Transcript: 'How Every's Team Used AI to Ship Its Biggest Launch Ever'

The transcript of AI & I with Brandon Gell, Yash Poojary, Austin Tedesco, and Douglas Brundage is below. Watch on X or YouTube, or listen on Spotify or Apple Podcasts.

Timestamps

1. Intro: 00:00:00
2. All Access explained: 00:00:35
3. Yash’s tech stack and how he’s automating testing pipelines: 00:03:01
4. The idea to execution loop: 00:08:02
5. How an agent turned an idea into $25,000: 00:10:25
6. The AI sandwich workflow: 00:17:50
7. Making AI tools accessible to solo builders: 00:22:03
8. Douglas on brand and design: 00:28:50
9. Tips on what to build rirst: 00:34:51
10. What’s next for All Access: 00:43:46

Transcript

(00:00:00)

Dan

We generated the largest subscription revenue increase in Every’s history last week.

We did it by launching this thing called the Builder Pack. Basically, we partnered with a bunch of the AI tools that we use every day, and subscribers to the Builder Pack get discounts and credits to those tools—$1,000 in Codex credits, 12 months of free Cursor access, all that kind of stuff.

The Builder Pack’s awesome, but I think the actual story behind it—how we did it, why we did it, how it generated so much revenue—is really interesting. For this episode I had the t

eam who worked on the Builder Pack, from conception all the way through execution and launch, come on the show to share how we did it and what we learned.

I hope you like it.

Brandon

Hello everyone, and welcome to AI & I.

This is a slightly different version of the pod today—we’ve got myself, Yash, Austin, and Douglas here, and we’re missing Daniel. We’re going to take you on a journey today, introducing you to All Access and the Builder Pack. This is a new tier at Every that we decided to launch recently, and we want to talk about why we decided to build it, why it’s important, how we use it, and what it means to be a builder today.

All of us are builders and use these tools every day, so hopefully by the end of this there’ll be at least one nugget you can take and use in your everyday life. So, to introduce this pod, first I’m going to introduce All Access.

We launched it Tuesday, on the 14th. It was a big launch for us—a big endeavor across the whole company. All Access is a new tier at Every: a $625 annual plan that gives you access to the Builder Pack, which includes 10 exclusive benefits from AI companies you know and love, enabling you to build and adopt these tools in ways you may not have been able to before.

All Access also includes a couple of other things: access to office hours we’re hosting that are exclusive to All Access members, unlimited use of Cora and Spiral, and a few other things up our sleeve that we’ll be launching in the next few months.

But we really want to talk about the Builder Pack, because that’s probably the meat of All Access. The Builder Pack is 10—soon to be more—exclusive benefits to AI tools that the Every team uses every day, so you can get access to these tools, use them for free for a while, and get comfortable adding them to your stack.

It’s early, exclusive access to the Everystack. This is the team that brought it to life. All right, Yash, how do you use the Builder Pack?

Yash

I run one of the products at Every called Sparkle, and I also work on the growth team now.

A large part of my work is using Claude. I know at Every we’re mostly Codex people, but I’m still very Claude—I’ve been using it for the last year, and I built out the entire Sparkle product, rebuilding it about 15 times, using Claude and PostHog. Those are the tools I reach for the most.

Recently I joined the growth team, and growth meant a lot of A/B testing and segmentation. I started optimizing in PostHog, but I find it really annoying to open dashboards and set all of that up manually—that’s where my Claude setup helps a lot, since I have custom skills built out.

My plan over the next few weeks is to automate the entire A/B testing pipeline. A lot of the work I did in my first month on the growth team felt very boring because it’s so straightforward—I should be working on more interesting problems, not clicking buttons and figuring out the right audience size manually. Claude is so much better at that. I have to prompt it, figure out what the right audience size is, when an experiment should go from a 10 percent test cohort to 50 percent—why should I be doing that manually? It’s a lot of fake work, and we spend a lot of time just observing these things instead of automating them. Setting up an experiment every time felt so boring to me—the only way I can run more experiments is by automating the entire pipeline.

Brandon

What’s interesting to you is less the work itself—the work is no longer setting up the pipeline. It’s actually: what should I test, and then, did it work? Great, implement it—what’s the next test?

Yash

Exactly. I’m more interested in the ideas—should we do a homepage takeover, and who should we show it to? Actually setting that up isn’t the interesting part.

We have so many experiments live right now—when should we cut off the paywall, when should we show it? Those are way more interesting, and they’re more about human psychology, so I have more fun with that.

I also feel like a large part of our work at Every is that everyone is very secure in what they do, so we’re okay with automating our own jobs. I spent one month on this stuff and thought, “Okay, I’m done—I want to do other things now.” At a different kind of organization, I might try to stretch that work out for a year, but here you want to move on to the next thing you can tackle. That’s fun, and it’s why we’re always at the edge of AI—we’re like, “Okay, this is done, what’s next? Is there a better model for this?”

Brandon

So what in the Builder Pack do you use every day?

Yash

Every day it’s Cursor. I use it because they have cloud agents that are really good—we did a lot of Fable 5 optimizations, and because I didn’t want to mess with my own setup while working on something else, the cloud agents just spin off a server, run there, and give me the results back. They also have a really good, really cheap model called Composer, which I love.

I also used Framer a lot while building the Sparkle site—it has good MCP support, so I could change headlines and see them live, experimenting more directly without going into Framer and setting everything up by hand. It’s very annoying to spend time getting the right font size for mobile—I used to do that manually years ago, and today I shouldn’t be spending my time on it.

And I’d say Claude Max is my go-to—it feels insanely underrated for the price. That’s my stack, along with PostHog. I use a lot of these products, and it’s so cheap that I’d pay way more for it. I couldn’t work without it.

Brandon

Right—Anthropic, I hope no one from Anthropic is watching this.

Yash

Get the deal, guys. Get the Builder Pack.

Austin

I like the way Yash is describing one approach to work—there’s an approach to being a general manager in the NBA where you just keep trading for future draft picks because you’re on an eight-year time horizon for contention. For someone as skilled as Yash at go-to-market engineering and growth, you can use this stack to operate at that level in a “set it and forget it” mode.

I run the growth team, and I look at what Yash is doing—he ships an experiment, it moves the needle by two percent, I figure he’s on it, he knows how to set up that Claude Code and PostHog loop. But the exciting part is what happens after: he looks at how we can win right now, how he can be creative and competitive to find the next thing, and how he can use always-on loops and automation so he can do the work that actually lights him up. I can see him in Slack going, “What if we made a three-hour, time-constrained discount targeted to these users, using this video, built in Framer?” That kind of work has always been the most fun way to work—it’s how smart, creative people differentiate themselves.

What the Builder Pack allows is for someone like Yash to automate and offload a lot of the work he’d otherwise have to manually click through in PostHog or code himself, so he can go be creative in the area where he actually has expertise. And it allows someone like me—who doesn’t really know what any of that stuff is, but is naturally more of an ideas person—to hand the model and the SaaS tools in the Builder Pack my outcomes and ideas and just let it run, which is the craziest part for me.

(00:10:00)

Brandon

Can you tell that story—you had an idea for something, went to bed, woke up, and had four emails waiting?

Austin

Yeah—it wasn’t even my idea. We launched the Builder Pack on Tuesday, and it was going really well. I’m on the West Coast; most of the team is on the East Coast. As they were getting ready to go to bed, Yash dropped a note in Slack saying we should send emails in the morning to a group of users who’d shown checkout intent but hadn’t converted, since we were running an early-bird discount for the first 24 hours of launch—so we should probably hit them again with some urgency. Yash said, “This isn’t actually my idea, it’s Dan’s, but I’m sharing it,” and everyone signed off. It was around 7 p.m., I’d been in front of my computer all day, and I really wanted to go to the gym.

So instead of copy-pasting the Slack link like I usually would, I took a screenshot of the thread, dropped it into a Codex session, and said, “Can you do this?” Then I ran /LFG—the harness in our compound engineering plugin that takes an agent through a brainstorm-plan-work-review-compound loop, which to me is the ideal way to work with agents. Then I went to the gym.

By the time I came back, it had basically already finished—it probably wrapped up around the time I was changing into my workout clothes, because it didn’t actually need that much time. It had created four different audience-based cohorts in Kit: people who’d gotten to the checkout page, people who’d visited the site but didn’t check out, our highest-engaged free users, and so on. It drafted specific emails for each audience, ran the copy through the MCP for our AI writing app, Spiral, so it matched a writing style we’d already approved. It looked through all our previous sends in Kit—both for language style and for what had driven click-throughs and opens—grabbed a social share image that had performed well before, embedded it, and scheduled everything for the next morning.

The crazy thing is, I wasn’t even thinking “holy shit, that’s so cool”—I was just like, “that’s basically what it should do.” My brain is so used to this now that it was literally exactly what I expected.

Brandon

I think it also posted about it somewhere?

Austin

Yeah—we have a Slack channel called All Emails. If you’re a subscriber to Every, I’m sorry, we send you a lot of emails, we’re working on it. One step toward working on it is having a channel where we build all our email sends through Codex or Claude Code, and the agent pushes a message saying, “Here’s what’s going to be sent, to whom, and why,” so the team can review whether we should actually send it to that audience at that time.

I pulled up the four drafts and made very light tweaks—partly because I thought I could write a slightly better headline, and partly because I’m still a control freak who doesn’t want to hand over everything to AI. Maybe ten extra minutes of work after I got back from the gym. Then we scheduled them and let it rip. In a couple of hours the next morning, those emails alone had driven more than $25,000 in revenue, plus multiple conversions, and it suggested things we could do to retarget the people who didn’t convert.

Two nice things happened: I got to go to the gym instead of sitting there building the emails myself, and while I was there—not looking at my phone, not looking at Slack—I had other ideas for what we could do the next day. Disconnected from being in the weeds, I’m working out thinking, “oh, we should try this experiment and that experiment,” because I don’t have to use my brain and my fingers to click around in PostHog to find the exact right audience segment—which I love using, but don’t need to operate manually anymore.

Brandon

It’s just so remarkable—going from “here’s an idea” to “go do this idea,” using /LFG to actually execute on it, in this case sending these emails, to generating $25,000 in revenue. That doesn’t happen every time you run /LFG, but it’s a real possibility, and then it compounds—you learn from it, and it recommends what to do next. A model like 5.6 Sol is expensive, but relative to the value you got out of that, it’s basically nothing. And best of all, you went to the gym and took care of yourself. If you’re really harnessing these tools well, for work and for life—I use one as my trainer at the gym—it can be incredibly fulfilling and effective.

Yash

It lets you get better at the things you actually want to get better at, and it frees up time. People are always pitching growth ideas to Austin and me—“do this, do this”—and to be honest, those are usually good ideas, but I don’t want to take time away from my main work to chase them. That whole thing should be automated, because the idea is good and we shouldn’t need to hire an engineer just to implement it. If I have a workflow—“let’s try this headline on our website”—that should be automated. That’s the goal: more time for deep work, more time in a flow state, without compromising on any of it. Building that loop is what puts me in the flow state, instead of just being told, “try this, try this, try this.”

Austin

I think what we’re circling is that this is the best way to build right now. Our colleague Kieran has described it as being at the top and bottom of the “AI sandwich”—you decide what the problem is, you frame the idea, and then you review the work. That’s been true for a while, but what’s exciting about this moment, about using the tools in the Builder Pack, is that you can go from idea to execution quickly, easily, reliably, and at a very high level—just by typing, or using a speech-to-text tool like Monologue and saying the idea out loud.

What I really like is that whether you’re primarily a Claude user like Yash or a Codex user like me, the tools all speak to each other really well now. Notion is a great example—I’ve hated it for years because my brain doesn’t work that way; I’m not going to sit and organize a database and make sure the growth hub makes sense to everyone else. It’s a lot of work for work. Up until about a year ago, Notion had too much of a barrier to entry around organization and maintenance—any time I tried using it with a team, we’d churn because no one was around to maintain it. Now all of that maintenance has been pushed off to Codex and Claude Code. Once a week I just ask Codex, “can you reorganize our growth hub?” and it does it perfectly, or “make sure this database is set up well.”

(00:20:00)

My favorite example: while I was doing some in-the-weeds Codex work—steering it, going back and forth—I had Claude, in Claude Code in the desktop app, using Fable, with the videos we’re releasing for this launch connected through the Descript MCP. Fable just runs in a loop, building out the edits, reviewing its own work, and interacting with Underlord, Descript’s AI tool. I’m not a video editor—I’m actually really bad at it, but it sometimes falls under my responsibilities. That loop runs so that by the time I sit down to edit, Fable and Descript haven’t produced something publishable, but they’ve gotten it about 70 percent of the way there while I was steering Codex. So I can go in and think, “okay, I know what to do now.” I can see the pieces—it does this incredible thing where it takes the transcript, builds a storyboard, writes a script itself, and pieces the video together, so I just have to do a few small things to get to a final product.

(00:21:07) [Sponsor break—ad content omitted]

Brandon

I think what I love about the Builder Pack specifically is that a lot of the time at Every we’re telling people, “go build, go build—there’s no excuses, be the builder you want to be.” And that’s great, but—how? One answer is: subscribe to Every, we’ll teach you how. But the other is that this stuff is genuinely expensive. The stack you use, Austin—the stack you use, Yash—the stack Douglas and I use—it’s very expensive. We spend a lot of money at Every on AI and credits.

Last night we had an event in the office, and a friend of mine’s younger brother came. He’s an engineer, and even so, he’s really struggling to get a job right now—and he’s not using AI at all. I introduced him to some people, and by the end of it he was really excited, but then they started telling him how much money their companies let them spend on AI. One guy, a single engineer, had spent $30,000 last month. I could see this kid visibly gulp, like, “how am I supposed to get access to this technology when it’s this expensive, and I’m just an individual, not at a company?”

That’s what I love about the Builder Pack. Take the OpenAI Codex benefit—three months free of Codex, plus $1,000 in overage credits. But those three free months include something like $5,000 of value. These subscriptions, like our Anthropic benefit, pack a lot of spend into a small subscription. It’s the same for OpenAI, Cursor, Gemini, and Anthropic—since they’re model companies, it’s structured as a one-month offering, but you can stretch the Codex credit across three months.

It’s so much value that it lets you go experiment with the technology and become the builder you want to be. I use all of it constantly—my stack is basically everything we use at Every. Personally, I’m in OpenAI and Anthropic. I’m more of a consumer of PostHog than a heavy user of their AI credits, and I’m a big Notion user because I do a lot of ops—also Framer, Render, and I really like experimenting with the Flora MCP. So I’m kind of all over the stack, but my biggest use is obviously the model companies—doing my work inside Claude or Codex. Ironically, for almost all of these tools, if they’re not a model company themselves, I’m using them through a model company—I almost never go directly into their own interface. Even Flora, which has a genuinely great interface, I only ever use through the MCP.

Austin

And that doesn’t devalue those tools—it actually pushes those SaaS operators to be the very best at the thing they do. A lot of us went through a phase from January to March where we thought, “let’s build our own CRM, our own dashboards, do all our own design.” You start doing it and quickly think, “man, this really sucks.” When you can instead say, “I really trust PostHog to run and automate my analytics, I really trust Notion to be the source of truth for the business”—even if you never open it up, even if your pages and databases look incomprehensible to a human—it doesn’t matter, because when I’m in Claude Code or Codex, I can say, “I need to make this plan, I need to do this thing,” tag @Notion, @Slack, @PostHog, and it finds what it needs.

That’s one reason this suite of tools is so powerful right now: the jump from 5.5 to 5.6 came with a real improvement in the model’s ability to go find the context it needs and maintain that context through whatever compaction happens as it works over the long term. If you looked at my prompts in Codex, you’d probably think I don’t know what I’m doing—I’m usually just saying, “look at all this stuff, set a goal for yourself, use compound engineering,” and then, “go.” There’s barely any thought or work in that moment. The real work is in having identified this as the right thing to do, and in knowing that I or my team have already built the harness the agents can work inside of and be trusted with.

I think to a lot of people that sounds like a ton of work up front, but the reason we’ve gotten here at Every is that we did it incrementally—one thing started working okay, and I’ve been trying to automate our marketing email sends for months. It keeps getting a little better, and through compound engineering the results keep compounding, so the project keeps getting a better understanding of the problem. It comes back to the idea that the best time to start doing this was a year ago, and the second-best time is right now—you can start this compounding flow with models that have really caught up to how fast you can work.

Yash

I’m curious, Douglas—what’s your stack? You’re the creative director behind all this, and someone actually asked me why this branding looks so different from Every’s—like, why is it green? I was like, “okay, I’m going to ask that.”

Douglas

I think it’s green partly because we haven’t used much green in our brand world yet.

But what’s interesting about everything you’re all describing—and what I’ve learned in my short time here at Every—is that language is fascinating to me, and the way model companies name themselves and their models is very literary. Codex is a book. There’s Haiku, Sonnet, Fable—these are all stories, or forms of putting words together. But the term people use for agents is “orchestration,” and that feels a lot more like music to me. I’m not especially musical, but I think everyone’s played an instrument at some point, and it feels like we can all take a step back—without knowing how to play the rest of the instruments in the orchestra, we can conduct it, as long as we understand what the piece needs to sound like. And you can be conducting several orchestras at once.

Creative direction and brand strategy—a lot of famous designers have described it as being like a maestro: you have color, type, shapes, symbols, forms, words, systems, photography, illustration styles, and you’re figuring out how to conduct all of it into something that sounds beautiful. I feel like this stack is that, for working with AI—and it applies to a lot more than just highly technical work. I use it for knowledge work too.

(00:30:00)

To answer your question: I do like Flora, so I can talk about that since I’m probably the deepest into it here. I’ve also learned, and I think Austin touched on this, what to use Claude for versus Codex, and how to use both together. I’ve built a set of marketing skills I use with compound engineering that really shortens the timeline on research when I’m working on strategy. You need data to back up an insight or a hunch. Do the tools come up with the insight on their own particularly well? Not always—sometimes they do. But you can give it the insight and say, “prove this out—what direction should we take this, here’s the competitive set.” It’s very good at competitive auditing, especially if you build a skill for it—you can push further and pull pricing, language, colors, whatever these brands are using. You feed that into a system and end up with a genuinely sharp strategic lens in hours, instead of what used to be a several-week process with humans researching.

From there, once you have a sense of what you want to create, a tool like Flora is really helpful—similar to Austin, I’m not a technical designer either, but I’ve been mood-boarding for fifteen years. You can productize a mood board now, especially with the right prompt—you can use Claude to write the prompt for you, put it into Flora, or use the MCP the way Brandon does, and have it analyze a set of reference images and tell you what you’re actually going for. With the Builder Pack, for example, it might say something like, “this is very high contrast.” On the note about green—this is, in a way, the most explicitly technical-facing audience we’ve spoken to with the Builder Pack. It’s for people who are really ready to start experimenting with all these tools, so it feels a bit more matrix-y and digital than the more Greco-Roman, heritage-driven parts of the Every brand. All of those were decisions these tools helped us land on, in terms of the whole aesthetic—and they also helped us design the website.

It’s been a mindset shift for me, moving from playing an instrument to orchestrating—once you learn how to get one group singing and playing, you move on to the next, and you can conduct several pieces at once. I’m just beginning to have everything plug into Notion—genuinely helpful. To Austin’s point, you can just say, “look in Notion, find this thing, do this for me,” and it has all the relevant context because someone on the team has automated it to keep updating itself. The way you work as an organization gets souped up to a crazy degree because we’re all orchestrating at the same time—it kind of hums. It’s been really exciting to work with these tools. It really is the sandwich: you have to put in a lot of work up front, doing real metacognition about how you think through a problem and what your process is. Once you’ve codified that, these tools can run with it, and you can ask them for help when you don’t understand why something happened—which I find genuinely revelatory.

Brandon

It’s remarkable—people say, “well, how do I use it?” and it’s like, “well, what are you trying to accomplish?” And then they say that, and I say, “great, just say that again”—to the model.

Austin

I think a good final thing here—if we’re speaking directly to people who either just got access to the Builder Pack or are thinking about it—is that this has all been pretty high-level thinking about how to work, with a few examples along the way. Let’s each say: if you had access to everything in the Builder Pack, what would you recommend someone actually do—one specific thing to try, or one mentality to take—if they’re a brand-new builder?

Brandon

If I’d just gotten the Builder Pack and was getting my feet wet, I think the first thing I’d do—if you’re really early on—is recognize that sometimes the hardest part about building is knowing what to build. You can sit at your computer and look for a problem, but that’s actually the worst way to find one. Going for a walk isn’t much better. You need to build the habit of noticing problems in your life, because usually we’re just numb to them. So if I were new to building and new to the Builder Pack, the first thing I’d do is pick something simple I really love and dupe it.

I’d build myself a very simplified MVP version, using as many parts of the Builder Pack as I could—launch it on Render using the credits, plug in PostHog so I could build in a loop or track how I use it and how my friends use it, build it with Anthropic, Claude, Codex, Cursor. I might work some AI into the product itself—even if it doesn’t need it, I might figure out how to add it. I might use Flora to make some unique icons or graphics. I’d just dupe something, because when you dupe something, you get moving on the process of making it, and very quickly you think, “eh, I don’t like that feature,” or, “I don’t like this version of the thing”—and you start to deviate. Very quickly it’s not duping anymore, it’s inspiration, and you’re off making your own thing. So if I were new to the Builder Pack, I’d dupe something, look through the pack, and figure out which of these tools I could actually start using in the build—and not second-guess myself for a second, just start doing it. Honestly, that’s the hardest part sometimes: just the first step, the first prompt.

Austin

(00:40:00)

I find that everyone learns differently—especially how to become an AI builder. Some people really want to understand what’s happening under the hood, and others just want to ship and learn from what doesn’t work. What worked best for me was: I actually had a lot of ideas for what I might build. Of those, I picked the one I was most excited about—for me, a movie app, like Fandango but for indie movies. I thought, “this doesn’t exist—if it worked, it would make my life a lot better, I’d be so stoked.” A good prompt to give yourself is: what would you be most excited to text someone about—“hey, I made this”? What would feel really good to send? I’d really encourage people to try to build something you don’t think you can—not necessarily complex for the sake of complexity, but something that stretches you—because you’ll learn so much through the process.

Building that app—which is now live, and a lot of my friends use every week—I’d never done any of this before. It walks you through: “okay, you need to connect Render, you need to go through Google OAuth,” whatever it is. Because it’s something you’re excited about and want to see through, it starts to feel like playing a video game, trying to get to the next level.

I agree with Brandon’s point about starting: connect the agent—whether it’s Cursor, Claude, or Codex—to the tools in the Builder Pack and anything else you have access to, and just say, “you have all this stuff, figure out where to go.” But what matters most is driving toward that goal of making something great enough that you’re excited to text a friend, “hey, go try this.” Along the way you’ll realize, “oh, it actually works, that’s crazy”—but the design looks bad, so how do I make it better? And you can sit and work through that—look at design work for inspiration, use the tools to check your instincts. That’s my best recommendation, and it’s worked really well for me. I’ve also learned a lot from seeing the first version the agent ships and thinking, “okay, that’s not good enough—how do I get this to a V2?”

Douglas

I have a good, specific one for marketers, writers, designers, creative technologists, and product designers. To Brandon’s point—I know your problem, and it’s probably your portfolio. This is a great stack to rebuild it with. It’s something everyone kind of hates about their own site, but you need it. And, to Austin’s point, you probably already have a dream version of it in your head. You can use Brandon’s duping strategy—find a site you really like and mimic it—or start from scratch, write the brief for what you’re looking for, add your references, and hand it to Codex or Claude connected to all these tools, and take it on as a real project.

It’s genuinely enriching, because most people aren’t happy with how they’ve merchandised their own work to the world, and maybe they want more interactivity or different visuals. You can use these tools to build something really striking that you feel good about—and that can also help you get more work or a better job, whatever you’re looking for. It also proves you’re AI-native. It’s a great project for anyone creative, or anyone running a personal side hustle or small business—do something impactful right away, because you can get people responding to it. To Austin’s point—his friends actually use his app now. People can use the thing, see it, and tell you how great it is, and that’s very encouraging as you start working with AI, at least in my experience.

Yash

To second that, or maybe close it out—what I’d do is use Cursor with Render. My mind was genuinely blown when I used Cursor cloud agents with Kimi K2, an open-source model that Cursor improved on. I’d say: dupe something, use Cursor cloud agents, and try different models. If this is what you’re going to be doing, you need to find your favorite model. With cloud agents you can run ten agents at once—we’re on the Pro Plus plan, above Pro, so you’re covered on spend. I’d use Render to deploy it and actually figure out which model you like working with. I’m picky enough that even when a new model drops, I don’t switch right away—if I already have a preference, I’ll test the new one out first before deciding.

You need to basically run your own evals—see whether a model’s doing better or worse for you—and build your own sense of what you like and don’t like. I think cloud agents are a really underrated way to figure out what you actually like to build, and to start feeling that out.

Brandon

Cool. In terms of what’s next for All Access—we’ve got some good stuff coming. In the next few days or weeks we’ll probably drop two more partners with pretty great benefits. And something that’s been really compelling for our readers, I’ve found, is that once they have access to these benefits, they love the guides on how to actually use each one.

Douglas

Yeah.

Brandon

So I think what we’ll probably do next is, for each of these benefits, pair the value you can capture from using it with a guide you can plug straight into one of your agents—something that leapfrogs you ahead with whatever net-new technology you now have access to. So: guides, or some kind of camp for each of these partners, is probably something we’ll build out.

Austin

Yeah. To be clear, I’ve never actually read one of our guides myself—I drop the guide into Codex and say, “use this to help build the next thing we’re doing.” I’d recommend everyone do the same. And I think the best way to think about All Access, and about getting the most out of an Every membership generally, is that we’re going to keep adding partners to the Builder Pack that we’re excited to use throughout the year, since you get these as part of a yearly subscription. We’re going to keep filling it up with more stuff you’re excited to use—there’s a lot more coming that will help you build.

Brandon

Awesome. I think now we’re handing it off to Dan’s outro. Dan, take it away.