The Biggest Goal Podcast
The Biggest Goal Podcast, with Alane Boyd and Micah Johnson.
Learn how to reach your biggest goal as we reimagine the workflows you run into scalable systems that use today's latest technology.
Every week, we take one real workflow and cover what you're likely doing now and how to improve it with today's latest technology. We'll look at where AI could be applied, what should be automated, and where you still need a person.
Alane and Micah are the co-founders of Biggest Goal. They grew an agency to 600+ active clients before their exit and have since helped 300+ teams put technology to work building scalable systems across every department.
(Formerly this podcast was called: Automate Your Agency)
The Biggest Goal Podcast
How to save hundreds of hours with Claude's most overlooked feature | EP 118
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
Alane Boyd wanted Stripe and QuickBooks in one finance dashboard, and it was quoted as months of development. With a Claude artifact, Alane and Micah built it in under an hour. In this episode, they explain why artifacts might be Claude's most overlooked feature.
If you're tired of copying answers out of an AI chat, artifacts are the next step: real interfaces with buttons, filters, and live data that Claude builds for you in minutes. Alane and Micah share the three ways they and their clients use them every day, and the mistake teams make once building them gets easy.
In this episode, you'll learn:
- How to build a working dashboard in about 15 minutes from more than one data source, starting with a rough prompt
- Why you should keep your data in ClickUp and put an artifact on top as your own command center, synced both ways
- What happens when you share an artifact externally and why connector data stays private
- How to chat with a folder of documents, from SOPs to a 400-page RFP, using cloud storage and an MCP connector
- The biggest obstacle with artifacts and how to decide what actually deserves to become one
If you want AI tools built around your own work instead of another chat window, press play and build your first artifact today.
This episode is brought to you by Biggest Goal, the company Alane Boyd and Micah Johnson co-founded.
Fifteen years of scaling their own companies taught them that adoption is a people problem long before it's a technology problem, so we train executives and their teams through workshops, cohorts, mastermind groups, and hands-on support.
Want to learn how to scale faster with AI?
Sign up for our free LIVE masterclass!
Thank you for supporting the podcast!
I wanted a Stripe and QuickBooks Online dashboard for seeing all the finances in one area. And we it was going to be, you know, months of development for us to build out that dashboard for just an internal use. And when Claude Cowork came out, we built it in less than an hour. Hi, you're listening to the Biggest Gold Podcast, and I'm your co-host, Alane Boyd.
Micah JohnsonAnd I'm Micah Johnson. In today's episode, we're going to talk about Claude artifacts and why they might be the most overlooked feature that Claude has.
Alane BoydI think it's one of the most fun features that Claude has.
Micah JohnsonOh, I'm loving them. Like they're absolutely insane. So in this episode, Alane, just so everybody's listening and knows why they should keep listening, we're going to talk about what artifacts are and three main use cases that both us and a lot of our clients are actually using artifacts for as well.
Alane BoydThey're such a great visualization tool.
Micah JohnsonYeah. So let's start off and just explain what an artifact is.
Alane BoydIt's hard to explain, you know, really, because I get that question a lot whenever I'm talking about it. And I'm always like, well, it's kind of like something you can go back to all the time. And it has a design visualization piece to it, but also interactive.
Micah JohnsonYeah. So I I think I would describe it as it's the kind of first steps we're seeing in getting out of that like text-in, text out, text-only interfaces that we've had to deal with with LLMs and AI for the last three plus years. And now we're actually getting real interfaces. The difference is nobody's building these interfaces. AI is able to build the interfaces in ways that we're used to, right? Like buttons and toggles and text fields and check boxes and all that kind of stuff. But it's on the fly. It's however we need it to look based on whatever we're talking to AI about.
Alane BoydIn seconds, like I wanted a Stripe and QuickBooks Online dashboard for seeing all the finances in one area. And it was gonna be, you know, months of development for us to build out that dashboard for just an internal use. And when Claude Cowork came out, we built it in less than an hour.
Micah JohnsonYeah. And all of the examples that we're giving today, dashboards we'll get into as the first one. It's legitimately go to Claude, ask it to build an artifact of something that you want, and within 15 minutes, you have an entire working version of what you want. So I love this first example that you gave, Alane, because it's multiple data sources. I think you mentioned QuickBooks and Stripe. So we grew, we grew, we grabbed multiple data sources, and then we said, Hey, Claude, build us a dashboard that displays this data in this way. I think that was about it for the first version.
Alane BoydFor the first version, it was, yeah. And I tweaked it a little bit by just typing what I wanted, telling it, okay, I want this to look like this, this to look like this. And then there was some nuance we had to build in there too, which is why I wanted the dashboard in the first place. But I mean, it it can read meta descriptions that we're putting in on how payments are coming through. It is remarkable.
Micah JohnsonWell, and one thing just to point out, Alane, is we didn't try, and I think you in particular, since you kind of own this, like you didn't try to get this right the first time. We gave it a really crappy prompt. It gave you a visual example, and then you were like, actually, I need this or I need this. And you iterated over I don't know, I don't know how many prompts, but not a lot.
Alane BoydYeah, not a not a ton. And then I this is the piece that I really like is that it's gonna create it in whatever like color and branding initially that it comes up with. But then I said, hey, I want this in biggest goal with the biggest goal design system. And it puts the entire design making it look like our piece of software.
Micah JohnsonYeah. And I think since then you've been using this dashboard, you haven't needed to make any other updates. It's pulling all the data live from the sources that it needs, and it's an amazing little tool.
Alane BoydYeah, I mean, and that's just one of our dashboards that we create. And and some dashboards I do create, and then I'm like, actually, I don't like it in a dashboard anymore. So, you know, these are so quick that you it's not like you're putting in this huge resource investment of time and money into to see if you like it in the dashboard or not. Because there's some things that I thought sounded really cool, like my morning brief dashboard dashboard. I thought, oh, this is gonna be so cool. I'll have um my upcoming meetings and everything summarized on one side, and then I'll have my unanswered emails that I need to get back to on the other side. And I found that that was just overkill, that I just liked working from the scheduled task originally, but it wasn't that I was mad at myself for ditching my dashboard I created because it took three minutes to create that dashboard originally.
Micah JohnsonYeah, I mean, you didn't have to pay anybody money to do it, it didn't take weeks or months. It was that three minutes, five minutes, look at it. I go, I actually I don't know, man.
Alane BoydYeah, I don't, this is overkill. It's more than I need to do. So I think this one's a really good one that you can, if for our listeners, just go and play around, build something that you think might be interesting.
Micah JohnsonYeah. All right, so we took this dashboard concept as artifacts have evolved and took it one more step deeper, I guess I would say. Now we built something for ourselves. As we're as we're planning for 2027 already, look at us go getters. Um, planning for 2027. Normally, what we do is have like a big old Google Doc, put all our ideas in there, eventually start building projects and tasks and project plans inside of tools like ClickUp. We did it completely different this year. We actually started in Claude. We went to Claude and said, Hey, this is what we want to accomplish in 2027. These are our goals. This is kind of our end result objective that we want to have done. And I think we had like kind of our broad strokes started with that, Alane. And we maybe gave it like five or six of our very next tasks. Now, the the thing with this is we didn't ask it to build it all in an artifact. We said build this in ClickUp and then build an artifact that gives us a custom interface that's synced with ClickUp.
Alane BoydIt's such an important piece of where your data lives, because I don't want another like quote unquote project management system in built-in Claude. I want that running from our main source of data in our project management system. But what is really cool is that we can manipulate how that looks, even though it's pulling from ClickUp. So everything's still housed there, and we can mess around with it in the dashboard view or our command center view, and it can pull from multiple places, which sometimes we're like managing tasks and projects in ClickUp, but we've got a Google Doc with some stuff. We've got the competitor research that we did leave living in another place. So now we can put that all in the one command center view.
Micah JohnsonYeah, and we did the same thing. Like we had our basic command center, it created six different projects. It created some of the tasks that we knew that we wanted to do. Other tasks we didn't even know yet, but that's just chatting with Claude and be like, hey, can you add some of these tasks to this work stream, which then updates the artifact because it's updating clickup and then it keeps everything in sync. But there's a couple of things in here where we started using this and working with it. And I'm like, you know, it'd be really cool. It would be a metric at the top of how far through this initiative we are. And it looks like five minutes later, we had it.
Alane BoydYeah. And if we close something out, it's interactive. So if we close something out in the dashboard, or if we close something out in the command center, it's going to close it out in clickup. So we're not managing two places.
Micah JohnsonYeah, absolutely. It's fully two-way synced, which would give me such a headache and heartburn thinking about how we would have done this otherwise. Claude does it, no problem.
Alane BoydMicah, I think we missed uh a piece of these that they can be shared internally with each other. So we can share these artifacts with each other. So we're not rebuilding it, or it's not necessarily private, even though it can be, it can be just for your view, or you can share it, but you can also share it externally.
Micah JohnsonYeah, so a couple of points on that. I'm glad you brought this up, Alane. The sharing is getting more and more powerful. All artifacts that you build start private, but then you can share it with one or more individuals on the team. You could share it your whole team that's on your Anthropic account. Um, but you can also share this externally. The catch with the external sharing is that you can't have connectors. So uh because when we leverage Claude to connect to ClickUp, it's through my account or it's through Alane's account. So if we share this externally, you can see the whole interface, but the data isn't there because that external person isn't connected to ClickUp, which is where the data exists. So there's no data to populate in it in it at all. What's cool is it doesn't crash, it doesn't give a big error, it just doesn't populate the data.
Alane BoydBut there's oh, I get we get shared on some of the ones that our clients build. And I'm even seeing for our AI committee or the AI champions at companies, like they're building um almost like a roadmap that they're sharing with me as an external member of their team while we're going through this AI adoption at the company. And so I can still see everything that they're sharing with me because it's not any private data that's in there. It's really just what is our plan? We want to share it with you in this visual way. I can share it with our leadership team so that they know what's going on. And here is like a home for us to go back to on what our plan is.
Micah JohnsonWell, and this isn't really um apparent when you're looking at artifacts, but there's a database behind every artifact. So you're actually able to store data, structured data, attached to any artifact, which I don't want to get too technical on it, but that means it can be self-contained, like you're talking about, Alane.
Alane BoydAnd they can be updated. So anytime you can go back and make changes to your artifacts. I even asked you this, Micah. Well, could this act as a um client portal if it can show the information and then so it can't whenever it's pulling from a data source like that. So, you know, you might be like, that's unfortunate. But on the other side, it is really nice from a security standpoint.
Micah JohnsonAbsolutely. And that's why they do it. So, yeah, so we took dashboards, which is kind of simply just viewing the data from different sources. We turn that into command centers where it's actually two-way syncing between other one or more platforms that Claude is connected up to and still leveraging all the benefits of the dashboard. Those are our first two. The third use case that we wanted to share that we use, we see a lot of other people using it this way as well, is being able to chat with your assets, your documents, your PDFs, your RFPs. Um, I don't know, there's an untold amount of things that you would want to chat with.
Alane BoydI've I've been seeing this as such a cool solution that companies are building out. I see it with your SOPs and having it's pulling from your Google Docs, if that's where you're keeping your SOPs. It's pulling from your original data sources, but you're able to chat and get the answer. Also see it on like 400-page RFPs or something like that, where you don't want to have to sit there and scroll and find these details. The AI can do it in two seconds.
Micah JohnsonYeah, absolutely. So, um, and it's really easy to set up. So you mentioned Google Drive, but it just has to be cloud storage. It doesn't matter if it's Google Drive or SharePoint or OneDrive or Dropbox or Box, put your assets in a folder, allow Claude access to that, and simply go to Claude and say, Hey, build me an interface that allows me to chat with this folder of documents. And it, like you said, Alane, it could be SOPs, it could be a single RFP. You can do all kinds of stuff with this.
Alane BoydAnd and just to like go back to what we're really doing here, it's a visual interface that can be branded with your branding. So if it's an internal SOP system and you want your team to have access to it, man, it can be branded as you. And it looks so incredibly beautiful.
Micah JohnsonYeah, and again, five minutes it takes to build this entire thing. Now, we do a lot of these live, uh, even building it live with clients, building it in our live masterclasses, different things like that, Alane. If you have the brand guidelines, if you have the brand kit already ready, it typically pulls from that. The trend that we're seeing with Claude right now, especially with like Opus, is that it will analyze the documents. It will typically build an entire index of the documents that are in there, give you hyperlinks to all the pages, allow you to search, and allow you to chat. And so it's like a full-on miniature tool that represents, and again, it's live data. It's not storing all that, it's live data from the documents themselves.
Alane BoydAnd with the cloud storage piece, the most important thing that you're looking for here is that it has an MCP server or a connector in Claude. That way, Claude has access to getting in and reading those documents so you can chat with it.
Micah JohnsonYeah. So those are three ways, and we're still coming up with more and more ways, but a lot of things kind of fit within those three categories. This is something that you can use today. Like after this episode, go to Claude Desktop, start in Claude Cowork or Claude Code and say, build me an artifact to do whatever you can imagine. Try one of the examples that we gave. The only thing that you have to do is connect Claude up to the data sources, the platform, uh the cloud storage, whatever it might be. As long as it has MCP access, like Alane said, you can ask it to build an interface and then just chat with it to change the interface, and you're off and running. All right. So this is time of the episode, Alane. We get into the normally we talk about the biggest goal, right? That's why we named ourselves that, what you can achieve. This part is the biggest obstacle that people face with implementing something like this.
Alane BoydWell, I'm starting to see kind of a trend around, I'll create an artifact for that. Oh, I'll create an artifact for that. And what happens is it's almost like Micah, you remember when you start creating Google Docs for a lot of stuff, and then people don't have access to it, and then they're like, well, that it doesn't exist if I can't see it. Well, I can start seeing where you get a little too artifact happy, and you've got so many artifacts that you can't manage them. Your team doesn't know that they exist, the people that need access to them don't have access to them. So, really just thinking about what should be an artifact, what is the data behind it? So we're not recreating the wheel because we are trying to build scalable systems for our teams to have access to that meet our needs, that meet the team's needs. And so, really just staying true to that data layer, where are we pulling things and not trying to reinvent the wheel of tools that we already have.
Micah JohnsonLove it. All right, everybody, thanks for listening to our podcast today on Claude Artifacts. Do go try to build an artifact yourself. Listen to Alane and don't build 100 of them because they're super fun to build. Do think about what would actually be valuable from a from the context of a dashboard command center, or even just being able to chat with specific assets or documents on one of your cloud storage providers.
