trace where the number came from).\n- The metric is stable and well-defined.\n- Errors are *bounded* (a bad run can’t quietly drive a big decision).\n- There’s a known review cadence (spot checks, thresholds, audits).\n\n**Require human judgment when:**\n- Incentives exist to game the metric (sales, branch rankings, bonuses).\n- The environment changed (new offer, new routing, staffing shifts, policy changes).\n- The decision is one-way (budget cuts, closures, major reallocations).\n\nPractical guardrail: automate the math, not the accountability.",{"id":128,"name":129,"type":89,"typeVersion":16,"position":130,"parameters":132,"category":94,"deletable":8,"connectable":8},"if_messy","If: messy_to_insight",[91,131],260,{"buttonId":75,"operator":93},{"id":134,"name":135,"type":98,"typeVersion":16,"position":136,"parameters":137,"category":51,"deletable":8,"connectable":8},"msg_messy","Messy evidence → usable insight",[100,131],{"text":138},"Turning messy evidence into insight isn’t “cleaning until it behaves.” Try this instead:\n\n1) **Separate ‘unknown’ from ‘no’.** Missing data is information—don’t zero it out and call it truth.\n2) **Keep a ‘reject bin’.** Track what you excluded and why. If exclusions rise, your process is drifting.\n3) **Slice by collection path.** Compare app vs manual entry, branch A vs branch B logging, new vs tenured reps.\n4) **Use ranges, not single points.** Decision-grade insight often looks like: “Likely 8–12%, worst-case 5%.”\n5) **Write the caveat as a decision rule.** Example: “If coverage \u003C 70%, treat trend as directional only.”\n\nThe goal isn’t a pristine dataset. It’s a decision you can defend after the fact.",{"id":140,"name":141,"type":89,"typeVersion":16,"position":142,"parameters":144,"category":94,"deletable":8,"connectable":8},"if_compare","If: compare_attrib",[91,143],360,{"buttonId":78,"operator":93},{"id":146,"name":147,"type":98,"typeVersion":16,"position":148,"parameters":149,"category":51,"deletable":8,"connectable":8},"msg_compare","Branch & attribution pitfalls",[100,143],{"text":150},"What teams routinely misread when comparing branches and attribution:\n\n- **Ranking ≠ learning.** A league table tempts you to copy tactics without checking if conditions match.\n- **Attribution is a policy, not a fact.** If the rule changes (last-touch, first-touch, assisted), the ‘winner’ changes too.\n- **Volume hides conversion truth.** A branch with better leads can look ‘better’ even if execution is worse.\n- **Conversation counts are not outcomes.** More conversations can mean lower quality leads, better logging, or a new script.\n- **Regression to the mean.** Last month’s hero branch often slides naturally—don’t punish them for statistics.\n\nBetter comparison question: “What changed in inputs, constraints, or definitions—before we claim performance changed?”",{"id":152,"name":153,"type":89,"typeVersion":16,"position":154,"parameters":156,"category":94,"deletable":8,"connectable":8},"if_culture","If: signal_culture",[91,155],460,{"buttonId":81,"operator":93},{"id":158,"name":159,"type":98,"typeVersion":16,"position":160,"parameters":161,"category":51,"deletable":8,"connectable":8},"msg_culture","Build a decision-helping signal culture",[100,155],{"text":162},"A healthy signal culture produces decisions—not prettier dashboards. A few habits that work:\n\n- **One owner per metric.** Not to ‘control’ it—so someone is accountable for definitions and breakage.\n- **Pre-meeting red flags.** Start with: what changed in tracking, routing, staffing, or incentives?\n- **Decision memos beat slide decks.** Require: metric definition, coverage, known biases, and what would change your mind.\n- **Celebrate corrections.** If people get punished for finding bad data, you’ll only hear good news.\n- **Keep a ‘known lies’ list.** Every org has numbers that look official but mislead. Name them.\n\nWit aside: the first thing that breaks is usually measurement. The second is everyone’s confidence in it.",{"id":164,"name":165,"type":89,"typeVersion":16,"position":166,"parameters":168,"category":94,"deletable":8,"connectable":8},"if_analyst","If: talk_analyst",[91,167],560,{"buttonId":84,"operator":93},{"id":170,"name":171,"type":172,"typeVersion":16,"position":173,"parameters":174,"category":178,"deletable":8,"connectable":8},"handoff1","Handoff to analyst","fallback",[100,167],{"handoffMessage":175,"departmentId":176,"departmentName":177},"Good call. This looks like it could benefit from human review. Please share: (1) which branch(es), (2) the metric name + time window, (3) what changed recently (routing, staffing, offer, tagging), and (4) the decision you’re about to make.","ops-analytics","Operations Analytics","terminal",[180,185,187,189,191,193,195,197,199,201,204,206,208,210,212,214],{"id":181,"source":35,"target":42,"sourceHandle":182,"targetHandle":183,"type":184},"c1","out","in","edge",{"id":186,"source":42,"target":53,"sourceHandle":182,"targetHandle":183,"type":184},"c2",{"id":188,"source":53,"target":87,"sourceHandle":182,"targetHandle":183,"type":184},"c3",{"id":190,"source":53,"target":104,"sourceHandle":182,"targetHandle":183,"type":184},"c4",{"id":192,"source":53,"target":116,"sourceHandle":182,"targetHandle":183,"type":184},"c5",{"id":194,"source":53,"target":128,"sourceHandle":182,"targetHandle":183,"type":184},"c6",{"id":196,"source":53,"target":140,"sourceHandle":182,"targetHandle":183,"type":184},"c7",{"id":198,"source":53,"target":152,"sourceHandle":182,"targetHandle":183,"type":184},"c8",{"id":200,"source":53,"target":164,"sourceHandle":182,"targetHandle":183,"type":184},"c9",{"id":202,"source":87,"target":96,"sourceHandle":203,"targetHandle":183,"type":184},"c10","true",{"id":205,"source":104,"target":110,"sourceHandle":203,"targetHandle":183,"type":184},"c11",{"id":207,"source":116,"target":122,"sourceHandle":203,"targetHandle":183,"type":184},"c12",{"id":209,"source":128,"target":134,"sourceHandle":203,"targetHandle":183,"type":184},"c13",{"id":211,"source":140,"target":146,"sourceHandle":203,"targetHandle":183,"type":184},"c14",{"id":213,"source":152,"target":158,"sourceHandle":203,"targetHandle":183,"type":184},"c15",{"id":215,"source":164,"target":170,"sourceHandle":203,"targetHandle":183,"type":184},"c16","automation",[28,30,29,218,219,220],"data-hygiene","automation-guardrails","attribution",[222,223],"Calypso Knowledge Base","WhatsApp","intermediate","Calypso","2026-05-29T11:04:05.663Z","/en/workflows/signal-research-decision-coach",{"en":227},{"title":9,"description":230,"ogDescription":231,"twitterDescription":232,"canonicalPath":227,"robots":233,"schemaType":234,"alternates":235},"Route teams to quick checks for trustworthy branch signals, dirty data detection, and automation guardrails—before bad numbers drive decisions.","A practical decision coach for branch metrics and messy evidence: trust checks, dirty signal detection, and guidance on when automation needs human judgment.","Turn messy branch signals into decision ready guidance: trust checks, dirty signal warnings, attribution pitfalls, and automation vs human judgment.","index,follow","HowTo",[236],{"hreflang":6,"href":227},1780761213302]