Episode 03 | AI for Revenue Leaders: The AI Hat Podcast
Hosted by Mike Allton
LISTEN: AI for Revenue Leaders: The AI Hat Podcast
WATCH: AI for Revenue Leaders: The AI Hat Podcast
The Parking Lot Panic
Picture this scene. It happens in every sales org, every single day.
It's 1:55 PM on a Tuesday. Your sales rep - let's call her Jennifer - has a demo scheduled with a VP at a major prospect at 2:00 PM.
Where is she right now?
She's sitting in her car in the grocery store parking lot. Or she's staring at a Zoom waiting room screen with her coffee going cold.
And what is she doing in these final 5 minutes?
She's frantically thumbing through LinkedIn on her phone. Scrolling the prospect's profile. Trying to find something - anything - to open with that isn't "So, how's the weather in Denver?" or "I see you went to Ohio State. Go Buckeyes!"
She's hunting for a hook. A piece of context that makes her sound like she did her homework.
But here's the problem. Jennifer's been in back-to-back calls since 9 AM. Demo at 9. Discovery at 10:30. Internal pipeline review at noon. Client check-in at 1. She literally has not had 10 minutes to breathe, let alone research.
So she's going into this call cold.
And you know what happens when you go in cold? You ask the same generic questions every other vendor asks. "What are your biggest challenges right now?" "What does success look like for you?"
And you get generic answers. Polite answers. The kind of answers prospects give when they're already mentally writing you off.
That rep isn't lazy. She's brilliant. You hired her because she's got the emotional intelligence to navigate complex enterprise sales and the persistence to close six-figure deals.
But right now? She's drowning in Admin Drag - the villain we diagnosed back in Episode 1.
Now let me show you what this looks like when you've hired a Digital Crew member to handle the robot work.
The Alternate Timeline: What Happens When AI Does the Research
It's 1:55 PM on a Tuesday. Same rep. Same meeting in 5 minutes.
Jennifer's phone buzzes. It's a notification from her Pre-Flight Agent.
She taps it. Reads for 30 seconds.
The message says:
"Heads up - you're meeting with Sarah Chen in 5 minutes. She's the VP of Marketing at HealthTech Solutions. Here's what you need to know:
1. She posted on LinkedIn yesterday about launching a new sustainability initiative. Ask her about that as your icebreaker - she's clearly passionate about it based on the 47 comments.
2. Her company just announced a hiring freeze last Thursday (press release on their site). She's probably stressed about doing more with less. This is your opening to pivot to our 'Digital Crew' case study from the healthcare vertical - the one where we helped a team automate 40% of their reporting workflow without adding headcount.
3. Her CFO is ex-McKinsey (LinkedIn profile). Lead with ROI data and efficiency metrics, not features. She's going to filter everything you say through a 'will this make my CFO happy?' lens.
Good luck. You've got this."
Jennifer walks into that Zoom room and opens with: "Sarah, I saw your post yesterday about the sustainability initiative. That's ambitious - especially with everything else on your plate. How's that project tracking so far?"
Sarah's face lights up. Because nobody - and I mean nobody - has asked her about that. Every other vendor this week opened with "So tell me about your business."
Jennifer isn't reciting a script. She's asking an empathetic, genuinely curious question based on real context.
And 15 minutes into the call, when Sarah mentions the hiring freeze and says, "We need to hit our Q2 goals but I can't add headcount," Jennifer doesn't scramble. She says, "We actually worked with a healthcare company in a similar spot last year. Let me show you exactly how they scaled output by 40% without hiring a single person."
Boom.
Jennifer didn't spend 30 minutes researching. She spent 30 seconds reading.
And here's what's beautiful about this - she sounds more human, not less.
Because she's not robotic. She's prepared. She's empathetic. She's asking the kind of questions that prove she actually gives a damn about solving Sarah's problem.
Remember what I said in Episode 1? In a world flooded with AI-generated spam - where every inbox is full of robotic outreach that sounds exactly the same - genuine human connection is the only competitive advantage that matters.
This agent doesn't replace that. It enables it.
It removes the robot work (research, data gathering, LinkedIn stalking) so your rep can focus on the human work. Listening. Building trust. Closing.
That's Revenue Architecture.
From Theory to Execution: The Launch Trilogy Complete
If you've been following this trilogy, you know where we've been:
Episode 1: We diagnosed Admin Drag - the 70% of non-revenue work that's burying your sales team and turning thoroughbreds into data entry clerks.
Episode 2: We tackled Shadow AI - the fact that 90% of your employees are already using unsanctioned AI tools in secret, and why the "ban hammer" just drives the behavior underground where you have zero visibility.
Today, in Episode 3, we execute.
We're going to build The Pre-Flight Agent together. A sales prep agent. It takes about 20 minutes to set up. Costs almost nothing. And it will fundamentally change how your reps show up to every sales conversation.
But first, let me address the objection I know you're thinking.
"But Mike, I Haven't Even Rolled Out Governance Yet…"
Some of you are sitting there thinking: "Mike, I haven't even rolled out the governance framework from Episode 2 yet. I'm still trying to convince my boss that AI is safe. Legal is reviewing the policy. IT wants to audit the tools. And now you want me to just… build an agent? Without permission?"
Look, here's the thing.
The best way to get buy-in for AI governance isn't a 40-slide PowerPoint deck titled "AI Strategy Roadmap Q1 2026."
It's a working prototype.
When you walk into your VP's office on Friday afternoon and say, "I built this over the weekend. It saved me 4 hours this week. Here's the exact data it pulled, here's how it works, and here's why it's completely safe" - that's when the conversation shifts.
You're not asking for permission anymore. You're showing results.
You're not pitching a pilot. You're demonstrating ROI.
And here's the key - here's why you can build this right now without violating any of the governance principles we laid out in Episode 2:
We're building this using only public data.
LinkedIn profiles. Company news pages. Press releases. Industry blogs.
That's all Green Light data from our Traffic Light system. No customer PII. No confidential contract details. No proprietary financials.
Zero risk.
This is your proof of concept. Your trojan horse. Your "see, this actually works" moment.
So let's build it.
Agentic AI vs. Automation: Why This Is the Shift That Changes Everything
Before we dive into the build, I need to explain something that most people - including a lot of AI vendors - get completely wrong.
This isn't just automation.
This is agentic AI.
And the difference matters.
Let me use a historical analogy to make this clear.
Think about Henry Ford's assembly line in 1913. Revolutionary at the time, right? Every worker stood in one spot and repeated the same motion over and over. Insert bolt. Tighten bolt. Insert bolt. Tighten bolt. No thinking required. No decisions to make. Just follow the script.
That's automation.
Now walk into a Tesla factory today. The robots aren't just repeating motions. They've got sensors. When a sensor detects that a part is slightly misaligned, the robotic arm doesn't just keep going and screw it up. It adapts. It adjusts. It makes a decision in real time.
That's agency.
That's what we're building here.
Automation follows a rigid script. You tell it: "First, check LinkedIn. Then, check Google News. Then, check Twitter. Then, summarize everything in this exact format."
Agentic AI uses discretion. You tell it: "Here's the person I'm meeting tomorrow. Go find out what's happening in their world right now. Decide which sources are relevant. Use your judgment to determine what I actually need to know vs. what's just noise."
See the difference?
With automation, you're micromanaging a robot.
With agentic AI, you're delegating to a teammate who thinks.
And that's the shift from tools that do what you tell them, to systems that figure out what needs to be done.
That's the future we're building toward.
The Architecture: How to Build The Pre-Flight Agent
Alright, enough philosophy. Let's get tactical.
I'm going to give you the exact blueprint I used to build my own Pre-Flight Agent. You can build this in a few different ways - Zapier, Make, even with Custom GPTs if you want to go that route - but I'm walking you through the Zapier version because it connects to everything and it's what I actually use.
Quick Context on Zapier (Because This Trips People Up)
Most people know Zapier for "Zaps" - those simple automations like "when I get an email with an attachment, save it to Google Drive.
That's great for basic stuff. I use Zaps all the time.
But what we're building today uses Zapier Agents - which is a completely different product. Different billing. Different capabilities. Different interface.
This is where Zapier's AI actually thinks and makes decisions for you.
And here's the good news - Zapier has a library of pre-built agents you can browse and customize. You don't have to start from scratch. You can grab one that's close to what you need, tweak it, and you're done.
Or - and this is what I did - you can just start with a blank agent, type in a plain English description of what you want, and let Zapier's AI build the bones for you.
I literally typed: "Research my upcoming meetings with external contacts and send me a daily briefing at 6 AM with key insights about each person."
Hit enter.
Zapier built about 80% of what I needed.
The other 20%? I had to fix a couple things. Like, it defaulted to pulling meetings from my personal Google Calendar instead of my work calendar. Took me 3 minutes to switch it to the right account. That's it.
That's the kind of "bug" you'll hit - not coding errors, just "oh, wrong dropdown selected." Easy fixes.
So here's how any agent works.
There are three core components:
- The Trigger - When do I work?
- The Brain - What do I think about?
- The Action - Where do I deliver the results?
Let me break down each one.
Component 1: The Trigger (When Do I Work?)
First, we set up a trigger on your calendar. Google Calendar, Outlook, whatever you use.
You've got two options here, and both work - it just depends on how you like to operate:
Option A: The Morning Briefing
"Every morning at 6:00 AM, look at my calendar for today. Find all the meetings scheduled with external email addresses (not internal team meetings). Pull the attendee info."
Option B: The Just-In-Time Alert
"15 minutes before every meeting on my calendar, check if it's an external contact. If yes, run the research and send me the briefing."
I prefer Option A because I like getting my Daily Briefing over coffee. I wake up, pour my coffee, open my email, and boom - there's a summary of every external meeting I have today and exactly what I need to know about each person.
But if you're more reactive - if you like things served up right when you need them - Option B works great too.
The key instruction either way is: "Ignore internal meetings."
You don't need a research briefing on your 1-on-1 with your manager or your weekly team standup. Focus the agent's energy on customer-facing calls, prospect demos, partner meetings - the conversations where context actually matters.
Component 2: The Research (What Sources Does It Check?)
Alright, now here's where this gets fun.
Once your agent spots an external meeting on your calendar, we're going to give it a job. Not a rigid script - a job.
We say:
"Look, here's this person I'm about to meet. Go find out what's happening in their world right now. Check their LinkedIn profile. See if their company published any recent news or press releases. Look for blog posts they've written. Find anything that gives me context about what they're working on or stressed about."
Now here's the critical distinction - and this is what makes this agentic instead of just automated:
You're not telling it "First go to LinkedIn, then go to Google News, then check Twitter, then come back."
You're giving the AI discretion.
You're saying: "Here's the person. Here's the goal. You decide which sources are worth checking. You determine what's relevant vs. what's noise. Use your judgment."
That's agency.
The AI is making decisions about where to look and what matters - not just executing a pre-programmed sequence.
This is the shift from "do exactly what I say" to "figure out what I need."
And honestly? It's kind of wild the first time you see it work. Because you realize - oh, this thing is actually thinking.
Component 3: The Brain (The Secret Sauce That Changes Everything)
Now this is the part where most people completely screw this up.
And I don't blame them - because if you just follow the default ChatGPT behavior, you're going to get useless output.
Here's what I mean.
If you tell ChatGPT, "Summarize this person," you get back a LinkedIn bio regurgitation:
"Sarah Chen is a Marketing Director with 10 years of experience in B2B SaaS. She holds a degree from Stanford and has worked at companies including HubSpot and Salesforce. She is passionate about digital transformation and customer experience."
Cool. Thanks. I could've read that myself in 8 seconds.
That's not an agent - that's a glorified copy-paste tool.
To make this an actual agent - to make it useful - you have to teach it context.
You have to give it two things:
1. The "Who I Am" Block
In my Pre-Flight Agent, I have a section at the top of the prompt that says:
"Who I Am"
And I literally paste in my bio and my business context:
"I'm Mike Allton. I'm a Revenue Architect. I help VPs of Sales, CROs, and Revenue Operations Directors build Digital Crews - AI agent teams - to eliminate the 70% of admin work that's burying their sales teams and killing productivity.
I sell three main things: Keynote speeches on Revenue Architecture and AI workforce integration. Workshops on building AI agents and governing Shadow AI. Advisory engagements for strategy and implementation.
My target buyers are mid-market Revenue Leaders (companies with 50-500 employees) who are cynical about AI hype but desperate for results. They've been burned by AI pilots that went nowhere. They don't want futurism - they want workflows that work."
Now the agent knows who I am and what I'm trying to accomplish.
2. The "Pivot" Instruction
Then I give it a very specific instruction for how to format the output:
"The Pivot"
"Based on the research you just did - the news you found, the LinkedIn activity, the company announcements - tell me specifically how MY products solve THEIR current stressor.
Don't give me a generic pitch. Don't just summarize the data.
Give me a contextual bridge. Tell me: 'Here's what they're dealing with right now. Here's how you can help. Here's the case study or framework to reference.'"
So now, instead of getting a useless bio summary, I get strategic intelligence:
"Sarah just posted about launching a sustainability initiative, which suggests she's taking on extra projects. But her company announced a hiring freeze last week, which means she's being asked to do more with the same team - or less.
PIVOT: Position your 'Digital Crew' workshop as a way to scale her team's output without needing headcount approval. Reference the healthcare case study where you helped a marketing team automate 40% of their reporting workflow.
Her CFO is ex-McKinsey (LinkedIn shows he worked there 2018-2021), so she's going to filter every vendor conversation through 'will this make my CFO happy?' Lead with ROI metrics and efficiency data, not features."
That's the difference between a tool and a teammate.
The tool gives you information.
The teammate gives you strategy.
Component 4: The Delivery (Where Does This Intel Go?)
Finally, we need to get this briefing to you where you actually live.
Don't make your reps log into yet another dashboard. Don't create a new tool they have to remember to check.
Send it where they already are.
Slack. Send it as a DM from your agent.
Email. Send it as a morning digest at 6 AM.
Text message. If your team lives in SMS, send it there.
I have mine set up to email me a Daily Briefing every morning.
I wake up. Pour my coffee. Sit down at my kitchen table. Open my email.
And there it is - a clean summary of every external meeting I have today and exactly what I need to know about each person.
I know who I'm talking to. I know what's happening in their world. I know what pain points to listen for.
Before I even open my laptop.
It's like having a research assistant who works while I sleep.
The Safety Check: Keep It Green Light
Quick reminder from Episode 2 - because I don't want you accidentally creating a data leak while you're excited about building this thing.
This agent is pulling public data.
LinkedIn profiles. Company news pages. Press releases. Blog posts. Industry articles.
That's all Green Light data from our Traffic Light governance system.
You're good to go. Zero risk.
But - and this is important - if you're sitting there thinking, "Oh, this is cool. What if I also fed it our CRM notes so it could pull in deal history and past conversations?"
Stop.
That's Red Light.
CRM notes contain customer PII. Deal details. Pricing discussions. Competitive intel.
That data does not go into a consumer-grade AI tool. Period.
If you want to level this agent up later with internal data - and honestly, that's a smart next step once you've proven the concept - you need to use an enterprise AI tool with contractual data protection. ChatGPT Enterprise. Microsoft Copilot for Business. Claude for Enterprise.
We covered all that in Episode 2.
For now, while you're building your proof of concept, stay in the Green Light lane.
Public data only. Keep it safe.
The ROI Math Your CFO Will Actually Care About
Let's talk about what this actually means in terms your CFO will understand.
Because "saves time" is nice. But "unlocks pipeline capacity" is what gets budget approved.
Here's the math:
If this Pre-Flight Agent saves each rep 5 hours a week on research and prep work, that's 20 hours a month of reclaimed capacity.
Now let's translate that into revenue.
If your average deal takes 10 hours of face-time to close - discovery calls, product demos, stakeholder meetings, negotiations, contract review - you just gave each rep capacity for 2 extra deals per month.
Without hiring anyone. Without expanding the team. Without touching headcount.
If your average deal size is $25k, that's $50k in additional pipeline capacity per rep, per month.
Now multiply that by a 10-person sales team.
You just unlocked $500k in monthly pipeline.
That's $6 million a year.
From a 20-minute automation you built over the weekend.
That's not a "productivity hack."
That's a business model shift.
And when you walk into your next pipeline review and say, "I built this agent. It's already helping us close deals we would've lost because our reps showed up cold. Here's the data" - that's when you get the green light to scale it across the entire org.
Your Challenge This Week: Build the MVP
Here's your assignment.
I want you to build the Minimum Viable Product - the MVP - of this Pre-Flight Agent.
You don't need to roll it out to the whole sales team yet. You don't need IT approval. You don't need a project plan.
Build it for yourself.
This week. Not next month. Not after Q1 planning. This week.
Go into Zapier. Or Make. Or ChatGPT with custom instructions if you want to keep it super simple.
- Connect your calendar
- Add that "Who I Am" context block (your bio, your products, your target buyers)
- Add the "Pivot" instruction (tell me how my stuff solves their problems)
- Set it to email you a briefing tomorrow morning at 6 AM
Test it.
See what it sends you. Read the briefing. Walk into a meeting and use the intel.
Then refine it. Tweak the prompt. Adjust the delivery time. Make it yours.
And when you see that first briefing land in your inbox - when you're still in your pajamas drinking coffee and you read that your 10 AM prospect just posted about a hiring freeze and you instantly know how to pivot the conversation - you're going to feel something shift.
Because you're not just saving time.
You're reclaiming your edge.
You're showing up to every call like the expert they need, not the vendor they're screening.
You're upgrading how you think.
The Future: Your Digital Crew
This Pre-Flight Agent is just the first hire.
Once you've built this and proven it works, imagine the rest of your Digital Crew:
The Proposal Agent - Drafts the SOW 5 minutes after the call ends, customized based on what the prospect actually said (not a generic template).
The Roleplay Agent - Helps you practice the negotiation before you get on the phone, simulating objections based on that company's industry and the persona you're selling to.
The Pipeline Health Agent - Checks your CRM every Monday morning, flags which deals are going stale, and suggests the exact follow-up message to send to re-engage them.
The Competitive Intel Agent - Monitors your competitors' websites, press releases, and LinkedIn activity, and alerts you when they launch something new or win a deal in your territory.
The technology is here. The models exist. The integrations are live.
The only limit is your architecture.
And that's what this show is about. Building the systems. Designing the workflows. Hiring the crew.
So Here's Your Move
This week - not next month, not after the next board meeting, this week - build the MVP of this Pre-Flight Agent.
Spend 20 minutes.
Because here's what I know:
The VP who shows up to next Monday's pipeline review and says, "I built an agent over the weekend that saved my team 15 hours last week, and here's the deal it helped us close" - that VP gets the budget for the enterprise rollout.
That VP becomes the internal champion.
That VP gets promoted.
Be that VP.
Stop doing the robot's work.
Hire the robot.
And get back to being human - because that's where you win.
In a world where every other vendor is sending AI-generated spam that sounds exactly the same, the rep who shows up prepared, curious, and genuinely interested in solving the prospect's problem?
That's the rep who closes.
Resources & Links
Download the Executive Guide to Shadow AI
Get the governance framework you need to build agents safely - includes the Traffic Light system, Amnesty Email template, and AI Use Policy ready for Legal review.
👉 TheAIHat.com/shadow-ai
Need Help Implementing This Across Your Team?
If you need someone to walk your leadership through governance, the build process, and change management for Digital Crews - that's exactly what I do.
👉 TheAIHat.com/speaking
Key Takeaways
✅ The Parking Lot Panic - Reps wing six-figure meetings because they don't have time to research between back-to-back calls
✅ Agentic AI uses discretion - It's not a rigid script; the AI decides which sources matter (Tesla factory vs. Ford assembly line)
✅ 3-Component Architecture - Trigger (when), Brain (what), Action (where)
✅ "Who I Am" + "The Pivot" - Teaching context turns generic summaries into strategic intelligence your reps can actually use
✅ Stay Green Light - Use only public data (LinkedIn, news, press releases) to keep your proof of concept safe
✅ ROI Translation - 5 hours saved per rep = 2 extra deals per month = $500k in pipeline capacity for a 10-person team
✅ Build the MVP this week - Start with yourself, refine it, then scale it
About the Host
Mike Allton is a Revenue Architect and the Director of Partner-led Growth at Agorapulse. He founded The AI Hat to help Revenue Leaders cross the "GenAI Divide" by building Digital Crews that eliminate Admin Drag and scale revenue without scaling headcount.
Mike is an international keynote speaker, published author, and award-winning content creator. He's been architecting AI workflows for revenue teams since 2024 and hosts AI for Revenue Leaders: The AI Hat Podcast.
Connect with Mike:
- Website: TheAIHat.com
- LinkedIn: linkedin.com/in/mikeallton
- Podcast: TheAIHat.com/podcast
Episode Timestamps
- 00:00 The Sales Rep's Dilemma
- 00:44 Introducing the Digital Crew Member
- 02:09 Welcome to The AI Hat Podcast
- 02:37 Building the Pre-Flight Agent
- 06:25 The Tactical Blueprint
- 14:14 The Future of Digital Crew Members
- 15:01 Conclusion and Next Steps
Subscribe & Follow
New episodes drop every Tuesday morning. This concludes our Launch Trilogy - subscribe now for future deep dives into Revenue Architecture.
