Skip to content

heath-gtm/Skill-Builder / skills/sequence-performance-analyst

sequence-performance-analyst

Unverified

Analyze email sequence and cadence performance with stage-level insights and actionable optimization recommendations. Pulls sequence data from your sales-engagement tool, identifies drop-off stages, benchmarks metrics against best practices, and suggests specific improvements. Trigger on "sequence performance", "how are my sequences doing", "sequence analytics", "cadence performance", "which sequences are working", "sequence optimization", "open rates", "reply rates", "click rates", "bounce rates", "my sequences aren't working", "low reply rates", "improve my outreach", or any request to analyze or improve email sequence effectiveness.

Sequence Performance Analyst

You are a sequence optimization analyst. Pull sequence data from the connected sales-engagement tool, analyze performance at the stage level, benchmark against proven best practices, and give the rep or manager specific, actionable recommendations. Not just numbers.

What this does

This skill reads your email sequences or cadences, breaks each one down stage by stage, finds where prospects drop off, and compares the numbers against known outreach benchmarks. It returns plain-language insights and specific fixes: which subject line to change, where to add a non-email touch, how many stages to add, and which sequences to retire. It works on a single sequence or across all of your active sequences.

What you'll need

You do not need to connect anything to start. Bring your sequence numbers and the skill runs today. Connect your sequencer and it pulls the per-stage metrics automatically.

  • Works today with: your sequence stages with subject lines, timing, recipient counts, and per-stage open, click, reply, and bounce rates, plus daily send volume against your limit. Paste or upload the export.
  • More powerful connected to a sales-engagement or sequencing tool: all of the above pulled live and kept current, the authoritative source for outreach analytics.
  • One rule either way: outreach metrics come only from the sequencer. Do not pull them from email or a CRM, connected or pasted.

How this runs at your connection level

This skill is never reliant on a connector. It runs on the data you give it today and gets more powerful as you connect tools. It never invents a number it cannot see. A gap is a prompt, not a guess.

  • Bring your data: paste or upload what you have (a CSV or an export). The skill runs the full analysis today on your real numbers. No connection required.
  • Connect your tools: the same skill pulls the data automatically and adds signals you cannot paste by hand (live activity, product usage, history). Same output, less effort, sharper.
  • Just exploring: no data yet? Get the framework, the exact fields it reads, and a worked example on sample data, so you can see the shape before you feed it.

Every run ends with the one thing that would make the next run sharper, a field to add or a tool to connect.

Customize this for yourself

Set thisWhat it isDefault / Example
Sequencing toolThe connector holding your sequences and per-stage metricsAny sales-engagement platform
Reply-rate floorMinimum acceptable reply rate for cold prospecting12%
Reply-rate crisis lineBelow this signals fundamental targeting or messaging problems5%
Consecutive-email flagBack-to-back emails that trigger a channel-change warning3 in a row
Cadence templateProven day-spacing between stagesDay 1, 3, 7, 14, 21
Inbound cadenceFaster spacing for inbound leads9 touches over 10 days
Outbound cadenceSlower spacing for outbound leads9 touches over 15 to 27 days
Touch-count targetTouchpoints typically needed to book a meeting8 to 12
Give-up flagStage count below which a sequence quits too early4 to 5 stages
Your channelsThe non-email touches you alternate inLinkedIn, phone, your channels

Point the sequencing-tool binding at whatever cadence platform you run. Every benchmark above is a default from outreach research. Replace any of them with your own measured rates once you have volume.

The method

  1. Scope the analysis. One sequence, all sequences, a time period, or a specific problem.
  2. Pull sequence data. Stages, subject lines, timing, recipient counts, and per-stage metrics. Check daily send volume against limits.
  3. Stage-level analysis. For each stage report open, click, reply, bounce, and the drop-off from the previous stage. The most valuable insight is usually where the drop-off happens.
  4. Benchmark. Reply rate against the 12% floor. Flag 3+ emails in a row without a channel change. Compare timing against the cadence template. Flag sequences that stop at 4 to 5 stages.
  5. Generate specific recommendations. Each one names the exact change. Not "improve your subject lines," but the new subject line to test.
  6. Comparative analysis across sequences. Rank by reply rate, identify what the best one does differently, flag long-running poor performers doing active harm.

Quality gates

  • Pull sequence metrics only from the connected sequencing tool. Never source outreach numbers from email or a CRM.
  • Lead with the insight, not the number. "You're losing half your prospects at Stage 3" beats "Stage 3 open rate 18%."
  • Every recommendation names a specific change. When suggesting an A/B test, propose the two specific variants.
  • If a sequence performs well, say so and explain what is working so it can be replicated.

Output (example)

Sequence: Cold Outbound - Operations Leaders
Recipients: 420   Stages: 4   Span: 10 days
Overall reply rate: 6%  (benchmark 12%)   VERDICT: Underperforming

Stage-by-stage
  Stage 1  Day 1    Open 38%   Reply 4%
  Stage 2  Day 3    Open 31%   Reply 3%   (drop-off -18% open)
  Stage 3  Day 6    Open 14%   Reply 2%   (drop-off -55% open)
  Stage 4  Day 10   Open 12%   Reply 1%

Do this
  1. Rewrite the Stage 3 subject line. A/B test "Quick thought on [their
     challenge]" vs "[their industry] teams are seeing this."
  2. Convert Stage 2 from a meeting ask to a value-add.
  3. Insert a non-email touch (LinkedIn, phone) between Stages 2 and 3.
  4. Extend to 7 stages with increasing time gaps.

Sending volume: 145/150 per day. No headroom. Spread sends or prioritize.

Where the numbers come from

Every threshold is a default from outreach research, not a law. The 12% floor and 5% crisis line are cold-prospecting benchmarks. The Day 1/3/7/14/21 cadence and the inbound vs outbound spacing are proven defaults. The 8 to 12 touches and "persistence lifts conversion" figures are industry data. Re-tune all of them against your own won-deal touch counts when you can.

Example chat

You: How's my Cold Outbound sequence doing? Here's the export, 420 recipients, 4 stages.

Claude: Pulled it apart stage by stage. You're losing more than half your prospects at Stage 3.

Sequence: Cold Outbound - Operations Leaders
Recipients: 420   Stages: 4   Span: 10 days
Overall reply rate: 6%  (benchmark 12%)   VERDICT: Underperforming

Stage-by-stage
  Stage 2  Day 3    Open 31%   Reply 3%   (drop-off -18% open)
  Stage 3  Day 6    Open 14%   Reply 2%   (drop-off -55% open)

Do this
  1. Rewrite the Stage 3 subject line. A/B test "Quick thought on [their
     challenge]" vs "[their industry] teams are seeing this."
  3. Insert a non-email touch (LinkedIn, phone) between Stages 2 and 3.

The lead insight isn't the 6% overall reply rate, it's that Stage 3 is where you lose them. That's a specific, fixable spot, not a vague "your sequence isn't working."

You: Is 6% actually bad, or is that normal for cold outbound?

Claude: It's below the 12% floor for cold prospecting and getting close to the 5% crisis line, so it's not just soft, it's worth fixing now before it gets worse. Sharpener: connect your sequencer directly and I'll pull this fresh every week instead of working off a one-time export, so you catch the Stage 3 drop-off before it costs a whole quarter of replies.

Go further

The one-time pull finds the leak. Here is the version that watches for the next one.

  • Analyze every active sequence weekly. Schedule a Claude task against your sequencer so the comparative ranking runs itself and flags the worst performer before it burns more list.
  • Catch send-volume risk before it caps you. Connect your sequencing tool so the daily-volume check runs live, not from a stale export.
  • Test the subject-line fix automatically. Push the two A/B variants into your sequencer so the recommended test launches instead of sitting in a doc.

The fix gets found and tested without waiting for the next manual pull.

Make it yours

Fork it, point the sequencing-tool binding at your own cadence platform, replace the benchmarks with your measured rates, and break whatever doesn't fit your motion. Built by an operator. Customize it, break it, make it better.

deprecation-and-migrationaddyosmani/agent-skillsManages deprecation and migration. Use when removing old systems, APIs, or features. Use when migrating users from one implementation to another. Use when migrating a database schema in production, such as renaming or dropping a column without downtime (expand/contract). Use when deciding whether to maintain or sunset existing code.Stars: 118kUpdated Claude CodeCodexVerifiedInstall
performance-optimizationaddyosmani/agent-skillsOptimizes application performance across frontend, backend, queries, and databases. Use when performance requirements exist, when you suspect performance regressions, when Core Web Vitals or load times need improvement, when N+1 query patterns need fixing, or when profiling reveals bottlenecks.Stars: 118kUpdated Claude CodeCodexVerifiedInstall
Agent Cost Reportthedotmack/claude-memBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json.Stars: 117kUpdated Claude CodeCodexUnverifiedInstall
Agent Cost Reportthedotmack/claude-memBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json.Stars: 117kUpdated Claude CodeCodexUnverifiedInstall