# Email Metrics > - Diagnosing underperforming cold email sequences (reply rate < 3%) - Skill: `uitbreidenos/email-metrics` (Agent Skill) - Install (CLI): `npx skillmds@latest add uitbreidenos/email-metrics` - Raw SKILL.md: https://api.skillmd.com/api/skills/uitbreidenos/email-metrics/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Productivity - Author: UitbreidenOS (https://skillmd.com/u/uitbreidenos) - Updated: 2026-09-22 - Page: https://skillmd.com/skills/uitbreidenos/email-metrics --- # Email Metrics ## When to activate - Diagnosing underperforming cold email sequences (reply rate < 3%) - Optimizing open rates (target: 28-35%) - A/B testing email campaigns (change validation, statistical rigor) - Benchmarking your performance against 2026 verified standards - Deciding where to spend optimization effort: deliverability vs. subject line vs. body copy ## When NOT to use - Warm outreach (reply rates are contextually higher; different benchmarks apply) - Transactional email (welcome sequences, password resets) - Newsletter campaigns (open/reply metrics are not comparable) - Single send analysis (minimum 100 sends per variant required for statistical validity) - Questions about email list hygiene (use list-specific tools; this covers optimization) ## Instructions ### 2026 Verified Benchmarks (Instantly Benchmark Report) Use these as your reference frame for all campaign analysis: | Metric | Baseline | Top 10% | Signal-Based | Multi-Signal Stacked | |--------|----------|---------|--------------|----------------------| | **Reply Rate** | 3.43% | 10.7%+ | 5-18% | 12-25% | | **Open Rate** | 28-35% | 40%+ | 32-45% | 38-50% | | **Meeting Rate** (from positive replies) | 40-70% | 70%+ | 50-80% | 60-85% | | **Show Rate** | 70-85% | 85%+ | 75-90% | 80-95% | **Key insight:** Open rate is *deliverability dependent*. If your domain is blacklisted, sender reputation is poor, or SPF/DKIM/DMARC is broken, you'll see 10-15% opens even with excellent subject lines. This is a platform issue, not a copy issue. --- ### The 3 Leverage Points (In Order of Impact) #### 1. Deliverability (Can They Even Receive It?) **Priority:** Check this FIRST if open rate < 20% **Diagnostic questions:** - Is your domain on any blacklist? (Check: MXToolbox, SURBL, Spamhaus) - What's your sender reputation score? (Gmail Postmaster Tools, Microsoft SNDS) - Are you sending volume-ramped? (Warm up sending: 50 → 200 → 500 → 2000 emails/day) - Do you have SPF, DKIM, DMARC configured? (All three required for ISP trust) - Are you using shared IP or dedicated? (Shared IP = reputation bleed from other users) **Corrective actions:** - Request whitelisting from recipient domain (legal/compliance) - Switch to dedicated IP with warm-up protocol (3-week ramp minimum) - Implement DMARC enforcement (p=quarantine or p=reject) - Add List-Unsubscribe header (improves inbox placement) - Reduce sending volume temporarily; rebuild reputation **You'll know this is fixed when:** Open rate jumps 15-20% without any copy changes. --- #### 2. Open Rate (Do They Open It?) **Priority:** If open rate is 20-30%, fix this next **Diagnostic questions:** - Does your subject line create curiosity or urgency without being clickbait? - Is sender name recognizable? (First name + company, or familiar person?) - Are you sending at recipient's timezone peak hours? (9-11am and 4-5pm convert best) - Is the preview text cut off? (First 40 characters of body shouldn't repeat subject) - Are you split-testing subject lines? (Minimum 100 sends per variant) **Subject line principles:** - Curiosity gap: "This one change increased [metric] by 40%" (creates information asymmetry) - Specificity: "Reduced MTTR to 8 hours" beats "Performance improvement" - Social proof: "Used by Figma, Stripe, Notion" triggers recognition - Avoid: ALL CAPS, multiple ???, "Free," "Act Now," "Limited Time" (spam trigger words) **Sender name optimization:** - Test: First name only ("Sarah") vs. "Sarah Chen @ Salesloft" vs. "Sarah Chen" - Recognition matters: If recipient knows you, use just your name. Cold? Use company context. **Send time optimization:** - Default: 9-11am in recipient timezone (most opens) - Test: 4-5pm for post-work browsing (finance, operations teams show higher engagement) - Avoid: Before 8am, after 6pm, Sundays (low intent) **You'll know this is fixed when:** Open rate reaches 30%+ consistently across variants. --- #### 3. Reply Rate (Do They Respond?) **Priority:** If open rate > 30% but reply < 3%, fix this **Diagnostic questions:** - Is your email copy too long? (Over 150 words loses readers) - Is it specific to their use case? (Generic beats no value, specific beats generic 3:1) - Does your CTA require a commitment? (e.g., "Let's schedule 30 min" fails; "Quick question about your X" works) - Are you using personalization tokens without research? ("Hi [firstName]" is not enough) - Does the email answer the reader's implicit question: "Why are you emailing me?" **Email body structure (tested, high-reply template):** ``` [OPENER: Reference their recent action or recognizable context] "I noticed you just launched [product] on [date]..." "You're using [tool] for [outcome]..." [HOOK: One sentence — why this might matter] "Most companies using [tool] miss [X gap], which costs [Y]" [SOCIAL PROOF OR SPECIFICITY: One example] "We helped [similar company] reduce [metric] by X% using [approach]" [CTA: Low friction, specific, single action] "Quick question: is [specific challenge] on your roadmap? Happy to share how we've solved this for others." [CLOSER: Soft, no pressure] "If not, no worries — just reply 'pass' and I'll remove you." [Signature: First name + title + calendar link] "Sarah Chen Growth Ops @ Salesloft [Calendar link]" ``` **Length rule:** 80-120 words is the sweet spot. Every sentence must do work. **CTA principles:** - Avoid: "Let's jump on a call," "Schedule a 30-min," "Buy now" - Use: "Quick question about [specific thing]?" "Are you exploring [specific need]?" "Worth a 3-min call?" - Reply rate jumps when CTA requires 5 seconds of thought, not a calendar commitment **Personalization depth (escalates reply rate):** 1. Basic: "Hi [first name]" — doesn't increase reply. Skip. 2. Surface: "I noticed you're at [company] in [role]" — +10% vs. non-personalized 3. Research-backed: "Your Q1 earnings mentioned [specific goal]; we help teams like yours..." — +25-35% vs. baseline 4. Signal-stacked: Combine company data + recent news + technographics — +40-50% vs. baseline **You'll know this is fixed when:** Reply rate reaches 5%+ with consistent open rates > 30%. --- ### Diagnostic Decision Tree ``` START: Analyze your last 100-email sequence ├─ OPEN RATE < 20% │ ├─ YES → DELIVERABILITY PROBLEM │ │ ├─ Check: Spam score (< 5), domain reputation, blacklist status │ │ ├─ Action: Implement SPF/DKIM/DMARC, warm up IP, reduce volume │ │ ├─ Retest: Wait 5-7 days, re-send to 100 cold contacts │ │ └─ Success metric: Open rate jumps to 25%+ │ │ │ └─ NO → SUBJECT LINE / SEND TIME PROBLEM │ ├─ A/B test: 3 subject lines (curiosity vs. urgency vs. specificity) │ ├─ Test: Send time (9-11am vs. 4-5pm in recipient timezone) │ ├─ Min requirement: 100 sends per variant, 7-day observation window │ └─ Success metric: Best variant hits 28%+ open rate ├─ OPEN RATE 20-30% (Acceptable deliverability; room to optimize subject) │ ├─ Action: Iterate subject lines (retest top performer + 2 new variants) │ ├─ Adjust: Sender name recognition │ ├─ Min requirement: 100 sends, 7 days │ └─ Target: 30-35% open rate ├─ OPEN RATE 30%+ BUT REPLY < 3% (Copy problem) │ ├─ Diagnosis check: │ │ ├─ Is email > 150 words? YES → Shorten, reduce ideas to ONE │ │ ├─ Is CTA low-friction? NO → Replace with "Quick question..." │ │ ├─ Is it personalized beyond [first_name]? NO → Add 1-2 research details │ │ └─ Does it answer "why email me?"? NO → Add context opener │ │ │ ├─ A/B test: One change only │ │ ├─ Option A: Shorten body (120 words) + tighten CTA │ │ ├─ Option B: Add specific personalization detail + lower CTA friction │ │ ├─ Option C: Different opener (news-based vs. use-case-based) │ │ │ ├─ Min requirement: 100 sends per variant, 7 days │ └─ Success metric: Reply rate reaches 4-5% ├─ REPLY > 3% BUT NO MEETINGS (Discovery problem) │ ├─ Diagnosis: │ │ ├─ Are people saying "interesting but not right now"? │ │ │ └─ Solution: Add urgency signal or timeline specificity │ │ │ │ │ ├─ Are people saying "we're not looking"? │ │ │ └─ Solution: Tighten targeting (use technographics + intent signals) │ │ │ │ │ └─ Are people asking questions back? │ │ └─ Solution: Build a strong discovery email → proposal sequence │ │ │ ├─ CTA optimization: │ │ ├─ Avoid: "Let's chat about your needs" │ │ ├─ Use: "Are you exploring [specific tool/approach]? We just helped [similar company]" │ │ └─ Include: Specific value prop before asking for time │ │ │ └─ Success metric: 40-70% of positive replies convert to meetings └─ REPLY > 5%, MEETINGS > 40% (You're in top 10%) └─ Hold pattern. Optimize: Reply response time, meeting follow-up sequence. ``` --- ### A/B Testing Rules (Rigor) **Violation = invalid data:** 1. **One variable only:** Change subject line, hold body. OR change body, hold subject. Never change segment + copy + sender simultaneously. 2. **Minimum sample:** 100 sends per variant (minimum). 200+ preferred for clarity. 3. **Wait 7 days:** Reply rate plateaus after 5-7 days. Reading results on day 2 is false signal. 4. **Track:** Open time, reply time, reply quality (positive vs. objection vs. negative). 5. **Statistical confidence:** If 3 replies from 100 opens (3%), variance is high. At 10 replies (10%), variance is acceptable. **Never run:** - "Test everything in one email" (confounds all variables) - "Read results after 2 days" (early replies bias sample) - "Test with your warm list" (benchmarks are cold outreach only) - "Combine segment change + copy change" (can't isolate driver) --- ### Prompt for Diagnostic Review Use this when you're stuck analyzing a campaign: ``` Campaign: [name] Sends: [count] Open rate: [%] Reply rate: [%] Meeting rate: [%] Benchmark comparison: - Opens vs. 28-35% baseline: [+/- gap] - Replies vs. 3.43% baseline: [+/- gap] Likely problem: [deliverability / subject / copy / targeting / discovery] Recommended test: - Change: [one variable only] - Variant A: [specific change] - Variant B: [control or alternate approach] - Sample size: [100+ per variant] - Timeline: [7-day observation] - Success metric: [target benchmark] ``` --- ## Example **Scenario:** SaaS sales team, 200 cold emails/month, reply rate stuck at 1.8% (below 3.43% baseline) **Diagnosis process:** 1. **Check open rate:** 22% (below 28-35% baseline) - Deliverability: SPF/DKIM present, domain reputation score is 6/10 (weak) - **Action:** Check IP warm-up. Team was sending 500 emails/day on a 2-week-old IP. Rolled back to 100/day. 2. **Retest after 7 days:** Open rate improved to 29% (deliverability fixed) - But replies still at 2.1% - **Diagnosis:** Body copy problem, not deliverability 3. **Copy audit:** - Original email: 240 words (too long) - CTA: "Would love to schedule a 20-min call to discuss how we could support your goals" - Personalization: Only "[Company name]" token - **Problems identified:** Length, high-friction CTA, weak personalization 4. **A/B test (100 sends each, 7 days):** - **Variant A (Control):** Original 240-word email - **Variant B (Optimized):** ``` Hi [First Name], Saw you're hiring for 3 new data roles at [Company]. Building a data org is hard—most companies we work with spend 4 months getting their onboarding process right. We helped [competitor] cut this to 6 weeks using [specific framework]. Worth chatting? Sarah Salesloft [Calendar] ``` - 95 words, specific opener, low-friction CTA, social proof 5. **Results (day 7):** - Variant A: 3 replies from 100 (3%) - Variant B: 7 replies from 100 (7%) - **Decision:** Roll out Variant B; reply rate improvement: +4 points (up to 5.2% across book) 6. **Follow-up optimization:** - Variant B now the control; test 2 new subject lines to push open rate from 29% to 32% - Test discovery email sequence: "Which of these 3 approaches fit your timeline?" **Result:** Within 3 months, campaign went from 1.8% reply / 22% open to 5.2% reply / 31% open (now in top 25% of performers for this segment). **Key takeaway:** The problem wasn't the message, it was the platform. Once deliverability was fixed, copy optimization could actually work.