How to do 4 Hours of Data Analysis in 10 Minutes with AI (Claude)

Summary

---

Podcast Summary: "How to Do 4 Hours of Data Analysis in 10 Minutes with AI (Claude)"

Podcast: Marketing Against The Grain
Hosts: Kipp Bodnar (HubSpot’s CMO) and Kieran Flanagan (Hubspot's SVP of Marketing)
Date: December 16, 2025

Episode Overview

This episode dives into a transformative workflow for marketers: turning unstructured business data into high-impact dashboards and executive-ready presentations with Claude’s Opus 4.5 AI model. Kipp and Kieran break down how this technology can collapse hours of manual spreadsheet work into minutes, making advanced data analysis accessible, actionable, and visually compelling. The conversation is practical, fast-paced, and focused on actionable tips—designed for marketers looking to gain an edge with new AI capabilities.

---

Key Discussion Points & Insights

1. The Power and Accessibility of Claude’s Opus 4.5 for Data Visualization

• Kieran introduces Opus 4.5 as a major leap for marketers, “changing the way that all of you think about your business, analyze your business performance and present that story to all of your execs. And it’s going to take you minutes, whereas it used to take you days.” (01:29)
• The workflow starts with unstructured data—messy CSV files or multi-sheet workbooks—and ends with world-class visualizations and decks, requiring virtually no technical analytics skills.

2. Example Workflow: From Raw Data to Insights

• Kieran recounts building a dashboard that clusters hundreds of SEO keywords by topic, then tracks the impact of Google’s “AI Overviews” on web traffic—all in one prompt with Claude.
  • “This was a single prompt. I did not need to edit anything.” (04:50)
• He points out a useful tip: If you don’t have real data for experimentation, you can ask Claude to generate synthetic datasets based on simple prompts. (02:14)

3. Lowering the Skill Barrier

• Kipp reflects on pre-AI workflows: “It used to be that you had to be really good at spreadsheets to get really good insights...Now Claude is the quasi analyst and takes how you want to look at this information and you basically just have to know that the data to look at it that way exists.” (05:27)
• The workflow lets marketers quickly iterate, ask creative or strategic questions, and shift from “data manipulator to business story owner.” (09:01)

4. Geographic Visualization & Benchmarking

• Kieran demonstrates Opus 4.5’s mapping capabilities across countries and cities:
  • “What I really want to do is just be able to visualize a global map and then click into anywhere on the map and get the performance of the country and get the performance of a city.” (07:31)
• The tool automatically benchmarks each region and city against company-wide averages, flagging both top performers and areas needing attention.
  • “It actually can tell you, hey, like, here you’re strong because that’s above the benchmark.” (08:08)

5. Actionable Insights & Recommendations

• Claude goes beyond displaying data—it suggests specific next steps.
  • “Based upon those insights, give me strategic action to take. Right? And if you give it enough of the data, over time, those actions it suggests are going to become much, much better.” (10:45)
• Example: Clicking into Miami reveals tailored actions to improve close rates, each framed by effort/impact and linked to revenue potential.
  • “Look at this sales enablement blitz… if you reach the average for this city, here’s the impact in Rev, dude, that’s dope.” (13:23, 13:35)

6. Effortless Executive Communications

• Creating executive-level decks from insights is now just a few clicks away.
  • “People don’t understand how good Claude is at creating decks… you’re basically saying, hey, let’s create a deck for our exec team.” (11:59)
• Claude can replicate existing deck formats, populate them with up-to-the-minute data and actionable summaries, and even update them routinely.

7. The Strategic Shift: From Analysis to Impact

• Kipp: “The person who could manipulate the data owned the story. And now it’s the person who can ask the best questions… and how you actually improve your business.” (09:01)
• The hosts emphasize how new AI flows allow for more strategic thinking and dynamic business decision-making, with hours of manual labor replaced by focused, high-level review.

---

Notable Quotes & Memorable Moments

• “This was a single prompt. I did not need to edit anything.”
  — Kieran Flanagan (04:50)

• “It used to be that you had to be really good at spreadsheets to get really good insights. And that was like a skill barrier…Now Claude is the quasi analyst…”
  — Kipp Bodnar (05:27)

• “What I really want to do is just be able to visualize a global map and then click into anywhere on the map and get the performance of the country and get the performance of a city.”
  — Kieran Flanagan (07:31)

• “It actually can tell you, hey, like, here you’re strong because that’s above the benchmark. So it actually benchmarks all of your countries and cities against each other…”
  — Kieran Flanagan (08:08)

• “The person who could manipulate the data owned the story. And now it’s the person who can ask the best questions, can own the story of how the business is working…”
  — Kipp Bodnar (09:01)

• “Look at this sales enablement blitz. Now, this is my first time asking [Claude] to do this. I can also improve this pretty drastically the more context I give it. Channel mix here and then projected impact… Oh, that’s actually pretty good. So if you reach the average for this city, here’s the impact in Rev, dude, that’s dope.”
  — Kieran Flanagan (13:23–13:35)

• “If you were good [at reports], it would still take you hours, and then it would have taken you another hour or two to get the PowerPoint right…this flow is saving people like half a day, a day’s worth of time, and condensing that to, like, really 20 minutes.”
  — Kipp Bodnar (14:30)

---

Timestamps for Key Segments

• 01:29 – Introduction of the Opus 4.5 workflow and why it matters
• 03:12–04:50 – Walk-through: Generating topic-cluster dashboards from SEO data with a single prompt
• 05:27 – Impact of AI on lowering the analytics skill barrier
• 07:31 – Global map visualizations and performance benchmarking features
• 08:08 – Insights into city/country benchmarks and identifying what “needs attention”
• 10:45 – Claude’s capacity to generate actionable business recommendations
• 11:59–12:34 – Building executive-ready decks, translated from visualizations
• 13:20–13:46 – Example: Claude framing actions for underperforming cities with effort/impact analysis
• 14:30 – Quantifying real-world time savings for marketers

---

Takeaways for Marketers

• With Claude’s Opus 4.5 model, comprehensive analysis, insights, and action are truly accessible in minutes—not hours.
• The ability to quickly pivot your lens on the business, ask better questions, and own the decision-making process is now democratized.
• Visualizing and narrating your data for internal stakeholders—from fellow marketers to the C-suite—is easier, more automated, and far more impactful.

---

In the hosts’ words:
“If you’re a marketer or entrepreneur, you know the hardest part is turning all that raw data into a story that gets buy in… we’ve showed you really how to go from lots of random CSV files and worksheets, turn that into, like, real data visualization in any way you want, give me real insights, tell me what I should do, and then update the exec team. And that is a workflow that I think everyone who listens to the show will like to use.”
— Kieran Flanagan (14:51)

---

Recommended for anyone looking to supercharge their marketing analytics, automate reporting, and move from spreadsheet jockey to trusted business strategist—thanks to AI.

Transcript

[A]

We are going to show you one of the most powerful workflows you can use in cloud today. Using their new model, Opus 4.5, we're going to show you how to go from unstructured data, data all over the place, to world class visualizations to incredible insights and how the AI can tell you what to do about those insights. Exactly what action you should take to get better. And better yet, we are going to show you how to turn that into exec level decks. You can update your manager and everyone else you work for and, and really turn yourself into a powerhouse and impress all of your colleagues, impress your manager, impress your boss with this one incredible AI powered workflow. All of that and more on this episode of Marketing against the grain.

[B]

Using only 20% of your business data is like dating someone who only texts emojis. First of all, that's annoying and second, you're missing a ton of context. But that's how most businesses operate today, using only 20% of their data. Unless you have HubSpot, where all the emails, call logs and chat messages turn into insights to grow your business. Because all that data makes all the difference. Learn [e-mail].

[A]

Okay Kip, I want to give you the thing that I, I am really obsessed with right now.

[B]

Yes.

[A]

And it is not watching poker on YouTube, which is like one of my obsessions. It is actually Opus 4.5 data visualization into exact level decks. So we are really going to change the way that all of you think about your business, analyze your business performance and present that story to all of your execs. And it's going to take you minutes, whereas it used to take you days.

[B]

Oh yeah. This show is for people who have to do reports, have to explain their work to their boss and their CEO and are sick of the time it takes to do all the manual work and the hours it takes to build decks, to build graphs in Excel. We're going to show you how to do all of that in minutes with the ultimate cloud flow.

[A]

So I'm going to show a couple of examples of just what we mean by how good Opus 4.5 is. A data visualization. For some of my team, watching this will look like things I was doing internally. Don't worry. Again, this is all fake synthetic data. One of the actual real tips you can get from this show, if you want to just try things out and you don't have actual data to turn into these data visualizations, just ask Claude to create some fake data for your business. Give it a little of Information and say, can you create synthetic data that I need to create this kind of dashboard? So me and the search team, we were riffing on some stuff this week, and really what we wanted to see was how is our traffic being affected by aios, Basically AI mode in Google, right? When that is triggered, how is that impacting our traffic? This is the exact same dashboard we built, but for this purpose. This is all fake. But you can see what we did was we just exported CSV files. Actually, one of the files was just a worksheet with multiple tabs, right?

[B]

A complex worksheet, basically.

[A]

Yeah, yeah, a bunch of CSV files. Now, in the olden times, the search team were like, in doing pivot tables, I was never actually very good at pivot tables. I'm ashamed to say, as much as good at data. I am. I was never good at that stuff. But now I don't need to be because I was like, okay, just give me all the data. And I was like, okay, well, I'll just build a dashboard for myself and Claude using Opus 4.5 to, like, figure this stuff out. And so what I basically did here was I asked it to cluster all of the keywords by topics. So think of the RAW file just had hundreds and hundreds and hundreds of different keywords that we get traffic for. So I asked it to cluster it into different topics. Then I asked it to give me the total traffic from those topics. When the AI overview is triggered, which we get from search tools, they can tell us, hey, AI mode was triggered for this search. And when it hasn't been triggered, right? So what you end up with is, I can go in here and I can say, okay, well, our total traffic for 2024 was this with AI overviews, because of that, we're seeing a decline in traffic, which is pretty huge here and this change. But the cool thing is, so I can say, wow, like, I get that per topic to see how it's impacting the different topics that it's clustered all of those keywords within. And then I can click and say, okay, let me see this one here. And then it gives me an actual breakdown of all of the keywords in that topic. So it gives me the keyword, it tells me the traffic for 2024, tells me the traffic for 2025, shows me the change in traffic, shows me the position we're in today, and then shows me the keyword rank, and then it gives me the total change in traffic for that actual topic. This was a single prompt. I did not need to edit Anything.

[B]

You can give an example of that prompt. We could drop that in the comments for this. But basically you gave it a prompt and you gave it some raw spreadsheets.

[A]

They told it to split them out by when the AI overview is triggered on a keyword. And look here. So this is when it's not triggered, right? So we can then see, well, what's been the impact here. We've still lost some traffic, it's a lot less. And then we can get into like, okay, well which ones are we losing traffic on even when the AI overview or AI mode is not active.

[B]

So I think what's mind blowing about this is a couple of things. One, it used to be that you had to be really good at spreadsheets to get really good insights. And that was like a skill barrier, right? It's like I had to know the thing, I need to know how to use spreadsheets really well and I needed to be basically a quasi analyst to do something like this. And now Claude is the quasi analyst and takes how you want to look at this information and you basically just have to know that the data to look at it that way exists. And once you know exactly that to look at that way exists, you can look at it a bunch of different ways. And what you've essentially built is like this kind of disposable like dashboard of like, oh, I'm curious, I want to change how we're going to do work for the next month. And based on this data, I can prioritize where we spend our time. And you're going to spend your time in this example on these keywords that have lost traffic without AI overviews because that means there's just some core basic search engine optimization work that you can do and you can win some of that traffic back. And that's like an amazing hack if you're a marketer. Now, Kieran, though, the question is you're probably going to go and do some stuff, you're going to change some of this data and then you're going to want to report on it to like an exec. And historically you would have taken that same data that you put into create this, and instead of creating this, you would have done a bunch of graphs in Excel and then done the copy and paste and screenshotting, put in the PowerPoint over to PowerPoint and everything. But I don't think you have to do that anymore.

[A]

No, I want to show you one more and then we can get into that one, which is this one, because this one is a Real one I'm working on right now. So we are again, this is a minimal viable version, fake data. But to kind of give you some insights on how we are starting to use these data visualization tools. Okay, so we are a global business, which means we care about our performance across countries and cities. And it's again, pretty complex to see all of that. Tons of different data reports, local reports, Google sheets. What I really want to do is just be able to visualize a global map and then click into anywhere on the map and get the performance of the country and get the performance of a city.

[B]

How we would look at this kind of regional lens of this business, right?

[A]

And so look how good this is, right? So I click in and I say, okay, here is the United States. I can see total traffic. I can see how much is coming from paid in search. I look at my average deal create rate and I average close rate. So average deal create rate is like how much traffic is turning into deals for the sales team and then how much of that is actually closing into customers. And it actually will highlight the different cities based upon these things. But I want to show you a couple of like real incredible things when I was going through this, why I want to do this for us. It can actually highlight where you're performing well and where you're not performing well.

[B]

Right.

[A]

So it actually can tell you, hey, like here you're strong because that's above the benchmark. So it actually benchmarks all of your countries and cities against each other and then will show you which countries and cities are above benchmark versus below benchmark for any of these different data points. And then you can see here, so what are my top deal creation rates? They're in San Francisco, Austin. What are my top close rates? And then what needs attention, right? So it can tell me each and every month, hey, this country in the city is underperforming the benchmarks. And so we can click in here to one of these cities, let's say Miami. And look what it gives me. It gives me the snapshot of how Miami is performing. It tells me the paid search, gives me the exact conversion funnel, and then shows me the deal create rate and the close rate, which I think it's the close rate that is underperforming. But the thing I can write into this is like where it says, need attention, I can ask it to give me some insights on what to do.

[B]

Yeah, exactly. I think what's fascinating about this is that it used to be that the person who could manipulate the data owned the story. And now it's the person who can ask the best questions, can own the story of how the business is working and how you actually improve your business. And this again is just fuel for the curious mind. If you know your business really well and you think you have some creative ways to look at it, right, this is a creative way to look at a business like HubSpot that's across a lot of geographies. You can see performance by geography and that's pretty mind blowing because it lets you know if you have regional gaps and areas of focus. And sometimes you might have gaps, and that's okay because that just aligns with your strategy. But if you're focused in this case like on some specific regions and one of them is not working, you want to do what Kieran's about to do and dig in and get the recommendations around. Why is that not working? And what should I actually do to go and fix this problem?

[A]

Look, if you're a marketer or entrepreneur, you know the hardest part is turning all that raw data into a story that gets buy in. That's why we built the Data Storyteller stack. It includes seven prompts that will help you turn raw data into killer interactive dashboards and executive level presentations. No coding, just copy paste and watch AI do the heavy lifting to save you hours of data prep time. Get the complete stack for free. Click the link below. Now let's get back to the show. Right. It does a couple of things. We can only really look at data and the way that a software allows us to see it, and this here allows you to visualize it in any way you want. You just need the data and the.

[B]

Brain power and the curiosity to want to look at it that way, to.

[A]

Want to look at that way, and then it turns that into insights. So it actually gives you real insights. So what it's doing here that's really valuable for us is it can benchmark every single data point against all of the countries and cities and show us like traffic page, SEO, what countries are below and above bar, and then that's the insight part. So it could easily show me insights very, very quickly and I can actually tell it whatever insights I want to see. And then it creates action. Because what I'm going to do now is going to say, based upon those insights, give me strategic action to take. Right? And if you give it enough of the data, over time, those actions it suggests are going to become much, much better. So in that way you have a real data analyst who has core insights about your business and can translate that into, like, real action for the sales team. But the thing I want to show is the thing that you said. So you have now this incredible interactive visual of your business. But every month I have to provide some sort of commentary to the exact team. And I would have to go and like, copy and paste the data, go and get all of this information for everyone. So I think what we can do and what you said is like, people don't understand how good Claude is at creating decks.

[B]

This is the underrated Claude superpower. Right? Literally everybody watching this video has this problem, which is you're doing some work, you have some data in a piece of software or an Excel or something from doing your work, and you need to ask some questions, which is what that visualization was just for. And then you need to show the key learnings from those visualizations to your boss. And your boss doesn't want to see you click around an app like this. Maybe, maybe he or she does, but a lot of times they're not. They're going to want to see it in a more visual slide format. And so that's what Kieran, you're doing right now is you're basically saying, hey, let's create a deck for our exec team.

[A]

Yeah. So let's do a truly short one with AI overview.

[B]

So you're doing a very simple deck here. It's like, hey, what's my performance? Without. And with AI overviews across both three topics we've lost traffic on, and five key actions we're going to take to win that traffic back.

[A]

All right, so look at this now. So now I have my interactive map. Now I can click into, let's say la. I like LA sometimes. You know what? Miami, I have not been there yet. I want to go there. And so let's click in there. I know. I mean, okay, look at this. It creates a whole plan to improve Miami. So it's told me the deal create rate is against average, close rate against average pay traffic. SEO is fine. So it gives me a bunch of things to do. Implement exit intent with a compelling offer.

[B]

Frames them in effort impact.

[A]

Yeah, that's what I was going to say. So here's the close rate ones. Look at this sales enablement blitz. Now, this is my first time asking to do this. I can also improve this pretty drastically the more context I give it. Channel mix here and then projected impact average in national average. Oh, that's actually pretty good. So if you reach the average for this city, here's the impact in Rev, dude, that's dope.

[B]

I'm sorry.

[A]

It's so good, right?

[B]

That's so good.

[A]

We can just click into the rest of the cities. Look at this, look at this. All of the ones that are underperforming. Pretty amazing, right? So again, we have gone from like, Excel spreadsheets and CSV files to incredible data visualization insights in action. It's so good.

[B]

If you're watching the show, one thing to know is that Claude, I think, is really nailing a lot of the use cases for business users.

[A]

Right?

[B]

Right. If you're working within a company, Claude really has a lot of the features and user experience. That's the leaning in there.

[A]

That is actually the amazing part is, if you think about what I asked it to do, is I find the amount of times I have to ask Claude to redo something to be dwindled into nothing. The other thing I'm finding is there's not much prompting to do with Claude.

[B]

If you were good, it would still take you hours, and then it would have taken you another hour or two to get the PowerPoint right. And so what we're really talking about is this flow is saving people like half a day, a day's worth of time, and condensing that to, like, really 20 minutes. And that's not active. Like, I'm doing work for 20 minutes because the model's got to write the code and do the design and everything.

[A]

I will say, like, if you visualize data in different ways, it really does make you think differently about problems. You can have the model turn that into, like, real insights, like, what are the things I should care about? And then you can tell that, what action should I take? And you could say, yeah, these are the right actions. And then you can say, great, just update the exact team deck and you can give it a format. The thing we didn't show is if you uploaded a format that you use with your exact team, it will just replicate that. So you can just say, replicate this deck and now update the exact team on this. And so we've showed you really how to go from lots of random CSV files and worksheets, turn that into, like, real data visualization in any way you want. You could have asked Claude to visualize it in any way you want, give me real insights, tell me what I should do, and then update the exact team. And that is a workflow that I think everyone who, who listens to the show will like to use.

[B]

I 100% think that's true. I think this is an invaluable use case. You and I are going to use this use case multiple times a week. It's super valuable and it's going to save you a bunch of time. I hope that everybody found today's show valuable. It's a quick hit of something everybody should be doing in Claude. We'll see you real soon on marketing.

[A]

It's the grain, Sam.