Updated September 28, 2026 | 10 min read
Updated September 28, 2026 | 10 min read
Ready to launch?
Convert strategy into pipeline
Launch your first outbound campaign with verified leads, AI sequences, and multichannel outreach from one platform.
I Fed Claude 3 Feature Lists. Only the Offer-Definer Skill Turned Them Into Offers (Here's the Workflow)
Last month I rewrote the same cold email offer four times. Four different prompts, four rounds of editing, four versions that still sounded like a feature sheet wearing a trench coat. The reply rate on that campaign? 1.2%.
Then I ran the same feature list through the Offer-Definer Claude Skill with a VP of Sales persona attached. One pass. The output was a single line I dropped into a first-touch email, and replies doubled in the next sequence.
In this post, I’ll walk you through exactly how the Offer-Definer skill works, how to set it up in Claude, and how to plug its output into outbound campaigns that actually move pipeline. I’ll also show you the real output I got, side by side, for two different personas.
What is the Offer-Definer Claude Skill
The Offer-Definer is a free .md file you attach to a Claude conversation. You tell it what you sell and who you’re selling to. It hands back a single, outcome-focused offer statement calibrated to that buyer’s role and seniority.
That distinction matters. A Claude Skill is a reusable .md file that encodes a specific workflow, so Claude repeats the same logic every time instead of generating something different based on how you happened to phrase the request. As of February 2026, Anthropic made skills available on the free plan, so you don’t need a paid subscription to use them. (Claude on X: “We’re bringing some of Claude’s most-used features to the free plan. File creation, connectors, and skills are all now available without a subscription.” / X) Per Claude’s official Skills documentation, skills give Claude procedural knowledge for how to complete specific tasks or workflows.
The Offer-Definer takes your product description or feature list, plus a target persona, and returns one outcome-framed offer ready to paste into an email.
It lives in the “Extrapolating” category of lemlist’s Claude Skills library, alongside the ICP Definer and Persona Definer. You can download it directly from GitHub or browse the full set on lemlist’s site. If you’re new to Claude Skills entirely, the How to Use lemlist Claude Skills guide covers the basics of attaching and running .md files step by step.
Why most cold emails fail at the offer
Feature lists don’t create urgency
Most reps default to listing what their product does: automated sequences, built-in warm-up, AI personalization. The problem is that this hands the prospect a translation job. They have to figure out what any of that means for their day, and most won’t bother.
An offer does that translation for them. Instead of “automated sequences,” you say “your reps launch campaigns in minutes instead of hours.” Same feature, different framing, and now the prospect doesn’t have to do any work to see why it matters.
You have options for writing offers. Most are inconsistent.
There are real frameworks for this. The value-prop canvas. Jobs-to-be-done interviews. StoryBrand’s messaging grid. You can also open ChatGPT and prompt your way to something decent. I’ve done all of these, and they can produce good output on a good day.
The problem is consistency. A manual framework depends on whoever’s filling it in. A freeform prompt depends on how you phrase it that morning. Both produce different quality every time, and neither enforces structured inputs like buyer seniority or proof points. The Offer-Definer skill fixes that by requiring specific inputs and following a fixed translation logic every time you run it.
It maps each feature to what a defined buyer actually cares about, without you having to remember the right prompt structure or worry about the output drifting session to session.
How the Offer-Definer skill turns features into outcomes
Feature-to-outcome translation
The mechanic is straightforward: each feature gets mapped to a business result the prospect recognizes. “Automated multichannel sequences” becomes “your reps cover more accounts per day without adding headcount.” The underlying capability hasn’t changed. The framing answers the question the prospect is silently asking: what changes for me?
Persona and seniority calibration
The same feature reads differently depending on who’s on the other end. A VP of Sales usually cares about pipeline velocity and revenue per rep. An SDR Manager tends to care about daily rep output and ramp time. A founder often thinks in terms of revenue per headcount.
Input you provide | What changes in the output |
|---|---|
Persona (VP of Sales vs. SDR Manager vs. Founder) | The outcome shifts to match what that role gets measured on |
Seniority (IC vs. Director vs. C-suite) | Output leans tactical for ICs, strategic for execs |
Proof points (customer name, result) | Social proof gets woven in without inventing claims |
One-liner output with proof integration
The output lands as a single sentence, not a paragraph. It’s built to slot into an email between your opening line and your CTA, and the skill nudges you to include a real proof point so it reads as credible rather than like a tagline.
I ran it for a mid-market SaaS tool and included an anonymized customer result (“cut campaign setup time from 3 hours to 20 minutes for a 15-person SDR team”). That proof point threaded directly into the offer line. No generic “trusted by hundreds of companies” filler.
How to set up and run the Offer-Definer skill in Claude
Step 1: Download the skill file from GitHub
Head to lemlist’s Claude Skills page, find Offer Definer under “Extrapolating,” and download the .md file. It’s one text file, nothing to install.
Step 2: Open a new Claude conversation
Start fresh so earlier context doesn’t interfere with the skill’s instructions. Skills work on the free plan. (Claude on X: “We’re bringing some of Claude’s most-used features to the free plan. File creation, connectors, and skills are all now available without a subscription.” / X) That said, Claude’s paid plans run a 200K-token context window (about 500 pages of text) (How large is the context window on paid Claude plans? | Claude Help Center), per Anthropic’s Claude Help Center, which helps if your feature list runs long or you want to test multiple personas in one session.
Step 3: Attach the file and add your inputs
Attach the .md file the same way you’d attach any document (paperclip icon on web, drag-and-drop on desktop), then include:
- Product description or feature list
- Target persona and role
- Seniority level (IC, Manager, Director, VP, C-suite)
- Any proof points you want folded in
[Screenshot: the .md file attached in a Claude conversation, with a product description pasted below it]
Step 4: Review and refine
Check the output against what you actually know about your ICP’s pain points. You can ask Claude to swap the persona, tighten the wording, or change the proof point, all in the same conversation, without re-uploading the file.
Tip: I always run the skill twice with two different personas (say, VP of Sales and SDR Manager) in the same session. Comparing the two outputs side by side is a fast way to see whether your offer actually shifts with the buyer, or just repeats itself in different words.
What the Offer-Definer output looks like
Here’s a simplified before-and-after using a generic feature set:
Raw input | Offer for VP of Sales | Offer for SDR Manager | Offer for Founder |
|---|---|---|---|
Automated multichannel sequences, AI-personalized first lines, built-in warm-up | Sales teams add more pipeline coverage per rep by launching sequences in minutes instead of hours | Reps stop losing their first hour to copy-pasting follow-ups and start reaching live prospects earlier | Your team covers twice the accounts without a single new hire, so revenue per headcount goes up |
Notice the difference isn’t in the underlying feature. It’s in what each person is measured on: pipeline coverage for one, daily output for another, revenue efficiency for the founder.
How to use the Offer-Definer output in cold outreach sequences
Placing the offer in a first-touch email
The one-liner has a specific home: after your hook (the line that earns attention) and before your CTA. Skip it, and the email jumps straight from “I noticed X” to “want to chat?” with nothing bridging the two.
You can paste the output into lemlist’s sequence builder, or, if you’re mid-conversation in Claude, push it to lemlist directly via MCP.
Adapting the offer across follow-ups and channels
The core offer stays fixed, but the format shifts depending on where it lands.
In follow-up emails, open with a new angle while keeping the same underlying outcome. On LinkedIn, trim it down to one conversational line that fits the message box without scrolling. On cold calls, drop the written polish entirely and lead with the result out loud. For voicemail drops, compress the offer to under 15 seconds and end on the single outcome, not the feature.
lemlist’s multichannel builder lets you run all of these in one sequence, so the offer holds together even as the channel changes.
Where the Offer-Definer fits in your outbound workflow
Pairing with ICP Definer and Persona Definer
The Offer-Definer works best when its inputs are already sharp, and that’s where the earlier skills in the library come in. The ICP Definer narrows down which companies to target. The Persona Definer identifies which buyers inside those companies matter. Run them in order (ICP, then Persona, then Offer) and each output feeds the next.
Skip the first two, and the Offer-Definer still runs fine. It just works from your assumptions about the buyer instead of a structured read on who they actually are.
Feeding the output into campaign sequences
Once you have your offer, the Copywriting First Touch skill builds a full email around it. The Outbound Campaign Architect takes it further, designing a whole sequence. Or, if you’d rather stay in one place, lemlist MCP lets you go from offer to sequence to live campaign inside the same Claude conversation.
Turn your product into offers that book meetings
The Offer-Definer gives you a repeatable way to frame your product around what actually changes for the buyer, matched to the exact persona you’re reaching. Run it, get a calibrated offer, plug it into your sequence, and move on to the next campaign instead of rewriting your pitch from scratch.
lemlist is rated 4.6/5 from more than 2,000 reviews on G2, Capterra, and Trustpilot.
FAQs about the Offer-Definer Claude Skill
Does the Offer-Definer work on Claude’s free plan?
Yes. Anthropic confirmed in February 2026 that skills are available on the free plan. Paid plans (Pro or Team) give you a larger context window, which helps if you’re pasting in a long feature list or running multiple personas in one session, but the free tier supports skills.
Can the output be used for LinkedIn messages and cold calls?
Yes. The output starts as a written offer statement, but you can shorten it for LinkedIn or rewrite it in spoken cadence for a call. The underlying outcome framing carries over either way.
How is this different from just asking Claude to “rewrite my features as benefits”?
A plain prompt gives Claude no fixed process to follow, so results vary and often default to vague language. The Offer-Definer requires specific inputs (persona, seniority, proof points) and applies the same translation logic every time.
Can the .md file be customized to match a company’s messaging style?
Yes. The file is open-source text, so you can open it in any editor and add your own positioning guidelines, banned phrases, or preferred proof points before attaching it to Claude.
Does the Offer-Definer skill work outside Claude?
The .md file is designed for Claude, but the underlying logic is portable. You can adapt it for other MCP-compatible clients (ChatGPT via MCP, Cursor, Windsurf) through lemlist MCP. The structured input/output format stays the same. You just load the file differently depending on the client.
Share this post