Questgen

In this Bubble app example interview Matt speaks with Ramsri founder of Questgen the AI powered quiz generator. We discuss in details topics such as Bubble vs traditional coding, user acquisition, pay as you vs subscription plans, MRR and growing app profitability.

Host

Matt

Guest

Ramsri

00:05
Matt

Today I'm joined by the founder of Questgen.ai. Would you like to start by introducing yourself, maybe saying where you are in the world and a little bit about how you've got yourself to where you are as a Bubble builder?

00:19
Ramsri

Sure, sure Matt. Thanks a lot for inviting me on here. First of all, I would like to introduce myself. So I'm Ramsri, I'm currently based out of Hyderabad, India. Prior to this I worked and also dabbled with startups for about ten years. So I graduated ten years back, stayed in the US for six years, then moved back to India and stayed for a little while in Singapore and then I'm currently back in India.

00:47
Ramsri

And coming to my background as such, I'm an AI developer mostly so building AI products and mostly on the coding side of things in the AI space. Most recently in the last two years or so, I dabbled into full stack development, especially with no code tools like Bubble because as a developer I wanted to build full scale apps that are out in the production setting and payments and everything. I wanted to do the fastest way, and that's how I discovered Bubble. And I think we can talk more about the journey as we progress further, but that's my short introduction.

01:31
Matt

Great, thanks Ramsri. Would you like to say what is Questgen like your elevator pitch? Tell us about the product you've launched.

01:42
Ramsri

Absolutely. So, Questgen is an AI powered quiz making app. For example, if you have middle school textbooks, then at the end of every chapter you have these assessment or quiz questions like multiple choice questions, true or false, fill in the blanks, all these kind of quizzes. Can you just give text and in one click generate these multiple choice questions, true or false using AI? That's exactly what Questgen solves and that's what I've built with Bubble and a fine mix of AI like GPT-3 and other stuff. Primary use case is teachers, schools and tech companies who want to create assessments or quizzes at scale.

02:36
Matt

And have you been building this Bubble app by yourself? Do you have a small team?

02:41
Ramsri

Yeah, so I've been building the Bubble app myself since the last one and a half years, since I launched. In some complex cases where I wanted to implement some complex drag and drop modules etc, I took freelancing help from Bubble forum, etc. But beyond that, mostly it's 95% of my work and when things get too complex, I take some expert paid help.

03:11
Matt

Okay, and do you mind sharing about your funding situation? Bootstrapped, investors, what are your thoughts on that?

03:20
Ramsri

Sure. The Questgen is solo-built by me and it is completely bootstrapped as of now. Currently there are 23 paying customers and $730 MRR. So about two or three B2B clients and then rest of them are B2C clients who use the app. I've been building Questgen on the side, almost like learning Bubble, learning marketing side of things and slowly experimenting and building and scaling it.

03:57
Ramsri

I launched subscriptions actively this March. Prior to that I was on a pay as you go model and that model is also because of my limitation of understanding Stripe subscriptions and implementing them in Bubble. I know how to just get $10 payment and add 100 runs or something like that. So I implemented pay as you go. But March I sat down and I went through all the tutorials available on planet Earth for Stripe and Bubble and finally got it done.

04:33
Matt

It's always good to hear about a start up that starts making some money. Would you mind sharing that journey that many founders are quite intimidated by which is getting the first paying users. What was it like going from maybe getting to an MVP and putting your app out there to getting that first dollar in? How long was that, what were your strategies?

05:05
Ramsri

So I think ideating the app and then building it and getting the first dollar, probably it took like three months or so, probably one and a half months of Bubble development. And my app when it launched the MVP was pretty simple. There is a text box, there is one click button and I show MCQs. And those MCQs are also not fancy. That means there's no drag and drop nothing. It's just another text box where it's just formatted like an MCQ, just some question and some text. That's it. That was the MVP coming to the idea as such.

05:46
Ramsri

I've been working on the idea for about a year and a half prior to the launch of the app even, but mostly like an on end of independent research project. I wrote some blog posts about how you can do question generation, quiz generation in etc space, so some technicality side of things etc for so that gained attention and from Medium blog post people started reaching out to me asking for hosted solutions and things like that. Then with few interns I put out a free open source library, GitHub library. So basically for those who are unaware, it's just like a packaged code where the research is packaged into small modules and it's open source for everybody to use.

06:37
Ramsri

And another interesting thing that happened was I also created a course. So making an app was in my last arena because I did not have the full stack skills. So I wrote blog post then I created this course on Udemy about question generation and everything and my developer life ended where I created an API where anyone can pass in some text and they get back some modified text with questions. It's like an API stands for Application Programming Interface over internet. You send some text and you get back. That was it. But then I was following the Twitter space and LinkedIn mainly these two social media apps and there was no code movement, picking up the tools like Softr, Bubble, Webflow. These things came up then. Primarily I am like an API developer and that's very intense. So I was curious to see who can provide a user interface for my API. Pretty simple. I put in text, there is an API call behind it and some text is returned back. Show the text on the right. Forget about payments. Forget about anything else. Can someone do left text box and right text box output with one click button?

08:03
Ramsri

I was trying to see that and I discovered Bubble was the kind of the best tool available, at least at that point of time and mostly even today because of the community and even when, to be very honest, I tried Bubble then I kind of you know, the learning curve seem little complex.

08:26
Ramsri

In the beginning it's just like driving a car. The first time you drive a car and you see someone else driving the car, you will think that how can they drive with music and people talking in the background? So many things coming in, how are they able to even concentrate? So it was just like me where when I first got into the driving seat of building a Bubble app, everything looked too complex. I even kind of paused it for a good month or so. Then something magical happened. I woke up one day on Twitter, I saw someone sold their app, GPT-3 based app on MicroAcquire using Bubble and GPT-3. It was an extremely simple app. One left text box, one right text box, one click button. Exactly like I was thinking. And they sold it for like $5,000 or something like that and they put it out on there.

09:25
Ramsri

And then I was like okay, someone is able to just sell this, build this and sell this for $5,000 and I do have some AI component and independent research that I did. There should be no way that if there are like let's say 10 million drivers driving on the road, it's not that you cannot learn it, right? Maybe you need to go through the upward learning curve. So again, I revisit Bubble. This time I took a Udemy course and I was in this mindset that just like driving ten days, 15 days, you'll get enough confidence to at least do the MVP version. So that's where I got in again into Bubble for the second time. And that tweet kind of changed. If someone sitting somewhere in the world is able to build this simple app and sell it for $5,000, there should be no way I'm not building a user interface for this independent research that I did. And then slowly I learnt what is textbooks and API and slowly I just launched with that just an MCQ text in, text out. And it took about almost one month for the first paying customer. And it was like a pay per plan.

10:40
Ramsri

Someone just pays $10 on Stripe and they get 100 more runs or 200 more runs. It's just like a counter increase. Pretty simple. And when that happened, that was my magical moment. Someone unknown to me paid money and this is actually working like someone paid money. It's out in the world. So I think it was kind of hard to believe that I as a solo developer built this and someone paid just seeing the output. And from there, slowly I started learning a little bit more. What are subscriptions, what are email marketing things that you can do in Bubble, integrating send grid and other things, logins, Google login, sign up and other things. So yeah, it started there.

11:35
Matt

Wow, thanks for going into such detail on that. Yeah, really fascinating journey and also just wonderful to see the joy of you remembering that first payment coming in. So, having moved from the kind of traditional, maybe more coded API onto Bubble, what would you say are now the biggest things you love about building a business?

12:01
Ramsri

With Bubble, I think the biggest things are speed because I know fullstack developer friends of mine where of course, even the best developers take probably good three or four days to implement a feature to its fullest form. But with Bubble, I think experience Bubble developers can just hit it off in like three or 4 hours, the same thing, and the ease of putting everything together. For example, if you have a traditional app, you have frontend, backend databases. There are quite a lot of moving things that you need to take care of behind the scenes in the database side, schema and other things. Whereas in Bubble it's there. But a lot of abstractions help quite a lot because you can run as a particular user by going in there and you can just store data, see the data right there in the Bubble app, just by scrolling, load more items, etc. And other than querying the database and doing other things. So that was very helpful. And secondly, beyond everything, if you are a solo developer or you don't want to outsource and spend quite a lot of money, you need to find a core and work on abstractions for everything else.

13:22
Ramsri

That's how you can outrun and compete with anyone out there in the world, which is my core is building an AI API, and my core ends there and everything else I need to work on abstractions. For the app I have Bubble, and for just the landing page I use another no code tool called a Typedream. And I learned SEO on the side. And with Typedream I implemented quite a lot of SEO things. So that I know right now, just from Google search, there are almost 3000 people finding Questgen in the Google search itself because of that SEO part taken care of. So I learned quite a lot of things and everything is no code. The full SEO side of things is built on Typedream with no code. The app is built on Bubble with no code. Only the API are taken care of by me on AWS. Yeah.

14:26
Matt

Why did you decide not to use Bubble for your landing page? Just out of interest.

14:33
Ramsri

So I realized that SEO needs quite a lot of things to be done, right. Secondly, for every minor change of the landing page, text, etc. For you don't want to disturb the Bubble app, which is update the Bubble app. Because if there is a new version of Bubble or plug in, et cetera that came up, it has the potential to break things right? Since it's a full stack app, there are changes going on all the time with Bubble. So working on perfect isolations can give you that comfort to operate at scale.

15:12
Ramsri

So now, let's say I have just the landing page isolated. Now I have the landing page isolated. I found someone on Upwork to write few more extra pages for me targeting, let's say, multiple choice questions true or false. And I am very confident that he or she will not destroy my main app. That is where paying users are and no database is stolen or anything like that. It's a perfect isolation for me. They go and do all the SEO things, change the copy and other things, and nothing happens to my backend data Bubble app. And also Bubble is not I mean, there are a lot of SEO capabilities in Bubble, to be fair.

15:59
Ramsri

But Bubble is not meant to be that Lamborghini for SEO, right? There might be some other companies who are hitting the nail right on the SEO side of things, whether it is FAQ markup or Schema markup, and there are many nitty gritty details to get right in the SEO to get highest page score and other things. So if you isolate that, then you can control the landing page speed and other things perfectly without disturbing anything about the Bubble and back in a database. And this can be in perfect isolation. Secondly, speaking, it's a funnel. There might be 10,000 people visiting your landing page as of this hour, but maybe 100 or 200 people will sign up on the Bubble app. So why increase the load of your landing page and keep it on the Bubble app? If you isolate that, then with the $29 base plan itself, you can comfortably scale Bubble without load or stressing it and you can keep the landing page isolated because of the funnel.

17:09
Matt

Yeah, I like to say that you can build 90% to 95% of MVP ideas with Bubble, but Bubble is not right for some application, some use cases. What would you describe as being the limitations? We touched a bit about SEO.

17:31
Matt

What other limitations have you experienced building with Bubble?

17:36
Ramsri

Right, so a good point. I think in the journey there were a few frustrating points where I encountered, for example, let's say if you're implementing drag and drop and you just drag something and drop somewhere else and reorder and other things. With a regular full stack app it's very fluid. But whereas with Bubble I had to find the right plug in to be able to do drag and drop within an ordered list, because for multiple choice questions, each MCQ will have further options. So there are nested things that you need to do drag and drop on. And that's where I found the biggest problem, where because of Bubble limitations, I could not implement the fanciest that I would have otherwise done with the regular app. And to be honest, they were fairly frustrating points where even midway I was thinking whether I need to switch to a regular app in order to implement that well, et cetera. But still, somehow every time I push something and I make it better and somehow I cruise through. So the plug in ecosystems as it is still not too mature. There are definitely limitations because the world has seen internet for the last 20 years and there are a lot of use cases that are taken care of in a regular full stack app, let's say React modules or other things that came up and someone puts out a library that you can directly integrate, but rather Bubble app cases, the ecosystem and everything although it is mature, it's not yet up to the mark to deliver the full completeness. So that's where I did encounter a few hiccups where I think I had to somewhat cut corners. But I think as a solo developer still, I'm happy to be running this, then implementing advanced features where it would have cost me more time and money.

19:54
Matt

Yeah, it's a balance, isn't it? A few months ago I was just toying with one of those side project ideas that just popped into your head and I thought, I'll have a good building this. And I wanted to do like a Trello board, simple kind of drag and drop to multiple repeating groups. And I tried all of the plugins I could find and none of them worked perfectly out of the box. I think on Twitter I'm really trying to push and encourage that. If Bubble was to launch like a native app, mobile app export, then that would be a huge game changer. Maybe scaling that back a bit, if they could just offer a drag and drop reordering with repeating groups, that would also be huge in its own way.

20:46
Ramsri

Absolutely. Similarly, people have been pushing the boundaries of designing canva like interface, right, put together elements, et cetera. But you know, like I mentioned, if you want to really build cutting edge things, then you hit some limitations. Oftentimes you need to push through by custom JavaScript, et cetera. So it becomes a little complex.

21:16
Matt

What are some of the challenges, whether Bubble or marketing or like user acquisition that you're currently working through to grow your business?

21:29
Ramsri

So I think definitely marketing is a big challenge, especially if you're a solo developer building the tech and also putting it out. So there are various marketing channels including SEO, email marketing, and each one is a behemoth within it itself, right. Digital marketing, SEO, email marketing and all these things. So I'm trying to see what works best. And also as a developer founder, you will have inhibitions about doing full scale marketing because it's not in your comfort zone. And secondly, you might not have the perfect skill sets to hit the nail on all these aspects. Right. Digital marketing, SEO and other things. So that's where I'm working on. So for the last one and a half year, what I focused on was whether there is some kind of product market fit in the sense that for every 100 sign ups, is anyone at least converting to paying user, what are the conversion rates and is the demand for this, etc. And fairly right now I have about 6000 registered users and from there I'm making like $730 MRR. And beyond that, I made one off money previously, even before subscriptions were introduced. So I'm somewhat confident that the PMF (product market fit) is there, people have this need, etc.

23:08
Ramsri

But the bigger problem is that my target audience are not creators or general folks, they're actually teachers and tech companies. And I need to, in an ideal scenario, I need to have a pathway to market to them which might not be online in most cases, some offline channels and other things, et cetera. So I think I'm trying to figure that out, where I want to get more exposure to that.

23:44
Matt

Sorry, I dropped you out for a moment. Would you go back and talk about pathways and pick up from there?

23:50
Ramsri

Yeah, so I'm exploring pathways where I can direct, where I can have a direct channel with teachers, or I take companies where I can market to them directly. One way is I actually bring in a tech cofounder and do that. But I wanted to take that route. Let's say if I take the VC funding route and want to play it big, then that's ideal for me at the moment with a little kid and a few other side projects. My commitment level might not be 100% dedicated to this as well. So that's why I'm trying to do it on my own. And SEO is something that's in my control and I can handle it. So I've been focusing more on that slowly. The SEO side has been growing. But if I have to admit, honestly, I think finding direct marketing channels to teachers or a tech audience might be much better for Questgen's growth.

24:53
Matt

So you said you started subscriptions back in March. Looking ahead now, maybe the next six months, twelve months and beyond, you got ambitions for your app. What can you share on those?

25:09
Ramsri

Yeah, definitely. So I think one immediate ambition is that right now the conversions are decent in the sense that probably I'm increasing my MRR by let's say seventy five dollars to one hundred dollars a month. So one pathway, my goal is to get to comfortable operating profitability just with Questgen and for example, let's have kids and family. So probably at least hitting $2,000 MRR will kind of be in the break even phase and I don't need to depend on other things like external consulting and other things. So that would be one perfect timeline and goal for me to achieve. And at this current stage, probably in seven to eight months I'll get there. But if there is any way to accelerate that without burning too much of cost and time, I want to do small experiments but short answer is $2,000 and above. If I hit that MRR goal, then Questgen is self sustaining in the sense that it can sustain myself and my expenses beyond any other side, apps that I do and consulting and courses that I do.

26:36
Matt

The final question what's the one thing you wish you knew about building with Bubble before you got started?

26:46
Ramsri

I wish I implemented subscriptions earlier. To be honest. There were no good public tutorials on Stripe and Bubble. If that was there, I think I would have added subscriptions a year ago and probably I might have been at higher MRR. That's one possibility. Basically knowing Bubble deeper and knowing how to build things faster. So far I've been slowly learning and adding things but if I was a pro Bubble developer I would have accelerated the journey a little bit.

27:23
Matt

Brilliant. Well, thank you very much for your shared and really hope that you get to increase that profitability in the next few months. It sounds like you're on a very positive trajectory.

27:38
Ramsri

Thanks a lot. Matt, thanks a lot for taking the time for this interview.

Start building a web app today

Join our community of no-code app builders. Learn Bubble from our Bubble tutorial library and schedule 1 to 1 calls with our Bubble experts.
Get started free
menu