Colin Henderson and Amela Djokovic came up with StarCrossed, a dating app that matches users based on an astrological algorithm, after reading about rising investor interest in astrology apps. They eventually chose to work with Crowdbotics based on Crowdbotics' competitive price point and collaborative development approach.
StarCrossed is built using the Crowdbotics Dating App Blueprint, a Tinder-style mobile app template that is fully functional out of the box. By making his own adjustments to the app's UI and leaving custom backend development to Crowdbotics' expert developers, Colin (a hobbyist React Native developer) estimates that he has saved "thousands of dollars."
StarCrossed is available on the App Store and will soon be launched for Android. StarCrossed users can currently contact expert astrologers for relationship advice directly within the app, and planned upcoming features include advanced astrological chart breakdowns and custom text suggestions. Colin and Amela are planning to apply to accelerators and scale StarCrossed in partnership with Crowdbotics later this year.
Crowdbotics: Tell me a little bit about your background.
Colin: I’m an accounting graduate at Pace University, and my girlfriend graduated in finance. So we both have a strong business background, and we live in New York City. I work at PwC in the tech and media assurance firm, so I audit a lot of tech companies and startups and media firms.
As a result, I’ve got a little bit of experience in that startup life, and I feel like it's a little unique to me, because I really get to see all the operations that happen in certain startups. And hopefully we'll learn from their mistakes with vendors and the ways they're reporting stuff and how they're budgeting all their money. I hope to use that knowledge to my benefit.
And then I also took a couple of coding classes, just to help make my additions to the app when need be.
Crowdbotics: Would you say you're a hobbyist developer? What's your level of technical know-how?
Colin: It doesn't really go past a hobby. I've only taken a couple of classes online. It's all React Native app development. I just took them specifically so I can make adjustments to the app we’re building.
I think what I really wanted to do was be able to understand everything that Crowdbotics was going to be telling me so that I could have a more productive conversation with them and communicate with them properly.
Crowdbotics: Have you found that to be helpful?
Colin: Yeah, it was helpful in that sense that whenever we're trying to figure out a solution to a problem that sometimes they didn’t necessarily understand, I could guide them in a way. Like, “Oh, hey, I noticed that this line of code might be causing something to happen, but I don't understand the solution to this. But maybe if you look at this problem, you'll be able to figure out the solution.”
And then we usually figure out what we need to do, and it does become a little bit more of a collaborative effort, which was what I was looking for with Crowdbotics.
We've saved probably thousands of dollars already by upgrading the UI ourselves and leaving the back end to Crowdbotics.
Crowdbotics: I hear that a lot, that customers find that everything is visible to them in terms of current challenges, things that are going well, and things that are going poorly.
Colin: I definitely have to advocate for my interests first. It's not like the team will show me everything, all the time, by default. I have to make sure that I ask the right questions to move things in my intended direction. But then once I do, it starts working pretty well.
Crowdbotics: How many apps have you built with us?
Colin: Just the one, StarCrossed.
Crowdbotics: How did you find us, and why did you choose us to build it?
Colin: I did startup school through YCombinator, and I saw that on their list list of YC alum discounts, there was a nice discount for Crowdbotics. I thought, “Oh, this might be a pretty good method to do it.” We were originally going to use a custom dev shop that would cost at least $30,000, and we weren't too happy about that. But then we saw the more affordable options available through Crowdbotics and we were ready to go all-in with that.
Crowdbotics: Was there anything specific to the way Crowdbotics does things that sounded appealing, or was it mostly the strong fit for your budget?
Colin: The budget fit was certainly the biggest thing for us, especially because we were still so new to all this, and we really didn't know what was the best pitch for everyone. We couldn't discern which pitches were genuine and which pitches were actually going to be beneficial to us. We really were pretty price-specific, and this route was the best fit for us.
But we also were attracted to the idea of working collaboratively with your team, rather than handing it off and waiting for results.
Crowdbotics: Great. So, tell me about StarCrossed.
Colin: StarCrossed started about a year ago, when I saw a couple of articles in one of my MBA classes talking about how venture capitalists are starting to invest a lot into astrology apps. I didn't really care about astrology, but my girlfriend was super into it. So I sent the article to her, and I knew she had a pretty good business acumen as well, so I felt like this would be a good person to work with.
I was like, “I know you're really into astrology. Let's think of some ideas that we can play around with and start a company.” And she had noticed that no dating apps have integrated astrology into them yet. And I was like, “Okay, that's cool. But why are you checking out dating apps when we’ve been together for a year?”
Crowdbotics: Yeah, you raise a good point.
Colin: But then I said, “You know what? I really like the idea. Let's get past that, and let's roll with this.”
So then we built our first MVP, which was this Excel spreadsheet that I made by having a bunch of friends in college fill out surveys. We had about 100 people create profiles where we get their birth dates, who they're interested in, and whatnot. Along with the birth dates, we collected birth time, birth location, and we were able to generate all the birth charts.
And then I played around with everyone's planets to figure out the best matchmaking algorithm based on astrological compatibility.
I started with their sun signs, which is the stereotypical zodiac sign, like, “Oh, I'm a Gemini,” or, “I'm a Libra.”
But there's also the moon rising and all the other planets. Those all mean something different, so if everyone in astrology believes that the Sun is your overall personality, and the rising is the way you present yourself, and Venus is love, and Mercury is communication, and things like that, then I felt like you could really make something with that on a compatibility scale.
I was able to write an algorithm to match people on a zero- to five-star rating system, and it worked pretty well. There was a good distribution among the hundred people of having one-star matches and zero-star matches, and five-star matches were pretty rare. There would be a couple of five-star matches per hundred people. And that sounded about right, so we started talking to as many people as we can about this thing that we've been working on.
The Crowdbotics Blueprint made the process a lot easier. We could implement a lot of our designs on top of the Dating App Blueprint and then start adding some astrology-specific features.
Crowdbotics: You've zeroed in on an intriguing thing about this from a technical perspective, which is that you're gathering a lot of unconventional data about people and using that to match them in a way that is highly qualitative in some ways, but also very advanced and sophisticated in others.
Colin: Yeah, exactly. When people are swiping through matches, the only things they’re really seeing are their name, their age, their sun sign, and then a star with a certain number on it. And that certain number is how compatible you are based on your birth chart.
But if you click on the profile, you get that full in-depth analysis. Are we compatible communication-wise? Are we compatible romantically, emotionally, sexually, personally? And you see their full birth chart, so you really get to see both aspects. You get that basic part in the very beginning and you get to see that advanced part once you do some digging.
Crowdbotics: That seems pretty intuitive to me. I know that we have a Dating App Blueprint, and I was curious about whether you used that for your build.
Colin: Yeah, we started off with that and we implemented our own colors.
I actually made a wireframe using InVision that I sent to Crowdbotics when I first started, for the kickoff. I think that made the process a lot easier. They were like, “Okay, I think we could implement a lot of this into the Dating App Blueprint, and then start adding some astrology features.”
They found a good, affordable astrology API that only charged $100 a month for all the astrology features. It gives us the horoscopes, the birth charts, and personality reports. We were really happy about that, especially using birth chart descriptions. Everyone gets to see what all their signs are, what that means, what their horoscopes are every day, and what their matching signs are.
This API seems to be the standard system that a lot of astrology companies are using, so that was huge that they found that.
Crowdbotics: So you had some InVision mockups. You had this earlier MVP. I'm curious to know about the total materials that you brought to the initial build.
Colin: We wanted to start off small with a website. Instead of a dating app, we'd be more like a dating service where people could sign up, make their profiles, and we’d email them back with their matches and how compatible they are with those matches. We’d more or less be emailing them the spreadsheet.
That was cool for a while, but we knew that the endgame was to have this be an app. We really wanted to use the website to test product-market fit. With InVision, we were able to make the mockups pretty similar to Tinder with our vision of how to integrate astrology into it.
From there, we started pitching it to all sorts of developers to say, “Hey, here's all the stuff we did, how would you go about building this for us?” When we brought it to Crowdbotics, we got the best answers from them.
Crowdbotics: That level of preparation makes it a lot easier for our team to give an accurate quote. What were the technical considerations that you knew that we would absolutely need to knock out of the park for this to be a strong product?
Colin: I think the most important things were to have a functional chat system and be able to pull up people's birth charts. We really didn't know how to do that yet, without having to do it manually in order for the algorithm to work properly in the feed.
We also wanted to know how to change the algorithm each day. If we said,”Hey, maybe the algorithm should work differently. Maybe this is what a two-star match really means,” then we wanted to know that we could make that quick change to the code, which we've been able to do.
The algorithm has definitely changed since the first day I wrote it. And I think that's one of the most important features that we were able to build.
If I want to make a fix for free, I can make the fix myself. And if I genuinely don't know how to do it, then I can ask Crowdbotics to make the adjustment.
Crowdbotics: That makes sense. I would assume that user feedback would guide those changes. Or is there an astrological component that you have to monitor to ensure that the algorithm is performing in a way that is valid?
Colin: Actually, the algorithm isn’t really based on user feedback at all. We talk directly to astrologists. We'll go to these astrology events in Brooklyn and meet up with some of them. I'll let them know that I'm working on this app, and then I'll run through the algorithm with them real quick on a conversational level.
I'll say, “Okay, let's say it's a one-star match for someone who's compatible with Mercury and Venus. And the only way you could be compatible is if your sign is conjunct, trine, or sextile.” And they'll be like, “Yep, that makes sense to me.” Or they'll say, “Maybe not. Maybe you have it be Venus and Mars, or Venus and Pluto,” or something like that.
I basically do what they say because they're the experts, not me, in that aspect. I just thought I could put the math into the astrology.
Crowdbotics: It's really cool that it's expert-driven, that you're seeking that out.
Colin: Yeah, we actually have a couple of them on call. One of our features is that if you want a more in-depth breakdown, that's where you can start paying money to talk to one of our astrologists about your real synastry breakdown with your matches.
Crowdbotics: So this is a much more personalized experience that you can pull up, and it’s actually kind of unique to this specific type of dating app.
Colin: Yeah, you get that real relationship consulting.
Crowdbotics: Are there any big wins or features that you can attribute directly to your team at Crowdbotics?
Colin: Yeah, I had no idea how to generate those birth charts automatically. And I really didn't know where to look. And I feel like Crowdbotics was able to find the astrology API really fast, and then that astrology API was able to do a lot more than we were even asking for it.
For example, we never thought we'd want to implement horoscopes in the beginning. And then we realized that implementing horoscopes was just going to be a quick addition, a couple lines of code, so we were like, “All right, let's go all-in on that then.”
Crowdbotics: So there was data in that API that you weren't really anticipating featuring inside the app? How did you re-envision the UI to incorporate this new trove of data that you had access to? Was that on your end?
Colin: It was more or less on my end because I have enough technical acumen to make these additions myself, so I edited the frontend components until I needed to work on the backend side of it. At that point, I just said, “Hey, I don't know how to bring this data out. Can you please do that part?” I would do most of the design stuff myself.
The frontend side, I've been really enjoying. The backend side is weaker territory for me, because I'm just not experienced in it. And that's usually where Crowdbotics is able to push.
Crowdbotics: So you have access to the GitHub repo, and the app code is just normal React Native. Do you find that having some React Native literacy lets you dive in and handle tasks that you don't want to bill out to Crowdbotics, and that's a cost-saving feature?
If I said, "I understand this isn’t in the Dating App Blueprint, but I feel like it would be good for this app to have," then they would do a great job of adding that feature.
Colin: Yeah, I feel like we've saved probably thousands of dollars already in UI upgrades. We've asked Crowdbotics for a couple of things and they've given us the estimated costs, and we’ve said, “Okay, let's learn how to do this ourselves.” Because many of these things are easy features to implement.
We've been pretty happy with what we've been learning to accomplish and how we've been able to do it. Even just changing the background color has been a cool win, and then kind of re-imaging a lot of the icons and the way the cards are set up in general. For example, we got rid of the white bar on the bottom of the cards and had a kind of black gradient.
We've been pretty happy with the UI aspect, but we know it's going be changing every day. I've been busy right now with accountant work at PwC because it's been a busy season, but once it’s over, I'll be back on the coding grind to make updates every day.
Crowdbotics: What was the day-to-day process like working with us?
Colin: On the first build, that's definitely when the most communication was happening. That was a daily conversation with our project manager. He would basically tell us what they accomplished today, what they plan on accomplishing tomorrow, and make sure I'm comfortable with their process and how much they're accomplishing.
If they had any bumps in the road, they would ask follow-up questions for me to hear my preference. Those calls went pretty smoothly. It was pretty consistent communication and keeping me updated.
Crowdbotics: So even when there was a snag, you felt like you were kept apprised of our progress?
Colin: Yeah, I was never surprised when something popped up because I received daily updates on where they stood with these specific tasks. So if a target date had to move, they’d let me know in advance. It was never like they said, on the day of, “We don't have anything for you.”
Crowdbotics: That’s reassuring. So what is the state of StarCrossed now? What’s the roadmap from here?
Colin: We're at a pretty good place right now. We have about 2,000 users on the App Store on iOS. We hope to get the Android build finished soon. We’re seeing strong potential for scaling our acquisition strategies.
Eventually, we're going to add some premium offers with monthly membership fees, like unlimited swipes and things like that. Normal dating app stuff. We’ll also add advanced astrology features, which will include a stronger in-depth breakdown, some text suggestions based on zodiac signs, and more. More stuff for astrology fanatics.
Crowdbotics: Did you say "text" suggestions?
Colin: Yeah, it'll be specific wording. For example, “This kind of phrase would work for a Gemini.” That kind of thing.
Crowdbotics: That's really interesting.
Colin: Another thing I’d like to add is that this process has helped us get ready for applying to accelerators, because we have been able to launch fast and build a user base by the time it's now application season. We're pretty happy with that.
Crowdbotics: That makes sense. It’ll also be useful to have a sense of realistic development timelines when entering an accelerator as an early-stage startup.
Colin: Yeah, at the end of the day, I don't really picture Crowdbotics as a dev shop. I know a lot of accelerators don't like it if you use a dev shop. I feel like this has been a collaborative effort to build out a Blueprint. Now that I have a good chunk of experience on this project, I know what kinds of problems we still have and how to solve those problems and improvements to make.
Crowdbotics: A lot of people don't immediately see any value in the fact that you own your code, it’s transparent to you, and this isn’t a proprietary platform. You can engage as much or as little as you want, and the way you’ve specifically taken a hands-on approach to this project is still kind of novel for many of our clients.
Colin: The main thing is that I don’t feel trapped in a situation where I’ll have to pay a lot of money. If I want to make a fix for free, I can make the fix myself. And if I genuinely don't know how to do it, then I can ask Crowdbotics to make the adjustment.
We pitched the app to all sorts of developers to say, ‘Here’s our design. How would you go about building this for us?’ Crowdbotics gave us the best answer to that question.
Crowdbotics: Along those lines: I'm curious to know if you as a more technical user have seen any advantage to the fact that we use React, APIs, and Django to build our templates and our apps.
Colin: When I was looking for stuff to learn, there were a lot of React and React Native principles that I was able to apply immediately to this project. So it's good that you use mainstream stacks. I don't have to do a lot of digging to learn how to make adjustments myself.
Crowdbotics: Yeah, that makes sense. That's a big reason we chose it. Is there anything else you’d like to add?
Colin: Overall, I think I really appreciated the transparency with the app. There was never a time that I couldn't see the code on GitHub, so there wasn't really anything hidden from me.
I was also really satisfied with the fact that the team was able to find the astrology API and implement that perfectly.
In fact, I feel like they were always able to accomplish anything that I asked to include in the builds. For example, if I said, “I understand this isn’t in the dating app template, but I feel like it would be good for this app to have,” then they would do a great job of adding that feature. It might be an extra fee, but they’re not limited when it comes to what we can have them do.
Crowdbotics has never said, “No, we can't. You can't accomplish that.”
To start building with Crowdbotics, contact us today.
May 6, 2020