Hello Everyone! Thanks for following me on this journey so far. It’s been really fulfilling sharing the journey of building Coco AI this way and I’m glad I decided to start writing this series of articles.
For those who are finding this article first, this is the fourth of a series of articles I began writing to document the build process for Coco AI. An intelligent mobile app that helps social media managers focus on creating content while taking away all the other tedious stuff like scheduling and coming up with ideas for different platforms.
I started building this app to solve a problem I had which was coming up with content to promote the apps I build on different social media platforms. Coco AI solves this problem by ensuring I always have ideas tailored to different platforms on demand and it has a calendar interface for scheduling as well.
I leveraged the notJustHack Hackathon which challenged indie mobile app devs to build and publish an app to either app store or play store within the month of December.
Although we were given the first two weeks of January to pass the app review process from either app store or play store, I unfortunately didn’t meet the deadline.
I chose to publish to play store because it is cheaper to purchase a Google developer account ($25/once) compared to an apple developer account ($99/year). I knew Google had more restrictions but I was confident I could scale through within the allotted time but there were unforeseen issues which I’ll be sharing in this article.
In the previous article, I shared all about how I built the app, the coding aspect mainly. In this article, I’ll be sharing what went down after I finished building Coco AI, my struggles with publishing and also my plans for Coco AI going forward.
My goal with this article is to give a progress update for those who have been following from article one. There’ll be things to learn, especially if you’re also a mobile app dev based in Nigeria.
So, let’s gooooo.
The first issue I ran into while preparing to publish the app was with Gemini API. Coco AI implements generative AI technology using Gemini flash model by Google to generate content ideas for different platforms for social media managers.
I ensured that Coco didn’t give the content completely just suggestions so that social media managers are not starting from a blank page when they get to work and they still maintain their creative human touch for the actual posts.
The issue I ran into with Gemini API was that since I was on the free tier, all of a sudden functionality started working intermittently. Most times it worked fine but sometimes it returned invalid API key error. For a product I was going to publish to the app store for general use this was a big issue.
The solution was to get on the paid tier and enable billing for the API but alas there I found my second issue.
All of the debit cards I tried to use to pay for the API didn’t work. I also tried virtual cards but no luck. Prepaid cards were no good either.
Eventually, I found a solution with Access Bank dollar card but it was a friends’ own and I knew if I’m thinking long term I needed to get mine. I tried to go through the process as quickly as I can but it still ended up taking me almost the entirety of January to get my Access Bank dollar card.
As it turns out, to be a publishing app dev from Nigeria you need a reliable dollar card because it’ll be requested at multiple steps of the journey. It could be to pay for an API or to purchase a dev account.
Once I get my card, I’ll be able to carry on with the process of enabling billing for Gemini API and publishing my app on both app stores.
So there you have it, the hackathon officially ended on the fifteenth of January and even though I completed the app I couldn’t complete the hackathon due to the limitations of being a Nigerian without an international dollar card. I see this as a positive though because if I didn’t go through this roadblock I wouldn’t have my card now and I’m sure I’ll be publishing many apps going forward thanks to a better understanding of what’s required.
Now that the hackathon is over, what’s next for Coco AI??
The plan was never to publish the app and forget about it. One of my goals for the year is to have a software product with at least a hundred paying users. That’s a tough ask but I’m fully committed to the vision and I intend to make it happen.
How I intend to do that for Coco AI is to first speak to at least thirty social media managers and know their critical pain points so they can pay me to solve it with Coco AI. So while I struggle to get everything ready to publish the Coco AI app, I’ll be simultaneously speaking to social media managers and tweaking what I’ve built currently to better serve them.
That’s what I have in front of me. Tweak the app to ensure it’s solving the most pressing needs for social media managers and ensure they’re more productive using Coco AI and also get everything ready to ensure I can publish mobile apps willy nilly.
In the next article of the series, I’ll share my learnings and updates on what has transpired in my bid to get my first couple of paying users.
Thanks so much for reading this article. I’ll keep writing them and let’s see what the future brings. Ciao!