Finished MVP
by Andreas TzionisIt took 11 weeks and a lot a helpless fights but finally there's something that can solve a big issue. I coded the whole thing in React / Node.js and send it over to Cloud Run with a Firestore database. Used Cloud Functions with Google analytics API to receive yesterdays traffic data in Slack. There were a some features such as MySql and Nodemailer integrations that I could have worked on yet I'm excited to start talking to customers and see what they think about the product so far!
-
2
That is an amazing job. Completing MVP is a great step towards validation and acquiring your first customers.
I would like to know when you mentioned 11 weeks it took for you to complete your MVP, have you implemented several features? Or just concentrated on a few of them that really matter?
The reason I am questioning this is that experts recommend sticking on to developing the most required feature of your product.
What are your views on this?-
1
Hi Reuben,
Thanks for getting in touch. 11 weeks is a long time and I was lucky to have some free time in that period.
I found out that it helps if you have a very specific product goal in mind. For me and my app it was to be able to "made a CRUD backend with no code in 3 minutes".
I drifted off a lot from that. I made it "manage Stripe Payments with no code" and "handle authentication with no code" that I probably shouldn't and I could have finished in 5-7 weeks.
On the other hand, this project is something that I really proud of but I wouldn't recommended it for anyone else. Build-Measure-Learn is the way to go!
Andreas
-
-
1
Thats a great achievement! Congrats!
Can you share some insights on using firestore at global scale where users can be from all parts of the world?
We see firestore having lots of delays. Can you share what are your response times and where to have firestore location when targetting users globally?
-
1
The landing page looks awesome! Very well done!
-
1
There is a typo on your faq section talking about pricing. You can also remove the x mark and make it something else on the pricing section.
The menu upon clicking navigates to a section but the menu itself doesn't close.
Last and foremost, you have a great product dear IndieHacker!
-
1
It’s a really good idea. Do you think that the usage of Firebase will bite you later given (imho) the pricing doesn’t scale really well?
-
1
Thanks. That is a great question and one that have been puzzling me for a long time. Firebase pricing is very different from traditional computer instances costs.
Diagram uses Firestore as my main database and Cloud functions for some analytics and reporting but that's all it has. The backend is a containerized Node.js app that runs on Cloud Run.
The current subscription model supports about 100 to 1 free to paid users (in my estimation) and that's mainly due to Cloud Run costs. If your business model supports it, I think Firestore (with Cloud Run - not with Firebase) is a great no-initial costs database solution to bootstrap!
-
1
Yeah Firestore to me seems the one that could scale without control. Are you creating a Firestore for each backend? Or there is only one Firestore where each collection is a backend?
-
1
That's the one!
There is one firestore for the app that stores the users and their backends. One the other hand when a user creates a new backend they can assign their own firestore instance to their backend so they can have full access to their data.
Currently, it only supports firestore (since was my favourite db :) ) but I am looking to add support for MySql and MongoDB if it is requested!
-
1
I see. That's a curious choice though. The winning point for Firebase regarding Firestore is that you don't need a backend to use it :D
Anyway, good luck!PS: it's not my intention of bashing your product, I was just curious about the architectural choice.
-
-
-
-
-
1
Wow, nice project! Had a quick play with the UI and seems like a cool MVP!
Will have a proper play later 👍
-
1
Thanks!
We were having a go with a friend earlier today. He was happy as well with the landing page but had some difficulty getting started with the main app.
His feedback was very constructive and I’d love to hear more when you get the chance to play more with it 💪
-
-
1
Congrats Andreas, that’s a huge milestone. All things considered, 11 weeks is a good amount of time.
I also don’t get too lost on the no-code stuff. There are many diminishing returns to that especially if you’re fully technical
Way to go and good luck!
-
1
That is very true. When I used a nocode project I aim to automated one thing such as recurring payments, user auth or CRUD without disrupting my code and keep the data in my database if possible. It’s a funny thing that this is one of the problems I was trying to solve with Diagram. It’s a hard problem that I think for technically easier projects no code can help and for more challenging projects code only is a no brainer!
-
-
1
This comment was deleted 5 years ago.
