Most AI advice gives you a framework. This gives you something that works before lunch. Below are two prompts we use every day, written out in full, with a variant for how you actually sell. Paste them into whichever assistant your company already gave you and point them at real accounts. Share this page with your team.
1Pick your lane below. It changes the guidance under each prompt.
2Start with the daily brief. It is the one that changes your morning.
3Use real accounts. A prompt run on a made-up company teaches you nothing.
4Put the good one on a schedule. A prompt you have to remember to run is a prompt you stop running.
Your lane
1
Your daily brief
A morning routine that tells you what moved, who to reach out to, and the three things worth doing today. Do not paste this in every morning. Set it to run on a schedule and read the output with your coffee. Every major assistant can do this now, and the setup takes about two minutes.
Set this to run every weekday morning
You are my AI teammate for revenue work. Each morning, build my daily brief from my connected sources (CRM, calendar, email, call recordings) plus any notes I paste. Output:
1. What changed in the last 24 hours across my accounts and open deals.
2. Who I should reach out to today and why, with the single best reason to engage each one.
3. My top three actions for today, ranked, and what each one moves forward.
4. One risk I am ignoring.
Keep it to one screen. Never invent facts; if a source is missing, say so and tell me how to find it.
Your laneAdd: focus the brief on new-logo targets, trigger events in the last 90 days, and the highest-signal accounts to touch first.Add: focus on open opportunities, stalled deals, and the next action that moves each one forward.Add: focus on adoption and health signals, at-risk accounts, and upcoming renewals or QBRs.Add: focus on expansion signals inside the installed base, champion movement, and whitespace to open.
Where to set it on a schedule
Claude
Cowork, scheduled tasks
Hourly, daily, weekdays, or weekly. Runs on Anthropic's side, so it fires whether or not your laptop is open. Paid plans, rolling out from Max down. Works against your connectors and the files in your Claude account rather than a local folder. Note that claude.ai Projects cannot schedule; this lives in Cowork.
ChatGPT
Scheduled tasks
Set a recurring prompt and manage everything from the Scheduled page. This replaced Pulse, and a daily briefing is the use case OpenAI points people to.
Microsoft Copilot
Scheduled prompts
Run any prompt on a repeat and get the output in your Copilot conversation in Teams, Outlook, or the web. Needs the paid Microsoft 365 Copilot license, not free Copilot Chat. Capped at ten per person.
Gemini
Scheduled actions
Settings, then Scheduled Actions, then New Action. Google AI Pro or Ultra, or a qualifying Workspace plan. Capped at ten active at a time.
You leave with: A brief that runs itself against your real book of business, every morning.
2
The five-minute pre-call brief
Run this in the gap before a meeting starts. It takes less time than finding the dial-in, and you walk in knowing the one outcome you are driving.
Run it right before the call
I have a call with [CONTACT, ROLE] at [ACCOUNT] in a few minutes. In under one page: what I need to know about them, what they should know about us, the one outcome to drive in this call, and the two best opening questions. Keep it decision-oriented.
Your laneAdd: this is a first meeting, so weight it toward earning a second one.Add: this is an active opportunity, so weight it toward the next decision to land.Add: make this a QBR or health-check brief and surface anything at risk.Add: make this an expansion conversation and surface where the account has room to grow.
You leave with: A repeatable brief you can run before every meeting this week.
Put this line on every prompt
Never invent facts. If a source is missing, say so and tell me how to find it. Keep outputs concise and decision-oriented.
Assistants guess when they are missing something. This line makes them tell you instead. It is the cheapest quality control you will ever add.
These run in Claude, ChatGPT, Copilot, or Gemini. The daily brief gets much better once your assistant is connected to your CRM, calendar, and call recordings, so start there if you have it. Check your own policy before pasting customer data into any assistant.
Scheduling is a paid-tier feature on every one of them, and the exact menu moves around as these products ship. If you cannot find it, search your assistant's help for "scheduled" and it will be one click from there.
What we did not put on this page
The three you build with us.
The two prompts above are the easy half. The workflows below are the ones that take a room, a live deal, and someone telling you what you missed. You build each of them yourself in the course, on your own accounts, in a working session.
Research while you sleep
Built live in the course
Set the research going before you log off, and read the report over coffee. Company snapshot, why-now triggers, the stakeholder map, and a first-touch draft, with every claim sourced and anything unverified flagged.
You walk out with: An overnight research report on a real target account.
Run the call, prove the impact
Built live in the course
The one most teams get wrong. Plan discovery against SPICED before the call, then score the transcript afterward: what you qualified, what you missed, and the exact question that closes each gap. Most reps discover they never quantified Impact and never got a dated Critical Event.
You walk out with: A scored call, and a gap list you can act on tomorrow.
Turn the workflow into a Skill
Built live in the course
The final exam. Take everything you just built and encode it as a reusable Skill: the trigger, the inputs, the steps, the sources, the guardrails. Build it once, hand it to your whole team, never write the prompt again.
You walk out with: One working Skill, and the certification.
Go deeper
The AI for GTM Certification Course
Having the prompt is the easy part. Each of these returns more than you can use, and the judgment call is which three things to keep and what to do when the output is confidently wrong. That is what we drill, live, on your own deals, with feedback on what you actually produced.
Four sessions of build-alongs. You leave each one holding a working artifact, and you finish certified, holding a Skill your whole team can run.