Podcast Summary: Marketing Against The Grain

Episode: I Used Gemini Code Assist to Build a Newsletter App (for Free)
Host: HubSpot Media
Date: November 6, 2025
Speakers: Kipp Bodnar (HubSpot CMO), Kieran Flanagan (HubSpot SVP of Marketing)

Episode Overview

In this episode, Kieran Flanagan dives deep into how Google's new AI coding assistant—Gemini Code Assist—enables rapid creation of code-driven marketing experiences, specifically demonstrating how he built an app to capture newsletter subscribers. The discussion moves rapidly from the tactical steps of app creation to broader implications for marketers: interactive tools are becoming as easy to make as static content, shifting how marketers capture leads and engage users.

Key Discussion Points & Insights

1. Rise of AI-Powered Code for Marketers

Trend Overview:
- The episode opens with the shift from static content (e.g., PDFs, ebooks) to code-powered, interactive tools as a lead generation mechanism ([00:00]).
- Notable Insight:
  - "It's as easy to create kind of interactive tools and interactive content as it was to create PDFs and static templates and Word docs or whatever else." – Kieran ([03:01])

2. Google AI Studio & Gemini Code Assist Overview

What is Google AI Studio?
- An integrated suite of AI tools, including image generation (Nano Banana), video (VO3 1), and now, a coding assistant.
Benefits of Integration:
- Seamless combination of different AI services, e.g., building an app with instant access to image and video generation tools ([05:25]).
- Example:
  - Kieran describes building a fitness coach app capable of having voice conversations.
  - “...It is best to chunk it into parts so the assistant does not get too confused. Now I will say Google's AI coding assistant seems to be able to just one-shot things." ([13:05])

3. Building the Newsletter Subscriber App

The ‘Cloud Skills Generator’ App:
1. Problem: Traditionally, growing a newsletter required creating gated content to collect emails.
2. Solution: Use Gemini Code Assist to build an app where users submit their job, get relevant “Claude skills,” and then, in exchange for their email, can download those “Skill MD” files ([10:29]).
Process Highlights:
- Kieran collaborated across AIs (Claude for prompt expertise, Gemini for code execution), showing a “stack” approach.
- Critical app features were developed step-by-step to ensure clarity: upload job → parse skills → generate properly-formatted Cloud Skills → enable download, gated by email ([14:42]).

4. Lead Capture Is Now Productized

Email-Gated Downloads (Live Demo):
- Kieran demonstrates how adding a download email overlay turns the app into a lead generation engine.
- “Now it has a popup to capture interest in this for my newsletter. And that’s what I mean. That’s the power of these code-powered experiences.” ([18:40])
- The pattern has flipped: Rather than a newsletter growing an app audience, the app itself becomes a tool for newsletter growth.

5. Tactical & Strategic Tips

Chunking tasks: Break app features into manageable pieces for better AI assistant performance ([13:05]).
Cross-tool prompting: Don’t hesitate to use different AI assistants for their strengths—Claude for specialized domain prompts, Gemini for execution.
App-first marketing: “You can really create code experiences as quickly or even faster than you can create content.” ([20:19])
Practical Prompts Offer: Kieran teases access to 10 ready-made prompts for Gemini code to build marketing tools at various buyer journey stages ([12:30]).

Notable Quotes & Memorable Moments

On Paradigm Shift:
- "What’s happening now is these code and assistants mean it’s as easy to create kind of interactive tools...as it was to create PDFs and static templates and Word docs." – Kieran ([03:01])
On Building With AI:
- “You have to really get experience about using AI across ChatGPT, Claude, Gemini, all these different tools. Look, we’re giving you 10 prompts you can plug into Gemini code right now to build interactive tools that actually convert.” – Kieran ([12:10])
On Power of Productized Lead Capture:
- “Now I can kind of market this app and I can create actual email signups for my newsletter. In what world prior to AI would I have created an app to grow a newsletter?” – Kieran ([19:04])
On the Future:
- “Google are rumored to be launching Gemini 3 very soon and that is going to be apparently an even better coding model. And at the moment this is using Gemini 2.5.” – Kieran ([21:21])

Timestamps for Important Segments

00:00 — Intro & context: The shift from traditional content to AI-generated code-powered experiences.
03:01 — Why marketers should care: Interactive tools as easy as static content.
05:25 — Exploring Google AI Studio’s powerful integrations.
10:29 — From gated ebooks to interactive skills generator apps.
12:10 — Cross-AI workflows and 10 ready-made Gemini dev prompts.
13:05 — Chunking app features for AI assistant effectiveness.
14:42 — Step-by-step: Building, testing, and refining the skill generator app.
18:40 — Live: Adding a newsletter signup overlay to the app.
19:04 — The paradigm flip: App drives newsletter instead of vice versa.
21:21 — Future outlook: Gemini 3 and unlocking even more app creation potential.

Episode Tone

The episode maintains an enthusiastic and pragmatic tone, balancing practical “how-to” tips with big-picture implications. Kieran’s language is informal, fast-paced, and candid, making the concepts accessible even to non-technical marketers.

Final Thoughts

Kieran’s walkthrough gives marketers actionable steps to embrace AI code assistants, democratizing app creation and reframing the process of building audiences and capturing leads. The future isn’t “content-first, sell a tool later”—now, marketers can flip the script with easy, interactive experiences tightly integrated into the lead engine.

Interested in the app or prompts mentioned?
Kieran invites listeners to comment for access and promises to release the tool to subscribers.

This summary captures the flow, practical advice, and the excitement of leveraging Gemini Code Assist for audience growth—all in the original voice and spirit of the episode.

Transcript

A (0:00)

Google have just launched an incredible AI coding assistant. Now, if you want to start to build applications, code powered experiences to help grow and market your business, this really is the episode for you. If you're thinking, how do I even grow an audience for my business anymore? How do I even get customers for my business anymore? Using these tools to create code powered experiences, apps that will capture demand that then you can convert into customers for your business, is going to be a tactic that you want to do. And Google's new coding assistant is one of the best ways to do it. I'm going to give you an app that I built, how I thought about building that app and how I'm going to use it to capture demand and grow an audience with that app. All of that and more on this episode of Marketing against the Grain. All right, I've been playing around with Google's AI coder. It's part of their AI studio. If you have not checked out AI Studio, it is a suite of AI tools from Google and it really is an underappreciated thing. It's an incredible, powerful suite of AI tools and they've just added AI Code builder. And what I want to show you is how easy it is now for people who want to market and grow an audience for their business or newsletter or services, how easy it is for them to create code powered experiences. So in the past we would create content, we would market that content and we would get people to consume that content. We would probably capture some contact details through a form and then we would try to convert them into customers. What's happening now is these code and assistants mean it's as easy to create kind of interactive tools and interactive content as it was to create PDFs and static templates and Word docs or whatever else. And I want to show you an example of a code powered experience I'm creating to market my newsletter to kind of bring you through how I think about this and then how you can kind of replicate this and then use a tool like this, Google AI Code and Assistant, which is actually very, very impressive to do the same for yourself. So let's start off with the Google AI studio. They have their Nano Banana, which is their Image Gen tool. It's very, very good. They have their VO3 one, which is their new video tool. I'm currently working on a mini sitcom using VO3 1. And so I will come back and do an episode and show you what I've done and bring you through that. So they have a ton of tools in here. What we're going to go into is build. The first thing you'll notice is it's actually integrated to Google's suite of tools, which it should not be underestimated. So if you want to build an app that has some image capabilities, it will just integrate with Nana Banana. If you want to create an app that's integrated with voice and ads in Gemini Lives API, which I have done. So I built this kind of fitness coach here, and this fitness coach can basically talk me through a plan each and every morning when I get to go to the gym, which is actually not too often given my daughter does not sleep that much. So it's like a little bit fragmented. But when I do get to go to the gym, I can basically just have a voice conversation and talk to my little coach. So you can see the value actually of being integrated here. You can integrate with VO3 if you want to actually have imagery. So lots of cool things here. And the app I want to show you is this Cloud Skills generator. And so I'm going to give you a few things that I've realized using Google's AI Coding assistant. So, first of all, how I think about things. One of the coolest features launched very recently was Cloud Skills. I did a whole episode on it. Cloud Skills allow you to create these little files, they're called Skill MD files, which are basically a prompt to teach Claude a skill. So if I'm me, I might want a skill to do competitive research. I will want a skill to do strategic memos, I will want a skill to do some sort of data analysis, right? A bunch of different skills. If you're a content creator or a salesperson or whatever it may be, there's tasks that you can do that you do each and every week that you can teach Claude to do for you. And then what you have is these little prompts that you upload to a Skill MD file, give it to Claude, and now it can use that skill whenever you're asking it a question that's relevant to that skill. Now, if you want to understand more about Claude skills, again, we did an episode. It's on the YouTube channel. Go check it out. So my newsletter audience is going to be very interested in Claude skills. In a world gone by, if I really wanted to go all in in Cloud skills for that audience, I would create a gated tutorial, right? I would create a whole ebook and I would put it behind a form and I would say, okay, now, if you want that ebook, it's really valuable. Download that ebook through a Form, give me your details, and it allows me to market to them. And if I had something to sell them, that's the way I could do it. Now what I can do is I can do much, much better, right? I can create an entire app to just allow you to upload your job, and that app in return will give you back those skills that you can just upload to Claude. And I did this through Google's AI code and assistant and I was really, really impressed. Very impressed. And a couple of tips. You can just work directly in this panel and ask the AI assistant to like build you things. What I did for this, because it's Claude skills, is I actually worked with Claude to actually build the prompts for Google because Claude would have the best knowledge about its own documentation and guardrails. And you know, how you actually build these skills in the right format? You don't need to do that. But I have quirks where I use different assistants for different things and I could actually use one to prompt another assistant. So you have to really get experience about using AI across ChatGPT, Claude, Gemini, all these different tools, look, we're giving you 10 prompts you can plug into Gemini code right now to build interactive tools that actually convert. Think of these as your dev team in your pocket. Each prompt builds a different tool for a different stage of the buyer journey, turning your tools into a lead gen machine if you want it. Scan the QR code or click the link in the description. Now let's get back to the show. All right, the other thing is, when you're creating an app, it is best to chunk it into parts so the assistant does not get too confused. Now I will say Google's AI coding assistant seems to be able to just one shot Things. When I say one shot things is like you can give it a pretty complex app and it can build it all from scratch with a multitude of different features and it doesn't get too lost. But to make sure this worked, I first of all built the functionality to upload a job or paste a description into this panel and it would parse out the different skills that could be created as cloud skills. And then I built the next feature which was converting those into actual Claude skills in the right format, which I'm going to bring you through. And then the last one is to be able to download those. And we're going to build a new feature here live to see how well it works. So how does this work? Well, let's say I can upload a job description. So let's do that. And so you can see I uploaded the job of a content writer and then it's given me all of these different skills. And so you can see each skill has a gradient for impact, the effort and then what that is for write in communication see analysis analyze competitors top content. That's a pretty good one. And so we would have the rationale for this. A competitive analysis identifies industries best practices. As a content marketer I want to analyze and competitive content to inform our content strategy. Talks a little bit about the skill and look at this. So now I can go to create and it creates the entire skill for me. How cool is that? Right? How cool is that? And then I can either copy this to clipboard and just create the scale MD file myself or I can download this now at the moment I can just download that skill. What I would do when I'm kind of building this app is I would add a little form. And if you want to download that skill, you have to give me your email address. Why don't we see if that will work? And let's do that. When a user goes to download a skill MD file, let's create a overlay form that asks for their email address with the following text. So we're back into our app here. We're in our app. Let's say we want to. It's what skill do I want? So many great skills, research industry. I come in here and then I want to download the skill. And look at this. Now it has a pop up to capture interest in this for my newsletter. And that's what I mean. That's the power of these code powered experiences. Now I can kind of market this app and I can create actual email signups for my newsletter. In what world prior to AI would I have created an app to grow a newsletter? Right? It would always be the other way around. An app would have a newsletter to actually grow it, but now it's been reversed because you can really create code experiences as quickly or even faster than you can create content. That's an example of Google's AI coding assistant. It's really powerful. I would definitely go try it out. Google are rumored to be launching Gemini 3 very soon and that is going to be apparently an even better coding model. And at the moment this is using Gemini 2.5. So this should actually get much, much better when Google launched Gemini 3. So that's an example of why you should build codepad experiences, how valuable they are. I hope this was helpful to you. I will actually be deploying the app at some point and giving you all access to create your cloud skills. If you want that app, put a comment in the description. I'll make sure I ungate that and send it to all of our subscribers. Until next time, Bye Bye Sam.