GoodApp Labs
Month one: time flies when you're having fun!
TL;DR: I can't believe it's been a month since I started GAppL. We're on track or slightly ahead of where I wanted to be. The business is fully set up, our first app is in App Store review, and half of HQ's toolbox is built. I didn't expect to be working on our 2nd and 3rd projects yet, so that's bonus progress.
What we got done this month
1The business
All the admin and infrastructure is taken care of: registrations, taxes, a business bank account, equipment, and moving to an organization account with Apple. We also announced the LLC on LinkedIn and YouTube. It's wild to think that a month ago I still worked at Hope Street and the LLC didn't really exist.
2HQ, our internal hub
We built HQ to run our internal operations. It hosts 7 tools across 3 categories:
- Designing and building products
- Product Idea Hub: generates product ideas
- Journey Maps: maps out user journeys
- Dream in a Jar
- Dashboard: tracks our success metrics
- Launch checklist: pre-launch tasks shared between me and our agents
- Running the LLC
- Expenses: tracks spending and tax notes
- Agent Usage: tracks what our agents cost
- Founder Time: analyzes how I spend my time
3Dream in a Jar
Dream in a Jar has been a joy to build. It's ready to ship and is in App Store review right now. Our main goal with our first app is learning: how to market, how to launch apps, and how to build with AI. By that measure, the project has already been a huge success! Of course, we're also hoping it succeeds in the market:
If you haven't seen it yet, here's the demo I posted two weeks ago:
4Projects #2 and #3
We've started exploratory designs on two more apps:
- NewJeff makes it simple for new cooks to build a cooking habit.
- An Ivy League coach (name TBD) helps A-students learn how to get into Ivy League schools.
Where AI shines, and where it struggles
We're starting to see which parts of building a product AI can handle on its own. Overall, it feels pretty good: big parts of the build can be handed off to agents. The big unknown is automating marketing. I think that's what will really make or break the business model.
Here's how much of each stage AI can handle today:
- Scouting100%
AI collects raw ideas and feedback from forums and app stores completely on its own.
- Inventing ideas66%
Still a struggle. AI gets close on the broader space, but picking out the truly insightful, magical opportunity is hard. I'm optimistic more tuning will get us there.
- Research95%
AI mapped user journeys surprisingly well, better than I could have on my own. Light feedback takes care of minor issues.
- Design50%
AI can't build good wireframes on its own and needs a lot of input. Luckily, we're good at this and enjoy it, and AI makes it much faster, more iterative and more fun.
- Build75%
Agents do basically all the building, but higher-level technical decisions need my testing and input. I hope to automate more with product manager agents that oversee the builders, though it hasn't slowed us down much so far.
- Marketing & iterationTBD
Not started yet, since nothing has launched. This is the big one.
Where my time goes
One reason I was excited about this experiment is having AI handle the parts of the work I don't enjoy, so I can focus on what I enjoy and do best. Here's how I spend my time today, compared with where I'd like it to be long term:
| Area | Now | Ideal | Notes |
|---|---|---|---|
| Idea evaluation | 5% | 5% | |
| Design | 5% | 40% | The main change I want in month 2. It puts pressure on producing more high-quality ideas and research. |
| Managing the build | 30% | 20% | Testing the product and problem-solving with coding agents. |
| Growth & marketing | 10% | 20% | We haven't started marketing, since nothing has officially launched yet. |
| HQ tools | 25% | 10% | The early days need more time on tools and infrastructure. I expect this to drop over time. |
Next month
- Build our marketing engine. Marketing Dream in a Jar is a great chance to set up our marketing operations in HQ. I'd feel really good if, by the end of next month, we have a highly automated marketing engine, with Dream in a Jar as the first mold. This is probably the single most important open question for the business.
- More time in design. That means improving the volume and quality of our product ideas. More inspiration means more design.
- Decide on Dream in a Jar. By the end of the month, I hope to know whether it's worth more energy, especially on growing virality. I'm also hoping to see a few tips come in, even if it's just a few bucks.
- A first build of the Ivy coaching app. I'm prioritizing it above NewJeff, even though NewJeff feels more fun, since I have a chance to test it with a high school student in December. The goal is a working intake and first experience by the end of the month.
David