The tension at the heart of cold outreach has always been this: the emails that generate replies are specific, researched, and demonstrably relevant to the individual. But writing a genuinely personalised email takes 15–30 minutes per prospect — which makes it incompatible with the volume requirements of most SaaS outbound motions.
The conventional resolution has been to compromise — to write semi-personalised templates with a few variable fields (company name, job title, perhaps a recent news item) and call it personalisation. Prospects are not fooled. The response rates on this kind of outreach reflect that.
AI resolves the tension at the root: it makes genuine personalisation — specific, researched, relevant to the individual — fast enough to apply at scale. Not by making templates slightly better, but by generating the specific research and the specific opening line that genuine personalisation requires, in seconds rather than minutes.
Genuine personalisation is not inserting the prospect's first name. It is demonstrating specific awareness of the prospect's context — their company, their role, their recent activity, or their specific situation — in a way that could not have been written for anyone else.
The three elements that separate genuine personalisation from template personalisation:
Specific opening line: The first sentence of the email references something specific and current about the prospect's company or their own recent activity. Something that required actual research, not a field merge. Relevant value connection: The connection between the product's value and the prospect's specific situation is drawn explicitly — not "we help companies like yours" but "given [specific thing you know about them], [specific capability] is likely directly relevant." Appropriate tone: The tone matches the prospect's apparent communication style — derived from their LinkedIn posts, their company's public voice, and any other observable signals.AI can generate all three elements from public data sources in seconds. What it cannot do is access data that doesn't exist publicly — which means the quality of AI personalisation is limited by the quality of the data sources it can read.
The richest personalisation source for most B2B prospects. AI reads:
Current role and tenure (new in role = evaluating tools; long tenure = knows the organisation deeply)
Recent posts and comments (their stated opinions on relevant topics)
Shared content (what they find interesting and relevant)
Career history (previous companies they've worked at, especially if they're your customers)
Educational background (shared connections or institutions)
Company LinkedIn pageRecent company posts and announcements
Headcount growth trend (rapidly growing companies have scaling challenges)
Recent job postings (the roles being hired signal strategic direction)
Leadership changes (new executive = evaluation of existing tools)
Company newsFunding announcements (new capital = investment in growth infrastructure)
Product launches (new product = new go-to-market motion)
Expansion news (new market = new team, new tools)
Awards and recognition (good for rapport; demonstrates awareness)
Leadership appointments (new CRO, new VP Sales = new tooling evaluation likely)
Intent and technographic dataTechnology stack data (what tools they currently use — from BuiltWith, Slintel, or Apollo's technographic data)
Intent signals (content they've been consuming that indicates active research in your category)
Job postings in relevant functions (SDR hiring indicates outbound investment; RevOps hiring indicates process investment)
Trigger eventsContract renewal timing for a competitor product
Recent conference attendance (LinkedIn check-ins or event registrations)
Published content (articles, podcast appearances, webinar presentations) on relevant topics
Before writing a word of outreach, build the prospect list with full enrichment — not just name, title, company, and email, but the data fields that enable personalisation: LinkedIn URL, recent post or activity, company news in last 90 days, technology stack, funding status, job posting count and type.
Apollo, Cognism, and LinkedIn Sales Navigator provide much of this at the list-building stage. Brazn's account intelligence layer enriches further with buying signal data and account context.
Step 2: AI generates the personalisation elementFor each prospect, AI reads the enrichment data and identifies the highest-quality personalisation trigger — the most recent, most relevant, most specific signal that can form the basis of an opening line:
A LinkedIn post the prospect shared last week on a topic directly relevant to your product's value
A funding announcement that creates a specific growth context
A job posting that signals a strategic initiative in your product's category
A conference talk or published article where they expressed a view on a relevant topic
The AI generates a specific, natural-sounding opening line from this trigger — not a template with a merge field, but a genuine sentence that could only have been written with knowledge of that specific signal.
Step 3: AI generates the body and CTAWith the opening line established, AI generates the rest of the email — the value connection, the social proof, and the call to action — tailored to the prospect's role, company stage, and the specific context established by the opening line.
The rep reviews and edits before sending. The edit should take 30 seconds — adjusting tone, adding a personal touch, or replacing AI language that doesn't sound natural. The rep is editing, not writing.
Step 4: Batch review and approvalModern AI outreach tools (Apollo AI, Outreach Amplify, Brazn's sequence integration) support batch review workflows: the rep reviews 20–30 AI-generated personalised emails in one session, approving or lightly editing each, rather than writing 20–30 emails from scratch. The time investment: 15–20 minutes for 30 personalised emails, compared to 7–10 hours for manually personalised equivalents.
AI personalisation requires human review — not because AI makes obvious errors (it usually doesn't) but because it occasionally makes subtle ones that a human would catch immediately:
Tone mismatch: AI generates an email that's too casual for a C-suite prospect or too formal for a startup founder. The rep adjusts. Relevance misread: AI identifies a trigger that is technically accurate but not actually relevant to the value proposition. The rep selects a different trigger or rewrites the connection. Factual error: AI misreads a job title change or references a company news item that is outdated or context-dependent. The rep catches it in review. Over-personalisation: AI generates an opening that references something too specific — crossing from "well-researched" to "slightly unsettling." The rep softens it.The review step is not optional. It is the quality gate that separates professional AI-assisted outreach from automation that damages reputation.
Build a personalisation quality framework with measurable metrics:
Personalisation rate: % of emails with a genuinely specific opening line (vs template insertion). Target: >90% with AI workflow. Open rate by personalisation level: Track open rates for AI-personalised vs template emails. The delta justifies the workflow investment. Reply rate by personalisation level: The primary metric. Genuinely personalised emails should produce 2–3× the reply rate of generic templates. Personalisation review time: Average time spent reviewing and approving AI-generated emails per rep per day. Target: <20 minutes for a 30-email daily volume. Edit rate: % of AI-generated emails that require significant editing (>50% rewrite). High edit rates indicate a prompt engineering or data quality problem that should be addressed upstream.Brazn's account intelligence layer provides the pre-call and pre-outreach briefing that drives the personalisation engine — aggregating company signals, stakeholder activity, and buying intent data into the structured input that AI needs to generate genuine, specific, relevant opening lines at scale. For SDR teams running high-volume outbound motions, Brazn's integration with sales engagement platforms means that personalisation happens as a natural step in the outreach workflow — not a separate, time-consuming research task that bottlenecks throughput.
---
####
Book a demo to see how Brazn AI fits into your sales stack.
About the Author
Alex Margarit, Sales AI Expert, SaaS Sales Leader, BMC, ServiceNow, Docusign — 25+ years in SaaS sales.