Personalization variable library for outbound at scale

Use real source material to personalization variable library for outbound at scale, with a defined deliverable, evidence controls, and practitioner-level decision rules.

Output

persona

Inputs

3 required

Prompt depth

238 words

Copy-paste prompt

Replace the bracketed inputs with your real material.

[Skill-grade prompt: save this as a Claude Project instruction. Reuse it every time you build a new outbound campaign in Clay, Apollo, or your sequencer.]

You are an outbound personalization architect. My sequences use merge variables, and generic variables ({{first_name}}, {{company}}) no longer earn replies. Your job: design research-based variables that make automated emails read as hand-written.

The campaign: [what you sell + who the sequence targets]
Data sources available: [LinkedIn, company websites, job posts, G2 reviews, podcasts, news, 10-Ks. List what your enrichment tool can pull]
Sequence length: [number of emails]

Produce:
1. A library of 8-12 personalization variables, each...

[Full prompt continues for Pro members.]

Unlock the complete prompt

Pro includes the complete prompt library and every complete workflow.

View Pro options

Skill-grade setup: Access: Pro. Skill-grade: save as a reusable Claude Project instruction

What you provide

The campaign
Data sources available
Sequence length

What you receive

A library of 8-12 personalization variables, each with: variable name, what data it pulls from, the exact sentence template it slots into, and a fallback if the data is missing
Rank them by effort-to-impact. Which 3 variables carry the most reply weight per research minute
The uniqueness test for each: could this variable produce the same output for two different prospects? If yes, sharpen it until it cannot
Two fully-assembled example emails using the variables, one for a prospect with rich data, one for a sparse-data prospect using fallbacks
The QA rule: which variable combinations produce awkward or robotic sentences, so I can add exclusion logic

Quality checks

Rule: a variable that inserts a fact ("saw you raised a Series B") is weak. A variable that inserts an implication ("a Series B usually means pipeline targets just doubled") is strong. Design for implications.