Updated September 28, 2026 | 11 min read

The Outbound Analyst Claude Skill: How I Benchmark Cold Email Against 244,000 Campaigns (Free .md File)

Here’s something I keep running into: a sales team sends 10,000 cold emails, gets a 3.8% reply rate, and has no idea whether that’s good or bad. They tweak the subject line. They swap the CTA. They burn through another list. Nothing changes, because they never diagnosed the actual problem.
That’s the job of an outbound analyst. And you can now run a version of that diagnosis inside Claude, for free, using a single .md file trained on data from 244,000+ real campaigns.
In this guide, I’ll show you exactly how this works: what an outbound analyst does, which metrics matter (and which ones lie), how to diagnose a broken campaign step by step, and how to use the Outbound Analyst Claude skill to do it faster. I’ll also walk through real prompts, setup instructions, and where lemlist fits once you know what to fix.

What an outbound analyst does in B2B sales

An outbound analyst reviews campaign performance data, compares it against real benchmarks, and pinpoints exactly where a sequence is breaking down, whether that’s deliverability, targeting, copy, or timing. From there, the job is to recommend a specific fix, not offer a vague guess.
Quick note: if you Google “outbound analyst,” you’ll find logistics roles, supply chain listings, and even Microsoft’s Copilot “Analyst” agent for Excel and Power BI. Those are data-science tools for spreadsheets. What I’m talking about here is different. In sales and RevOps, the outbound analyst function is purely diagnostic: you’re figuring out why your outbound results look the way they do, and what to change next.
The work usually covers four areas:
  • Reviewing send, open, reply, and conversion data across sequences to spot where engagement drops off
  • Comparing your numbers to real baselines, so you know whether a 4% reply rate is strong or a warning sign
  • Finding which stage of the funnel is broken and flagging it with specifics
  • Surfacing patterns for sales leaders and RevOps so decisions come from data, not instinct
Without this feedback loop, teams tend to burn through prospect lists with no real sense of what’s working. That’s the gap the function closes, and it’s also what a Claude skill can now do for you inside a chat window.

Key metrics every outbound analyst should track

Before you diagnose anything, you need to know which numbers actually matter. Some outbound metrics are misleading when read on their own, and open rate is the clearest example.

Email deliverability and bounce rate

Deliverability is whether your email reaches the inbox at all, versus landing in spam or bouncing outright. Deliverability issues rarely happen all at once. They build up gradually, and without a single source of truth, teams often notice problems too late, after reply rates drop. (Monitor deliverability with lemlist Deliverability Hub | lemlist Help Center) According to lemlist’s benchmark data, a bounce rate under 2% is good, under 1% is great, and under 0.5% is excellent. Anything higher usually points to unverified contacts or a stale list.

Open rate and its limitations

Open rate tracks how often people open your email, and it sounds useful on the surface. But since 2022, Apple Mail Privacy Protection has been pre-loading tracking pixels, which inflates open data without reflecting real engagement.
Where open rate still helps is in comparing subject line variants sent to the same audience. The absolute number might be inflated, but the relative difference between variants is often still informative.

Reply rate and positive reply rate

Reply rate counts every response, including “not interested” and out-of-office replies. Positive reply rate filters those out and only counts replies that move the conversation forward.
Per lemlist’s data, a positive reply rate above 5% is good, and above 8% is excellent. This is the number worth building your analysis around, since it reflects actual interest rather than inbox noise.

Meeting booked and opportunity conversion rate

A solid reply rate that never turns into meetings usually points to a weak CTA or poor qualification earlier in the sequence. If open rate is high but replies are low, look at your subject line and opening message. If replies are strong but meetings aren’t happening, the CTA is likely the problem.
Tracking these two together tells you whether you’re struggling to get attention, or struggling to convert the attention you already have.

Cold email benchmarks from real campaign data

Numbers only mean something in context, which is why a benchmark table is more useful than any single metric on its own.
Metric
Good
Great
Excellent
Positive reply rate
3–5%
5–8%
8%+
Meeting booking rate (of replies)
30–40%
40–60%
60%+
Bounce rate
<2%
<1%
<0.5%
Deliverability (inbox placement)
95%+
97%+
99%+
Source: lemlist help center benchmarks
Quote Icon
💡 Tip: lemlist recommends at least two weeks and 100+ sends before you draw any conclusions from these numbers. Anything less, and you’re reacting to noise.

How to diagnose an underperforming outbound campaign

Diagnosis works best in a specific order, because each layer depends on the one before it. Fix copy before deliverability, and you’re polishing a message that was never going to arrive.
  1. Check deliverability first. Verify SPF, DKIM, and DMARC, the three protocols that tell inbox providers your emails are legitimate. If any are misconfigured, your message can land in spam before anyone reads a word.
  2. Evaluate list quality and targeting. Are you reaching verified contacts that actually match your ICP? A reply rate under 3% almost always traces back to targeting, unless you’re working a small, high-value enterprise list where lower volume is expected.
  3. Audit your copy and CTA. If deliverability and targeting check out and replies are still low, the message itself is the problem. Common culprits: leading with who you are instead of a relevant pain point, or asking for a meeting before showing any value.
  4. Review send timing and cadence. According to lemlist’s analysis of millions of cold emails, overall reply rates increase with every follow-up, but after 9 or more follow-ups, the benefits are negligible and you could be marked as spam. (Prospecting Email Sequence for 2026 [+ Examples]) In cold sales outreach, you should be sending anywhere from 4–9 follow-up emails. (Prospecting Email Sequence for 2026 [+ Examples]) Past that point, returns flatten out and the spam risk starts to outweigh the benefit.
I’ll be honest: the first time I ran this diagnostic on a campaign with a 4.2% reply rate, I assumed the copy was the issue. It wasn’t. SPF was misconfigured on one of the sending domains, and half the emails were landing in spam. The fix took five minutes. The lesson took longer.

How the Outbound Analyst Claude Skill works

A Claude skill is a structured set of instructions that gives Claude specialized knowledge and a defined workflow for a specific task, instead of a general-purpose answer. Anthropic defines skills as “organized folders of instructions, scripts, and resources that agents can discover and load dynamically.” For this particular skill, that’s a single Markdown (.md) file. More complex skills can also bundle scripts and resources.
The Outbound Analyst skill is built specifically for the diagnostic process above. You feed it your campaign stats, and it compares them against benchmark data pulled from over 244,000 lemlist campaigns and 249 million emails. Instead of a vague “it depends,” you get a structured verdict tied to the exact metric that’s off, whether that’s a healthy result or something more critical that needs immediate attention.
Worth noting: this is an analytical layer only. It reads your data and diagnoses the problem. If you want Claude to actually execute inside lemlist (search leads, launch campaigns, pause sequences), that’s what lemlist MCP does. The skill diagnoses. MCP acts.
Quote Icon
💡 Tip: The skill is free to download from lemlist’s Claude Skills Hub. You can preview the full .md file before loading anything.

How to set up and use the Outbound Analyst skill

lemlist publishes companion skills in the same library, including a Copywriting First Touch framework and a Copywriting Analyzer, so you can stack a few together and cover more of the outbound workflow in one place.
  1. Download the skill file. It lives in the l3mpire/claude-skills GitHub repository. These are .md files you upload to a Claude conversation. They are not Claude Code skills that require npx, a terminal, or any developer setup. If you can attach a file to a chat message, you can use these skills. See lemlist’s official setup guide for the full walkthrough.
  2. Load it into Claude. For Claude Desktop, drag the .md file straight into a conversation, no terminal commands required.
  3. Attach your campaign data. Export stats from lemlist (or any outbound tool) as a CSV, or paste the numbers directly into the chat. More data generally means a sharper diagnosis.
  4. Ask your first question. Something direct works best: “Is my reply rate good?” or “What’s wrong with my campaign?” Claude reads the skill’s instructions and returns a structured answer.

Sample prompts for outbound campaign analysis

Once the skill is loaded, these prompts work as-is.
Benchmarking your reply rate — “My campaign has a 4.2% reply rate across 1,200 sends. Is that good? What’s the benchmark for B2B SaaS outbound?” Claude places your number against the benchmark table and flags whether your send volume is large enough to trust the result.
Diagnosing low open rates — “My open rate is 22% across two campaigns. My bounce rate is 1.3%. What’s going on?” Claude checks whether the open rate reflects Apple MPP deflation or a real engagement issue, and whether your bounce rate points to a list problem.
Comparing multichannel performance — “Here are my stats for email, LinkedIn, and calls across the same campaign. Which channel is performing best, and where am I losing the most pipeline?” Useful if you’re running a multichannel sequence and want to see where each channel adds or drops value.
Getting a full health report — “Run a full diagnostic on this campaign. Here’s my CSV export from lemlist.” Claude reviews every metric, flags what’s off, and ranks the issues by impact. Close to a one-click campaign audit.

Turn campaign data into better prospecting with lemlist

Knowing what’s broken is only half the job. Fixing it is where pipeline actually comes from.
The Outbound Analyst skill tells you what to fix. lemlist gives you the tools to fix it.

Over to you

If your outbound campaigns feel like guesswork, they probably are. The Outbound Analyst skill replaces gut feelings with a structured diagnosis built on real data from 244K+ campaigns, and it takes about two minutes to set up.
Download the skill, load it into Claude, paste your numbers, and see where your sequence actually stands. Once you know what’s broken, lemlist can help you fix it, from list quality to deliverability to multichannel execution. Trusted by teams who’ve left 1,300+ reviews on G2.
→ Start a 14-day free trial
(no credit card, cancel anytime)

FAQs about the Outbound Analyst Claude Skill

What data format does the skill accept? It works with CSV exports and with raw metrics pasted directly into the Claude conversation. Any outbound tool that lets you export campaign stats to a spreadsheet will work.
Can someone use it without a lemlist account? Yes. The skill file is free and works with data from any outbound platform. A lemlist account only matters if you want to use its built-in analytics export or enrichment tools.
How is this different from lemlist MCP? The skill gives Claude analytical knowledge: benchmarks and diagnostic logic. lemlist MCP goes further, letting Claude act directly inside lemlist, searching leads, launching campaigns, or pausing sequences. The skill diagnoses, MCP executes.
Does it work with both Claude Code and Claude Desktop? Yes. Drag the .md file into a Claude Desktop conversation, or load it in Claude Code. See the official setup guide for step-by-step instructions on both methods.
Share this post