> ## Documentation Index > Fetch the complete documentation index at: https://docs.mavera.io/llms.txt > Use this file to discover all available pages before exploring further. # Email A/B Winner → Focus Group Why > Pull Mailchimp A/B test results and run a Mavera Focus Group to understand the psychological drivers behind why one subject line outperformed another ### Scenario Mailchimp ran an A/B test on subject lines (or content) and one variant won. You know *which* subject line had a higher open rate, but not *why*. You pull both variants and the performance results, then run a Mavera Focus Group asking personas which subject line they'd open and why — the psychological drivers behind the click. ### Architecture ```mermaid theme={"dark"} flowchart LR A["GET /campaigns/{id} A/B results"] --> B["Extract both variants + metrics"] --> C["POST /api/v1/focus-groups"] --> D["Why did one subject line win?"] ``` ### Code ```python Python theme={"dark"} import os, requests, time MC_KEY = os.environ["MAILCHIMP_API_KEY"] MC_DC = os.environ["MAILCHIMP_DC"] MV = os.environ["MAVERA_API_KEY"] MB = "https://app.mavera.io/api/v1" MH = {"Authorization": f"Bearer {MV}", "Content-Type": "application/json"} MC_BASE = f"https://{MC_DC}.api.mailchimp.com/3.0" mc_auth = ("anystring", MC_KEY) AB_CAMPAIGN_ID = os.environ.get("MAILCHIMP_AB_CAMPAIGN_ID", "") if AB_CAMPAIGN_ID: r = requests.get(f"{MC_BASE}/campaigns/{AB_CAMPAIGN_ID}", auth=mc_auth) r.raise_for_status() campaign = r.json() ab_settings = campaign.get("variate_settings", {}) else: ab_settings = { "subject_lines": [ "Your Q1 numbers are in — here's what they mean", "We analyzed your Q1 data (3 surprises inside)" ], "winner_criteria": "open_rate", } campaign = { "report_summary": {"open_rate": 0.32, "click_rate": 0.048}, } subject_a = ab_settings.get("subject_lines", ["Subject A", "Subject B"])[0] subject_b = ab_settings.get("subject_lines", ["Subject A", "Subject B"])[1] if len(ab_settings.get("subject_lines", [])) > 1 else "Subject B" report_r = requests.get(f"{MC_BASE}/reports/{AB_CAMPAIGN_ID}/ab", auth=mc_auth) if AB_CAMPAIGN_ID else None if report_r and report_r.status_code == 200: ab_report = report_r.json() a_stats = ab_report.get("a", {}) b_stats = ab_report.get("b", {}) else: a_stats = {"open_rate": 0.28, "click_rate": 0.038, "unique_opens": 1400} b_stats = {"open_rate": 0.36, "click_rate": 0.052, "unique_opens": 1800} winner = "B" if b_stats.get("open_rate", 0) > a_stats.get("open_rate", 0) else "A" PERSONA_IDS = os.environ.get("AB_PERSONA_IDS", "").split(",") if not PERSONA_IDS[0]: for name, desc in [ ("Newsletter Regular", "Opens emails 2-3x per week. Skims subject lines quickly. Values clarity over cleverness."), ("Busy Executive", "Gets 200+ emails/day. Opens only what seems urgent or directly relevant. Reads on mobile."), ("Curious Explorer", "Opens most marketing emails. Enjoys discovering new ideas. Clicks through to articles."), ]: p = requests.post(f"{MB}/personas", headers=MH, json={"name": name, "description": desc}).json() PERSONA_IDS.append(p["id"]) time.sleep(0.3) fg = requests.post(f"{MB}/focus-groups", headers=MH, json={ "name": "Mailchimp A/B Subject Line Investigation", "persona_ids": [pid for pid in PERSONA_IDS if pid], "questions": [ f'You see two emails in your inbox:\n A: "{subject_a}"\n B: "{subject_b}"\nWhich would you open first? Why? What about the winning subject line caught your attention?', f'Subject A got {a_stats.get("open_rate", 0):.0%} opens and Subject B got {b_stats.get("open_rate", 0):.0%}. Does the winner ({winner}) surprise you? Why do you think it performed better?', "What elements in a subject line make you most likely to open: (1) Personalization, (2) Curiosity gap, (3) Urgency, (4) Social proof, (5) Specificity (numbers/data)? Rank and explain your #1.", "Write a subject line that would outperform BOTH of these. Explain your strategy.", ], "context": f"""Mailchimp A/B test results: VARIANT A: "{subject_a}" Open rate: {a_stats.get('open_rate', 0):.0%} | Clicks: {a_stats.get('click_rate', 0):.0%} | Unique opens: {a_stats.get('unique_opens', 0):,} VARIANT B: "{subject_b}" Open rate: {b_stats.get('open_rate', 0):.0%} | Clicks: {b_stats.get('click_rate', 0):.0%} | Unique opens: {b_stats.get('unique_opens', 0):,} Winner: Variant {winner} (by {ab_settings.get('winner_criteria', 'open_rate')}) Margin: {abs(a_stats.get('open_rate', 0) - b_stats.get('open_rate', 0)):.0%} difference""", "responses_per_persona": 2, }).json() for _ in range(24): time.sleep(5) data = requests.get(f"{MB}/focus-groups/{fg['id']}", headers=MH).json() if data.get("status") == "completed": break print(f"Focus Group: {data.get('id')} — {data.get('status')}\n") for resp in data.get("responses", []): print(f"[{resp.get('persona_id','?')}] {resp.get('question','')[:70]}") print(f" → {resp.get('answer','')[:300]}\n") ``` ```javascript JavaScript theme={"dark"} const MC_KEY = process.env.MAILCHIMP_API_KEY; const MC_DC = process.env.MAILCHIMP_DC; const MV = process.env.MAVERA_API_KEY; const MB = "https://app.mavera.io/api/v1"; const MH = { Authorization: `Bearer ${MV}`, "Content-Type": "application/json" }; const MC_BASE = `https://${MC_DC}.api.mailchimp.com/3.0`; const mcAuth = "Basic " + Buffer.from(`anystring:${MC_KEY}`).toString("base64"); const AB_ID = process.env.MAILCHIMP_AB_CAMPAIGN_ID || ""; let subjectA = "Your Q1 numbers are in — here's what they mean"; let subjectB = "We analyzed your Q1 data (3 surprises inside)"; let aStats = { open_rate: 0.28, click_rate: 0.038, unique_opens: 1400 }; let bStats = { open_rate: 0.36, click_rate: 0.052, unique_opens: 1800 }; if (AB_ID) { const camp = await fetch(`${MC_BASE}/campaigns/${AB_ID}`, { headers: { Authorization: mcAuth } }).then((r) => r.json()); const subs = camp.variate_settings?.subject_lines || []; if (subs[0]) subjectA = subs[0]; if (subs[1]) subjectB = subs[1]; const abReport = await fetch(`${MC_BASE}/reports/${AB_ID}/ab`, { headers: { Authorization: mcAuth } }).then((r) => r.json()); if (abReport.a) aStats = abReport.a; if (abReport.b) bStats = abReport.b; } const winner = bStats.open_rate > aStats.open_rate ? "B" : "A"; const personaIds = []; for (const [name, desc] of [ ["Newsletter Regular", "Opens 2-3x/week. Skims subjects. Values clarity."], ["Busy Executive", "200+ emails/day. Opens only urgent/relevant. Mobile."], ["Curious Explorer", "Opens most marketing emails. Enjoys discovering."], ]) { const p = await fetch(`${MB}/personas`, { method: "POST", headers: MH, body: JSON.stringify({ name, description: desc }) }).then((r) => r.json()); personaIds.push(p.id); await new Promise((r) => setTimeout(r, 300)); } const fg = await fetch(`${MB}/focus-groups`, { method: "POST", headers: MH, body: JSON.stringify({ name: "Mailchimp A/B Investigation", persona_ids: personaIds, questions: [ `Two emails:\n A: "${subjectA}"\n B: "${subjectB}"\nWhich first? Why?`, `A: ${(aStats.open_rate * 100).toFixed(0)}% vs B: ${(bStats.open_rate * 100).toFixed(0)}%. Winner ${winner} surprise you?`, "Rank: (1) Personalization, (2) Curiosity gap, (3) Urgency, (4) Social proof, (5) Specificity. Explain #1.", "Write a subject line that outperforms both. Explain your strategy.", ], context: `A/B test:\nA: "${subjectA}" — ${(aStats.open_rate * 100).toFixed(0)}% open, ${(aStats.click_rate * 100).toFixed(0)}% click\nB: "${subjectB}" — ${(bStats.open_rate * 100).toFixed(0)}% open, ${(bStats.click_rate * 100).toFixed(0)}% click\nWinner: ${winner}`, responses_per_persona: 2, }), }).then((r) => r.json()); let data; for (let i = 0; i < 24; i++) { await new Promise((r) => setTimeout(r, 5000)); data = await fetch(`${MB}/focus-groups/${fg.id}`, { headers: MH }).then((r) => r.json()); if (data.status === "completed") break; } for (const resp of data.responses || []) { console.log(`[${resp.persona_id}] ${(resp.question || "").slice(0, 70)}`); console.log(` → ${(resp.answer || "").slice(0, 300)}\n`); } ``` ### Example Output ```text theme={"dark"} Focus Group: fg_mc_ab_5k8n — completed [Newsletter Regular] Which would you open first? → B: "We analyzed your Q1 data (3 surprises inside)". The number 3 tells me this is a quick read, and "surprises" creates a curiosity gap I can't ignore. Subject A feels like a report I should read — Subject B feels like a secret I want to know. [Busy Executive] Does the winner surprise you? → No. B has specificity (3 surprises) and implied exclusivity ("your data"). A sounds like a generic newsletter. When I'm scanning 200 subjects, the one that says "we did work FOR YOU" wins over "here's information." [Curious Explorer] Write a subject line that outperforms both → "Your Q1 had 3 anomalies — #2 changes your strategy." It combines personalization ("your"), specificity ("3"), curiosity ("#2 changes"), and stakes ("your strategy"). The incomplete list (naming #2 but not #1 or #3) creates irresistible curiosity. ``` ### Error Handling Mailchimp stores A/B test variants in `variate_settings.subject_lines` (for subject tests) or `variate_settings.contents` (for content tests). The winning criteria is in `variate_settings.winner_criteria`. The `/reports/{id}/ab` endpoint returns variant-level stats. Not all campaigns have this — only those created as "A/B Test" type. Regular campaigns return 404 on this endpoint. 3 personas × 4 questions × 2 responses = 24 total. Allow 60–120s. The loop provides \~120s. *** ## What's Next Back to Mailchimp integration overview Train brand voice from winning campaigns Generate content by audience interest Full reference for POST /api/v1/focus-groups