At a glance
WHAT IT’S REALLY ABOUT
Claude Cowork automates sales account research and post-call follow-ups
- The speaker explains that critical account intelligence is fragmented across systems like Salesforce, a data warehouse, call recordings, Slack, email, and the web, making manual prep slow.
- Claude Cowork lets her create a reusable “skill” (a plain-text instruction file) that defines the data sources, signals, and output format for account research.
- Running the “account strategy builder” skill concurrently pulls recent calls, spend and revenue trends, open opportunities, communications activity, and web/funding signals to generate a strategy document saved to a folder.
- After a customer call, a separate “call transcript processor” skill turns the transcript into personal action items, an internal Slack recap with owners, and a customer follow-up message, each requiring approval before sending.
- The workflow reduces pre- and post-meeting overhead from hours/30 minutes to minutes while improving conversation quality by arriving with concrete context and next steps.
IDEAS WORTH REMEMBERING
5 ideasCodify “what I need to know” into a reusable skill.
Instead of reinventing prep for every account, the AE describes her ideal meeting context and Claude drafts a skill file that standardizes sources, signals, and structure.
Parallel data pulls replace manual tab-hopping.
Cowork runs multiple retrievals at once—recent calls, pipeline in Salesforce, email/calendar/Slack activity, and web signals—cutting research time from hours to minutes.
Plain-text skill files increase transparency and iterative improvement.
Because the skill is just a text file, she can open it, verify behavior, and refine it after testing on a sample account.
Pre-meeting briefs shift the first call from orientation to strategy.
The generated doc includes spend, stakeholder mapping, inferred build/use cases, open deals, and risk signals, enabling more informed discovery and positioning.
Post-call automation improves speed and thoroughness without losing control.
The transcript processor generates action items, an internal Slack summary with owners, and a customer follow-up draft, all presented for approval before sending.
WORDS WORTH SAVING
5 quotesBefore I meet with a customer for the first time, I need to get smart on them fast.
— Brittany
Pulling this together used to mean hours of manual research, but with Claude Cowork, I have a skill that does it for me in minutes.
— Brittany
Claude drafted a skill file from that conversation, which is basically a text file that tells Claude how to approach the task, what data sources should it be looking at, what usage signals matter, and how to actually present the analysis.
— Brittany
Now I have the full landscape of the account before I've even talked to the customer, and it makes for a way more strategic first conversation.
— Brittany
With Claude Cowork, I'm pulling insights from multiple different systems and turning them into real account strategy.
— Brittany
High quality AI-generated summary created from speaker-labeled transcript.
