Enterprise AEs writing cold outreach face a persistent tension: the volume required to build pipeline demands speed, but the quality required to get replies demands specificity. Generic outreach at scale gets ignored. Specific outreach at scale used to be impossible because it required manual research per account. Claude Cowork for personalised outreach resolves this tension by making the research and the writing happen together, automatically, with the specificity quality gate built into the process.
This is part of the broader Claude Cowork for Account Executives deployment — the same platform that handles pre-call research and post-call CRM updates. The outreach workflow sits at the top of the funnel: the first touchpoint with accounts that don't know you yet, and the follow-up sequence with accounts that have gone dark. Both require genuine personalisation to get movement. Cowork provides it.
AEs using this workflow are writing 3.1× more personalised outreach messages per day compared to manual sequencing — not by sacrificing quality, but by compressing the input-to-output cycle from 25 minutes per account to under 6 minutes. If your organisation is deploying this at team scale, our Claude Cowork deployment service configures the outreach workflow with your ICP context and sequencing tool connectors as part of onboarding.
Why Most AI-Generated Outreach Fails
The failure mode of AI-generated outreach isn't that it's obviously AI. It's that it's generic. An AE who feeds Cowork "write a cold email to the VP of Sales at Acme Corp" will get an email about Acme Corp. An AE who feeds Cowork the CFO's earnings call quote about "compressing the sales cycle," the 3 revenue operations hires from last quarter, and the fact that a competitor just won a deal at their biggest account will get an email that reads as if someone did their homework.
That distinction — between account name substitution and account intelligence synthesis — is the entire difference. Cowork can do both. But it only produces genuinely personalised output when you give it genuine account intelligence as input. The workflow below is built around getting that input right.
The Specificity Input Framework
Before running any outreach prompt in Cowork, build a specificity input for each account. This doesn't require extensive research — it requires the right 3 elements:
The Hook
One specific, observable, recent fact about the account or prospect that connects to the pain you solve. Sources: earnings call quote, press release, LinkedIn post from the prospect, recent hire pattern, competitor win/loss, industry news. Not "they're in the [industry] sector" — that's not a hook. A hook is: "Their CFO said on the Q3 call that 'revenue predictability is our number one priority for 2026.'"
The Pain Connection
One sentence connecting the hook to the problem you solve. Not your solution — the problem. "That focus on revenue predictability suggests pipeline visibility is a current pain point" or "3 RevOps hires in Q1 typically signals they're rebuilding their data infrastructure — which is exactly where we help."
The Proof Point
One specific result from a comparable customer — same industry, similar role, similar pain, with a number. "We helped [similar company] reduce forecast variance by 34% in the first quarter." Cowork uses this to build credibility in the second touch without making the first email a case study dump.
What Good vs Bad AI Outreach Looks Like
Subject: Improving Sales Performance at Acme Corp
Hi Sarah,
I noticed you're the VP of Sales at Acme Corp. I wanted to reach out because we've been helping companies like yours improve their sales performance and close more deals.
Our platform helps sales teams be more efficient and productive, which leads to better results. I'd love to show you how we can help Acme Corp achieve similar results.
Would you be open to a 15-minute call?
Subject: Your Q3 note on revenue predictability
Hi Sarah,
Your CFO mentioned on the Q3 call that revenue predictability is the company's number one priority for 2026. Three RevOps hires in Q1 suggests you're already building the infrastructure for it.
We helped [COMPARABLE COMPANY] cut forecast variance from ±23% to ±8% in their first quarter using [PRODUCT]. Their head of RevOps wrote up the implementation if you'd find it useful.
Worth a 20-minute conversation?
The second email references a real earnings call quote, a real hiring observation, and a specific numeric proof point. None of that is fabricated. All of it was provided to Cowork as input. The writing itself — the structure, the concision, the framing — is what Cowork contributes. The specificity is what the AE contributes via the input framework above.
The Cowork Personalised Outreach Workflow
Compile your target list with specificity inputs
For each account on your list, add the 3-element specificity brief: hook, pain connection, proof point. This can be a simple table in the Cowork canvas or a CSV you paste in. For a list of 15-20 accounts, compiling these inputs takes 15-20 minutes total (1 minute per account if you've done pre-call research; see our pre-call research guide for the research step).
Set your messaging context and tone reference
Tell Cowork your ICP definition, your value proposition in one sentence, your tone (enterprise-formal, SaaS-direct, consultative, etc.), and your banned phrases. Upload 2-3 of your best-performing emails as style references. This context is set once and reused across all accounts in the session — or saved as a Cowork skill for your entire team.
Generate sequences for the full list
Run the batch outreach prompt (below) to generate 3-touch sequences for every account on your list in one pass. Review and approve each one — most require zero editing; some need minor tone adjustments. Export directly to Outreach or Salesloft via the sequencing tool connector, or paste into your email client. For the Outreach/Salesloft connector configuration, see our dedicated Cowork + Outreach/Salesloft integration guide.
Outreach Prompt Templates
Prompt 1 — Single Account 3-Touch SequenceOutput at Scale: AE Outreach ROI
For an AE who runs 3 prospecting blocks per week, the Cowork outreach workflow changes the output profile significantly. Manually, a disciplined AE writes 3-5 genuinely personalised emails per session. With Cowork, the same session produces 12-20 personalised sequences — and the quality is comparable because the specificity inputs are the same; only the writing time is eliminated.
At 14 hours of recovered time per week (our aggregate figure from deployed AE teams), outreach is the category that recovers the most time — typically 4-6 hours weekly from eliminating manual sequence writing. That time goes directly back to calls and pipeline development. Combined with the pre-call research savings covered in our research guide, the aggregate weekly recovery is enough to materially shift an AE's ratio of selling time to admin time.