Home · Business · Sales · Prospecting & outreach

PARTLY

As of 13 August 2026, AI can only partly personalise cold emails for each prospect.

Most people should hand this to a purpose-built tool.

Can you do it?

5 minutesto a draft.

30 minutesto something you’d act on.

Cost, all in£0

Skill neededchat-fluent

Who has to check ityou

What the alternative costsApollo.io is a purpose-built alternative with a prospect database, AI outreach sequences and enrichment.

If this goes wrong: the email uses a false or intrusive detail, damages your credibility and wastes the prospect relationship.

What to actually do

  1. Use a tool built for this

    The route this page recommends

  2. Do it yourself

    Second choice

    A chat interface, chat-fluent skill, and roughly 30 minutes until you can act on the result.

    How to actually do it

    1. Open a spreadsheet or CRM view and make one row per prospect with the name, role, company, company website and the business trigger you can verify.
    2. Open each company website and the source for its trigger, then paste only relevant, current facts and links into the matching prospect row.
    3. Write a short description of your offer, the outcome it can credibly support, permitted evidence and the call to action you want to use.
    4. Paste the prompt into an AI chat and replace its bracketed fields with the information for one prospect, then ask it to produce the subject line, email and checking list.
    5. Compare every claim in the draft with the company pages and source links, removing any claim that the sources do not support.
    6. Read the email as the recipient, remove intrusive or irrelevant details, and check that the opening gives a real reason to contact this person rather than generic praise.
    7. Ask a colleague to check the final wording and targeting where the message could affect an important account, then send it through your approved outreach system.

    Prompt

    Personalise a cold email for this prospect using only the facts supplied below. Do not invent achievements, needs, company priorities, funding, technology, job changes or personal details. Do not infer sensitive characteristics. If a useful fact is missing or uncertain, say so and use a neutral alternative rather than guessing.

Prospect name: [name]
Prospect role: [role]
Company: [company]
Company website or source links: [links]
Relevant business trigger or recent change: [verified trigger]
What we offer: [product or service]
Specific outcome we can credibly support: [outcome]
Evidence or customer example we are allowed to mention: [evidence]
Call to action: [preferred next step]
Sender name and role: [sender]
Tone: plain, concise and professional British English.

Write one subject line and one email of no more than 120 words. Open with the prospect-specific reason for contacting them, explain the relevant problem or outcome without hype, and use a low-pressure call to action. Do not use generic praise, empty personalisation or claims that are not supported by the supplied facts. After the email, list each factual claim that needs checking and identify the source supplied above that supports it.

    Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.

  3. Hand it to a person

    The distant third

    A person who owns the outcome does this end to end, worth it when the failure is dear.

What it gets wrong

  • AI cannot access private buying context, internal priorities or the relationship history unless you supply them.
  • AI cannot tell whether a public fact is genuinely relevant to this particular recipient.
  • AI turns thin research into plausible-sounding filler unless you restrict it to verified sources.
  • AI cannot take responsibility for the targeting, compliance, sender reputation or relationship consequences of the campaign.

What caps this at PARTLY: private data access, judgement under ambiguity and stakes of error.

How we scored this

Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.

AxisScore (0–2)
Output2
Inputs1
Verification1
Liability1
Effort delta2
Total7 / 10

The methodology and its thresholds are published in full.

FAQ

Can AI write personalised cold emails?
Yes, it can produce a distinct draft for each prospect when you provide verified company, role and business-context information. The personalisation is only as good as that research, so check every factual claim before sending.
What information does AI need to personalise a cold email?
Give it the recipient's role, the company, a verified business trigger, the relevant problem, your credible outcome and the evidence you are allowed to mention. Do not give it permission to fill gaps with guesses or personal details.
Can AI personalise cold emails at scale?
Yes, outreach tools can generate variations across a prospect list, but scale does not make weak research reliable. Keep a human check for source accuracy, relevance, targeting and the final send.
Will AI-generated cold emails get better replies?
AI can make emails more specific and easier to produce, but it cannot establish that the recipient has the problem or wants contact. A generic claim, incorrect detail or intrusive reference can reduce trust instead.

Nearby answers

All Prospecting & outreach rulings

Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.

The newsletter

AI news, new answers and product picks, straight to your inbox.