What If ChatGPT Sends Customers to Your Competitor?
AI search is changing how people discover businesses.
Every day, millions of people ask ChatGPT and Perplexity for recommendations. If your business isn't showing up, someone else is getting those customers.
AutoSEO helps your business rank in AI search by publishing expert content, building backlinks automatically, and improving your online visibility—without hiring an SEO agency.
On paper, Diego Roshardt left college with nothing. No job lined up, no internships, no successful side projects - while his friends walked into corporate roles.
What the paper missed was the education underneath. Diego had spent his college years learning to code from YouTube and shipping project after project. Most failed. Some made money intermittently. All of them taught him something that no internship covers: how every part of a business actually fits together.
A year ago, he quit his software engineering job with only a few months of runway, moved into a studio apartment in Austin, and went all in on AppAlchemy - a browser-based tool for building native mobile apps without downloading Xcode, owning a decent Mac, or fighting through complicated setup.
He launched the first version in about two weeks. Today, AppAlchemy has thousands of users and generates more than $10k in monthly recurring revenue.
Here is how he did it, including the most specific Reddit marketing playbook we have featured in this newsletter.
The Failed Projects Were the Degree
Diego knew at his first college career fair that the traditional path was not for him. Launching scalable online projects sounded more exciting than anything on offer at the booths, so he taught himself to code and started building immediately.
The results through college were unremarkable by any conventional measure. But his read on that period is worth sitting with: he does not consider any of it wasted time. Building complete projects - even failed ones - forces you to touch every area of a business. Product, marketing, pricing, support, distribution. There is, in his words, no better way to learn all of those skills at once.
After graduation, he cold-emailed a startup founder with a portfolio of launched projects and got hired as a software engineer. The job lasted a few months. Something felt off - he was unfulfilled, and he knew he wanted to build something of his own. But this time, he decided, he would do it differently than the college projects that went nowhere.
Spotting the Gap Lovable Left Open
The timing of the idea mattered. When Diego went all in, products like Lovable were going viral on Twitter - AI tools that let anyone build web apps from a chat interface. What he noticed was the gap: nobody was doing the same thing for mobile apps.
He knew the pain firsthand. Creating mobile apps traditionally requires downloading Xcode and owning a reasonably powerful Mac just to get started. The setup friction alone filters out most people before they write a line of code. A browser-based alternative removed that entire barrier.
So he built a simple MVP and started posting on Reddit to test interest. With only a few months of runway, there was no room for a long exploratory phase. It had to work, and the only way to find out was to put it in front of people immediately.
Two Weeks to Launch, Charging From Day One
Diego launched the first version in about two weeks - a pace he credits partly to having far more mental energy available than he did while employed. The focus of full-time commitment compounded his speed.
The two-week timeline was also a corrective. On past projects, he had spent too much time building features when he should have been marketing - and the features he did build were based on his own assumptions rather than what actual users wanted. This time, he flipped the sequence: launch fast, post immediately, and let real users drive the roadmap.
He also charged from the start - monthly subscriptions with discounted yearly plans. His reasoning is one of the cleanest statements of the principle we have seen: paying customers are what ultimately prove demand for a product. Free signups measure curiosity. Payment measures need.
Growth was slow at first, with a few sporadic paid users. That is the normal, unglamorous beginning that most success stories edit out.
The Reddit Playbook
Reddit has been AppAlchemy's primary growth channel, and Diego's method is specific enough to replicate. His core argument: Reddit is underrated because it solves the targeting problem that kills most early-stage marketing.
The standard SaaS advice says launch on Product Hunt and similar directories. But if your target customer is not an entrepreneur or early adopter, that audience is the wrong one, and the results show it. Reddit has a subreddit for every imaginable niche - which makes it the best way to reach the specific people your product actually serves.
His method, in three parts:
Never post ads. Most people fail at Reddit marketing because their posts are pushy and feel foreign to the platform. Instead, provide genuine value in your niche, then casually mention your product in the middle of the post - not the headline, not the close.
Become an active user first. Join subreddits related to your niche and to things you are genuinely passionate about. You learn what kinds of posts do well, and you warm up your account in the process. Posts from brand-new accounts almost always get filtered out automatically.
Show, do not announce. Diego posted product updates, videos of himself actually using AppAlchemy, and commentary on vibe coding and AI tools. The content was interesting on its own; the product came along with it.
Beyond Reddit, word of mouth did significant work. Users who liked the product shared it on social media and with friends. And last summer, a major Twitter influencer posted about AppAlchemy, producing a significant influx of users overnight - the kind of moment you cannot plan for, but that only happens when the product is already good enough to talk about.
A Stack Chosen for AI Compatibility
The original version was a simple React MVP that let users chat with an AI to build and preview a mobile app. The current stack:
Next.js as the main framework
Supabase for authentication and database - a recent switch from Firebase, which he describes as a clear improvement
Claude Code and Cursor used extensively for development
The selection principle is worth noting: because AI tools write a large share of his code, he deliberately chooses open tools that AI can easily interact with. Tooling decisions that once hinged on personal preference now hinge on how well an AI assistant can work with the stack. That is a quiet but meaningful shift in how solo founders should evaluate their choices.
The Isolation Problem
Diego's biggest challenge so far has not been technical or financial. It has been isolation. Indie hacking gives you enormous freedom, and the cost of that freedom is spending most of your time alone.
His retrospective on this is a reframe of common advice. Most people recommend building in public to get users. Diego thinks the real benefit is different: building in public is how you meet like-minded people working on interesting projects. If he started over, he would share far more of his journey online - not primarily for distribution, but for community.
"Surround yourself with people who are building things. It makes the journey more fun, and you'll learn a ton."
He adds one more piece of advice that founders rarely discuss: eat clean. Indie hackers make important decisions every day, and in his experience, eating badly robs you of the mental clarity and energy those decisions require. It sounds minor next to marketing strategy and tech stacks. He does not think it is.
What Comes Next
Diego's goals from here are refreshingly uncomplicated: grow AppAlchemy, level up as an entrepreneur, and enjoy the process along the way. After a college full of failed projects and a year of proving the model works, the plan is simply more of what is working.
You can follow his journey or connect with him on Twitter. And if you want to build mobile apps fast, without Xcode or setup friction, take a look at AppAlchemy.
Stop losing Sundays to SEO
Over 2,500 businesses show up in Google, ChatGPT, and Perplexity with AutoSEO. No blog posts. No backlinks. No agency reports. Just set it up and step away.
