The B2B Podcast Index
Index
All categories
MarketingSalesSaaSFinanceHROpsLeadershipCustomer SuccessAI & DataProductStartups & FoundersRevOpsEngineering & DevTools
MethodologySubmit
Best of:MarketingSalesSaaSFinanceHROpsLeadershipCustomer SuccessAI & DataProductStartups & FoundersRevOpsEngineering & DevTools
An independent project byFame
SearchBest episodesGuestsInsightsMethodologySubmit a podcast
Index/AI & Data/OwnerRX Podcast with Alan Pentz
OwnerRX Podcast with Alan Pentz artwork

From Spreadsheets to AI Automation: Real Business Results in 30 Days

OwnerRX Podcast with Alan Pentz · 2025-12-12 · 14 min

0:00--:--

Key moments - from our scoring

Substance score

29 / 100

Five dimensions, 20 points each

Insight Density8 / 20
Originality6 / 20
Guest Caliber4 / 20
Specificity & Evidence7 / 20
Conversational Craft4 / 20

Alan Pentz walks through a real-world case study of implementing AI automation for lead generation and business mapping, using a trash bin company (Liz's Citibin) as the central example. The episode unpacks why current AI tools fall short for non-technical business owners and introduces Claude (Anthropic's model) as the emerging B2B standard, contrasting it with ChatGPT's consumer focus. Pentz explains three critical developments: Claude Code as the coding interface that's replacing ChatGPT in production environments, MCP (Model Context Protocol) as an open standard enabling direct connections to business software like QuickBooks, and Claude Skills as a memory-management system that isolates company-specific processes (client reports, deliverables, etc.) so teams can invoke them without manual setup. The episode culminates in how Slack integration - allowing teams to trigger Claude workflows directly from chat - solves the integration friction that currently requires manual spreadsheet uploads. B2B operators, especially those managing operational spreadsheets, lead generation, or client delivery workflows, will find tactical value in understanding why Claude outpaces ChatGPT for enterprise automation and how to architect hands-off systems using Skills and Slack connections.

Key takeaways

  • →Claude and Claude Code have emerged as superior to ChatGPT for B2B enterprise use cases and are becoming the standard interface for business automation.
  • →Model Context Protocol (MCP) servers enable Claude to directly connect to business systems like QuickBooks without requiring traditional API integrations, making automation faster and more dynamic.
  • →Claude Skills allow you to encode company-specific processes and workflows that Claude can selectively apply, reducing token usage and improving accuracy for repeatable business tasks.
  • →The integration of Claude into Slack and other communication platforms will soon enable non-technical business owners to trigger complex automations directly from their existing workflow tools.
  • →Real business results - like generating 30 qualified leads daily or converting spreadsheet data to actionable maps - are achievable within 30 days using current AI tools, despite their user experience still being imperfect.

In this episode

  1. 1AI Automation Success: Jackie's Lead Generation System
  2. 2Liz's Trash Bin Company Challenge with Spreadsheets
  3. 3Claude vs ChatGPT: Enterprise vs Consumer AI
  4. 4Model Context Protocol and AI Connectors
  5. 5Claude Skills: Building Custom AI Capabilities
  6. 6Claude Integration with Slack and Future of Business Automation

Mentioned

OwnerRXClaudeAnthropicChatGPTOpenAISlackQuickBooksZapier21HatsAlan PenceClaude CodeMicrosoft 365

Topics in this episode

ZapierChatGPTAnthropicClaude CodeModel Context Protocol (MCP)Claude SkillsClaude AISlack integrationQuickBooksAI for Business Owners cohort

Questions this episode answers

What is Claude Code and why is it becoming the standard for business automation instead of ChatGPT?

Claude Code, built by Anthropic, is the backend used by tools like Lovable, Replit, and Bolt for production coding and automation. Alan argues Claude has invested heavily in B2B enterprise applications while ChatGPT focuses on consumer use, and that Claude Code will become the standard interface because business teams increasingly need to 'talk to computers' - coding is just writing symbols in files to instruct computers what to do.

How does MCP (Model Context Protocol) simplify connecting Claude to business software like QuickBooks?

MCP is an open standard that allows Claude to dynamically connect to external systems without changing API code. Instead of manual API maintenance, you can tell Claude 'build an MCP connector to QuickBooks, get this information' and it figures out the connection requirements on its own, making integrations faster and more flexible than traditional APIs.

What are Claude Skills and how do they solve the problem of giving Claude company-specific knowledge?

Claude Skills are small files that encode how your company does specific processes (e.g., building PowerPoint decks, creating client reports, project deliverables). When invoked, Claude loads only the relevant skill into memory, keeping it isolated so Claude doesn't get confused with too much information, and teams can reuse skills across the organization without retraining.

How did Alan solve the Citibin company's spreadsheet-to-map problem and what's the next step?

The trash bin company wanted to convert spreadsheet location data into dynamic maps and directions for field teams. Alan initially mentioned Zapier as a band-aid solution but explained that with Claude integrated into Slack (a forthcoming feature), they can set up a skill that lets team members request maps directly from Slack, with Claude generating the output automatically.

What real business results did Jackie's marketing company achieve using the lead automation Alan built in his cohort?

Jackie's team receives 10 fresh leads every morning, delivered to her inbox and distributed across 3 team members (30 leads total), generated by scraping and enriching data from lead services. Within less than a month of running the automation, she had acquired two new customers with more pipeline in motion.

What our scoring noted

Our reviewer’s read on each dimension, with quotes from the episode.

Insight Density

8 / 20

There are a few concrete, usable ideas - the lead-scraping-and-enrichment workflow for Jackie and the explanation of MCP and Claude Skills - but large portions of the episode are throat-clearing, cohort plugs, and very high-level AI commentary that any attentive tech reader would already know. The ratio of actionable insight to filler is low for 14 minutes.

We scraped them off of a couple of services out there, and then we enriched them with some research data to figure out who they were, more detail about them, and then plot them in her email inbox every morning. So she has 10 fresh leads every morning for three people on her team.
it's basically an open standard that helps connect two bots together. So like Claude can go talk to QuickBooks through an MCP connector and all you have to do is say, hey, Claude, build an McP connector to QuickBooks

Originality

6 / 20

The framing of 'we are all becoming coders' and 'ChatGPT should become the next Google' are recycled takes circulating widely in AI commentary; there is no contrarian or first-principles argument, and the Claude-vs-ChatGPT positioning is a standard narrative. The small-business application angle is mildly differentiated but not developed into a fresh thesis.

my philosophy is all of us who work in businesses, we all just, we are becoming coders, we talk to computers
I think they have a huge lead on the consumer side. They have the brand recognition and they should just try to become the next Google

Guest Caliber

4 / 20

This is a solo episode by a small-business AI cohort instructor; no practitioner guest appears. The host references two business owners (Jackie, Liz) only anecdotally, and the host himself openly admits he hasn't tested the tool he's recommending, which undermines practitioner credibility.

It's gonna be a solo episode today. The lovely Tanya is on some sort of Spanish boondog all holiday.
I haven't used it yet and I don't know if it's there or coming. I think they said it was coming

Specificity & Evidence

7 / 20

A handful of concrete details appear - two customers in under a month, 30 leads per day across three reps, the trash-bin-to-map use case, January 22nd start date - but most AI claims are asserted without data, and the host admits uncertainty about whether the Slack integration even exists yet, undermining the episode's evidential grounding.

she's been running it for less than a month, got two new customers, more on the way
So she has 10 fresh leads every morning for three people on her team. So 30 leads, I think total.

Conversational Craft

4 / 20

Being a solo episode, there are no guest questions, follow-ups, or pushback by design; the host meanders between topics without tight structure, interrupts his own narrative for promotional plugs, and never stress-tests any of his own claims. The format eliminates the primary vehicle for conversational craft.

So anyway, that's giving Claude access to all these things outside of it. Right now everybody's adopted that.
I got to go test Slack. I haven't used it yet and I don't know if it's there or coming.

Conversation analysis

Computed from the transcript - who did the talking, and the words that came up most.

Share of words spoken

  • Alan Pencehost96%
  • Speaker B4%

Most-used words

claude22slack12code8build7start7cloud7spreadsheet6skills6files6owner5podcast5figure5model5chatgpt5second5everybody5

Episode notes

Small business owners are achieving measurable results with AI automation in weeks, not months. Alan Pentz shares two case studies: Jackie's marketing agency now gets 30 qualified leads delivered daily through automated web scraping and data enrichment, landing 2 new customers in under a month. Liz's trash bin manufacturing company transformed manual spreadsheet chaos into automated map generation and customer directions. Alan reveals why Claude Code is dominating enterprise AI (42% of coding workloads) while ChatGPT focuses on consumers, explains MCP servers that connect AI to business systems like QuickBooks, and previews the Slack integration that will finally make AI accessible for non-technical owners. Essential listening for business owners ready to implement AI automation with real, measurable outcomes.

Full transcript

14 min

Transcribed and scored by The B2B Podcast Index.

1, 2, 3, 4. You're listening to Owner RX with Alan Pence, the podcast for business owners who want to scale without sacrifice. Each episode delivers battle tested strategies from the Owner RX Playbook library, showing you exactly how to build systems that run without you. Here's your host, Alan Pence.

Hey, welcome back to this OwnerX podcast. It's gonna be a solo episode today. The lovely Tanya is on some sort of Spanish boondog all holiday. Who knows, you know, they can just invent stuff whenever they want, those Europeans.

So I wanted to talk today a little bit about something that came up actually on another podcast. So if you never listened to the 21Hats podcast with Lauren Feldman, definitely recommend it. And he interviews small business owners about their experiences and challenges. And he has a couple of panels of people.

One of them attended my first AI for Business Owners cohort, which is now open. So go to ownerrx.com, you'll see a link for signing up for the cohort. We're starting January 22nd, I think for six weeks, so definitely check that out.

So a woman, Jackie, had been in the cohort and we had helped her build an automation where she was getting leads for her marketing company. We scraped them off of a couple of services out there, and then we enriched them with some research data to figure out who they were, more detail about them, and then plot them in her email inbox every morning. So she has 10 fresh leads every morning for three people on her team. So 30 leads, I think total.

And that was sort of like they want that human contact. They don't want automated thousand leads, that kind of thing. Right. So they reach out to them and it was really successful.

She's already, you know, she's been running it for less than a month, got two new customers, more on the way. It's been a really effective mechanism for her to generate marketing business. And so as part of that, she was talking to another panelist, Liz, who owns a company that makes trash bins for cities and now for national parks. She was talking about a big gnarly spreadsheet that her husband, who's her kind of CFO coo, was using to track a lot of stuff in the business.

So I emailed her right after. I was like, okay, let's see what we can do with AI. And she kind of expressed her own lack of knowledge about AI and then her husband's reluctance to use it. So I met with them and was able to figure out exactly what their problem was.

So, like we, we actually put the Big spreadsheet aside, because it seemed to be working pretty well. But there was another spreadsheet when they want to show customers where their trash bins and enclosures are in a city so that they can walk around, take pictures of them for marketing, or send a customer over there who might want to take a look at their stuff in the wild. So they want to, from a spreadsheet, they want to generate dynamic maps and then directions and stuff like that they can send out.

So I thought it demonstrated three really cool things that are trends going on right now, this problem. So number one was the fact that you didn't know how to do that, right? And the tools are still, I mean, they're not terrible to use, right? But they're not quite there yet for the everyday user, right?

So our first step was really just to figure out what model would be best, because they were doing this in a ChatGPT project where they would download the spreadsheet every time they needed to do it, because they're obviously adding stuff to the spreadsheet all the time. It's dynamic uploading to ChatGPT, having to run a query and then send the result out. So that was very laborious for them and wasn't always accurate. So I think the second thing that I saw, beyond the fact that the tools to connect all this make it easy aren't there yet, the second thing was a theme I'm going to start hitting a lot, which is the difference between the different models.

And I think what we're seeing emerging is that Claude code, Anthropic, which is sort of always been the baby brother of OpenAI, has really focused hard on B2B enterprise applications. So they have a model that you use. You can use their Chatbot Chaplaud AI, but they run that same model through another interface, through the API interface, and it's called Claude Code. And it's been the key coding tool that everybody's using now.

So even when you use, like, if you've heard of things like Lovable Replit, Bolt, those kinds of things, base 44 or whatever that is, they're using cloud code in the background, like they're a wrapper or skin over cloud code. And it's Anthropic has really kind of foregone the consumer market, you know, so like ChatGPT is your buddy, right? That's what you go to every day to get your, you know, you know, for it to pump you up and tell you all your ideas are good, right? And people do it for search and Google Replacement, Right.

So that's great, but really CLAUDE has gone farther on the enterprise use case. So they've started building other capabilities into the models that you're not quite seeing yet, but will be really important for you in the future. 2 I would point to our MCP server. So MCP means model context Protocol and you'll hear about it from time to time.

It's basically an open standard that helps connect two bots together. So like Claude can go talk to QuickBooks through an MCP connector and all you have to do is say, hey, Claude, build an McP connector to QuickBooks, get this information, it can go talk to QuickBooks and do it. So it's sort of like an API was at one point, but it's a lot faster and more dynamic. So you don't have to keep changing the API code.

You can just tell Claude to go figure out what QuickBooks needs and it'll build, it'll create the connection with QuickBooks. So anyway, that's giving Claude access to all these things outside of it. Right now everybody's adopted that. So like if you look AT connectors and ChatGPT or Google has something similar, those are all using MCP, right?

So Claude kind of built that, they developed this model, open sourced it, now everybody's adopted it. So that's one key thing. The second thing that just came out like a month, maybe a month and a half ago is called Claude Skills. And that was an effort to take some of the things.

Claude's a great general purpose tool, but as Anthropic says, it has no expertise. If you wanted to write a Word document or create a slide deck the way you want it to be created, CLAUDE has no capability to do that. So what they started doing is building just like little files that say, this is how you should build a PowerPoint deck or a Google Slide deck or a sheet or whatever. By the way, you can do this with non technical stuff, right?

It could be like, how do we do project deliverables, how do we do client reports? And when you create it as a skill, CLAUDE has sort of like in its memory, it has just like what skills it has. And then when one gets called, if you go, hey, build a project report, it goes and gets all the information from that file and brings it in only when it needs it and then forgets it again. So it's a way to keep CLAUDE from having too much stuff in its memory at one time because it gets confused and it isolates that stuff.

So those two things are amazing. You can now figure out you can take somebody else's skill and now duplicate what they do. So you can even take somebody I was talking to took a programming book, a textbook, fed it in, turned it into a bunch of skills and uploaded and now they don't have to take the programming course. It's all built in.

So it's pretty mind blowing. Right? So going back to what we were talking about is like that with Liz, who had the Citibin company is one. She can't access those tools in an easy way for her right now they exist.

Right, right. And Claude is the way to go. So the first thing I'm telling everybody is you gotta start working with Claude. Right.

It is, it's the future for now. And I think they have a substantial lead that at least for the next six to 12 months, no one's gonna catch up. Now ChatGPT, I think is gonna have to decide what it's gonna do. I think they have a huge lead on the consumer side.

They have the brand recognition and they should just try to become the next Google, which with advertising, free use, all that kind of stuff. Anthropic clearly has a lead in coding. And my philosophy is all of us who work in businesses, we all just, we are becoming coders, we talk to computers. Cloud code is the thing that talks to computers.

You talk to computers with. Right. So that's, everyone's going to be using it. And I think a third development that really showed me where we're going, and I'll link this back to the Citibin conversation is I just saw that Claude code is now going to be in Slack.

So you're going to be able to use Slack to call Claude to have it go do stuff. So now what we have is an interface that everybody knows and I'm sure they'll build it into teams and other stuff. Whatever you're using as your chatbot or your communication tool and you can say, hey Claude, go do this thing, it will then be able to go to files because what is code? Coding is just writing symbols and files that tell the computer what to do.

That's all it is. When you're coding, you're writing words and numbers and symbols into files that you save on a computer. And then in the cloud, that's all we're doing with knowledge work too. You're writing a document, you're creating a PowerPoint, those are just words and symbols in files.

So you are going to have connect. Your Microsoft 365, your Google workspace. I think even those could go away, right? You could just have it running in some cloud service that you don't need like that structure anymore.

But anyway, for right now you're going to talk to Slack, tell it to do stuff and tell Claude to do stuff and it'll do it and produce it in your files or just produce the file for you, right? So now you can, now you can start using these skills, right? So you can start uploading the skills to Claude and they can be ways that your company does stuff, they can be things, you know, how you do deliverables, how you create something, anything. You can start putting these in the cloud, communicating through Slack and having the output in the place where you're used to having it.

So I think those three things that I was talking about, so first is like the tools aren't there yet but then you see the Slack thing and you're like ooh, they're getting there, right? And then second thing is, you know, somebody like Liz, she can't really like the zapier thing. I think I can teach her how to do that, right? That's a good bubble gum and duct tape way to solve the problem.

But now that she has Slack, think about that Slack connection. If she has Slack or teams or whatever. Now we can hook and hook and hook. We can set up how do they want maps to look, how do they want the searches to be run in Skills, right?

We can have the message Claude say create this map through Slack. Claude will go do it and pop a file wherever you want or send it back to you in Slack. So now I feel like just today I read about the Slack thing, like now I know how to solve our problem, right? So I got to go test Slack.

I haven't used it yet and I don't know if it's there or coming. I think they said it was coming but I think this is part, you know, this is a real world example where I'm seeing the things change in real time and it's very soon that you're going to start seeing it in your day to day. And I think the one thing if you take away from this is it's coming soon. Second thing is get to know Cloud cause it is the future for business.

So I think that's gonna be it for me today definitely. Go sign up for the AI for Business Owners cohort. If you haven't starting in January just go to ownerrx.com and check out my newsletter.

It is also@ownerx.com newsletter you can read about this kind of stuff. Too. All right.

Appreciate it. Talk soon. You've been listening to Owner Rx with Alan Pence. Want to apply what you just lear?

Try our AI business advisor@ownerrx.com it has all the insights and lessons from this podcast, plus hundreds of playbooks. Ready to solve your specific business challenges. Owner Rx, Stop being the bottleneck.

Start being the owner.

Related episodes across the Index

Other episodes covering the same guests and topics, from across The B2B Podcast Index.

  • Telstra’s Time Sync Outage, DoorDash’s 1.5M RPS Cache, Stateless MCP, GitHub and PyPI Supply Chain Delays, and Why the Quietest Infrastructure Often Has the Biggest Blast RadiusShip It Weekly · on Model Context Protocol (MCP)97 / 100
  • Episode 023 Deepfakes and the Liar's DividendAI Tools for Practicing Lawyers · on Claude AI96 / 100
  • The 18x Midas Lister Betting $3B on AI (and calling most of it fake) | Navin Chaddha, MayfieldThe Peel with Turner Novak · on Anthropic91 / 100
  • 183: Why Trusted Data is the New AI Moat (w/ Rick Kranz @ AI Marketing Automation Lab)Move The Needle · on Anthropic91 / 100
  • Why Your Enterprise AI Pilot Won't Scale (with Nate B. Jones)CXOTalk · on Claude Code87 / 100
  • The Terminal as an Agentic InterfacePodcast Archives · on Claude Code87 / 100

More from OwnerRX Podcast with Alan Pentz

All episodes →
  • Stop Getting Scammed: Why Business Owners Must Learn AI w/ Alan Pentz52 / 100
  • Outpace Everyone With AI: Why Owners Must Lead AI Adoption w/ Alan Pentz & Tonya Berenson51 / 100
  • The AI Productivity Tsunami: How Owners Must Adapt Now68 / 100
  • Blocked Data, Rising Costs: The Truth About Today’s SaaS Market w/ Alan Pentz & Tonya Berenson77 / 100
  • Claude Code, MCP, and the Future of Enterprise AI w/ Tonya Berenson
Explore the best B2B AI & Data podcasts →
All OwnerRX Podcast with Alan Pentz episodes →