When running a startup, it can often feel like you’re on the right track one week and the completely wrong one the next. It can be difficult to share this experience or to explain the ebbs and flows of your startup journey. In the no code community, being open about your product’s traction, monthly recurring revenue (MRR), and user sign ups is the norm. It’s actually one of the things I find most exciting about the no code community. You get to learn what’s working and what’s not across the industry and see patterns within the growth of successful companies. In the world of tech startups, sharing this type of information is a bit frowned upon. Throwing caution to the wind, we decided that we wanted to share with you where we are at in the process of building Metaranx and where we’re going for the next big iteration as we launch Metaranx 2.0!
When Metaranx launched the first closed beta invitations
As many of you know, the actual platform for Metaranx has changed a lot since the early days. What started as more of a rapid development tool for developers and data scientists slowly evolved into a platform where anyone could have the benefit of artificial intelligence (AI) without having to understand what was happening under the hood.
When we sent out initial invites to a segment of our closed beta sign ups, things seemed to be going well on the platform. We have a 29% conversion rate from invitees to platform onboards. We had onboarded, at the time, a significant portion of our closed beta sign ups and this list continues to grow at a 4% rate per week - all organically. Daily active users (DAU) of the Metaranx platform were hovering around 24% in the initial weeks. Not bad! We hadn’t released a paid model at the time because we wanted to see how and if people were going to use Metaranx but the numbers, at that point, didn’t generate any worry. People were entering text from blogs to social media posts to famous speeches to get sentiment analysis from the AI.
The big block: How and where to get data for AI?
During the launch process, we received a lot of important feedback from our onboarded users on how to iterate and what was missing or frustrating for them. Finding and collecting data was still a problem and we really didn’t have an appropriate solution for it at the time. No one has a plethora of data lying around, perfectly formatted on a spreadsheet, ready for use with AI. Some users were pasting in their text-based data on the platform, but that wasn’t engaging enough and still meant that the user had to think about what they wanted to analyze before jumping into the fun.
This led us to build a Tweet Analyzer. The no code community is typically a daily active user of Twitter, so we thought that’d be a great jumping off point. There is a lot you can do with Twitter data and artificial intelligence analysis and predictions. So we built our Tweet Analyzer on the platform for users to run sentiment analysis on any user or hashtag. We wanted to test if users were interested in combining their Twitter experience with AI.
<iframe width="560" height="315" src="https://www.youtube.com/embed/lj6pHtQLCVU" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe>
The dreaded churn
After a few weeks, our platform use had dropped to 5% DAU. Yikes. Our Tweet Analyzer was by far the most popular component. With so many sign ups inactive on the platform, we decided to halt invitations. Clearly, we were on the wrong path and needed to rethink our strategy. The Tweet Analyzer seemed to be somewhat working or striking a cord, but we had to rethink how we were going about this.
We were left with a few questions for our next iteration:
- Is it the artificial intelligence tool itself? Do people not care about sentiment analysis or the information it provides enough to use it regularly?
- Is the hassle of coming to a platform to use artificial intelligence versus being available on other platforms a block?
- Do people not have enough need for AI to implement it regularly and should we be a business-facing company, like many others in our space?
Do people not care about sentiment analysis or the information it provides? We couldn’t ask users their thoughts on this question, as they didn’t have a comparison on our platform. Sure, there may be more interesting language-related models (known in the AI world as Natural Language Processing tasks), but without being able to use it in a no code environment, there wasn’t much space for comparison.
Is it the hassle of coming to a platform to use artificial intelligence? To us, this seemed the most obvious. The data a user would care to analyze is out on the web - it’s part of their journey as they explore the Internet. Whether they’re on Twitter, Reddit, HackerNews, the content is out there, not in our platform. Having a user be responsible for bringing data onto our platform created extra blocks and challenges. We needed to meet people where they are online - wherever their data lives is where we need to be.
To solve this, we built a Chrome Extension. A quick Google search told us that the process of uploading your first extension can take time, as can uploading later iterations. This has proven a challenge as we would like to launch the extension but are waiting on it to go live on the Chrome Store. Unpredictability is not suitable for the fast-paced environment of a startup or for our iterative process but sometimes, you have to go with the flow! The Metaranx extension is coming soon, but it is taking longer than we originally expected and planned for.
<iframe width="560" height="315" src="https://www.youtube.com/embed/NLbLsPYXfl4" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe>
Should we make Metaranx business-facing? This question has loomed over me since day one. All the data in the artificial intelligence field points towards business needs and use cases. Venture Capitalists (VCs) are quick to tell us we should be business-facing. There is plenty of competition in this space and some really successful players and new investments in no code AI, as well. But I don’t agree that we should be a business-to-business (B2B) company. If anything, the more involved I am in the fields of artificial intelligence and no code, the more certain I am that AI needs a bigger solution than just another business tool. Seeing how excited people are to learn that they can be involved in AI, and they can impact the field and lower bias in machine learning models and datasets, makes me so happy to have pursued this route. Zeroqode’s incredible Zero Code Conference taught me that people see the importance of being able to understand and use AI and they’re excited about the possibilities of combining it with other no code solutions, too.
The answer to whether or not we should be business-facing is both yes and no. We are not switching our go-to-market strategy and we will still be working with you in this journey. I believe we are building the best platform possible and will create a truly no code solution to AI only because we are business-to-customer (B2C). It is very difficult to get feedback from enterprise clients as they have both the team and earnings to pour money into solutions, such as hiring or outsourcing, to achieve their goals in AI.
But, most of the wonderful no code community runs their own business. Whether you’re freelancing your digital skills, launching a startup on no code tools, or running a no code platform yourself, many of you are business owners. You run e-commerce stores or newsletters or communities that you want to optimize, each with your own goals in mind. This is exactly what artificial intelligence is great at - patterns within varied data. So we did some research to see where it would be easiest to meet you and your data online and as a result we want to launch Metaranx AI tools on the following marketplaces (and this is just the beginning!):
Using Metaranx artificial intelligence tools with your data
We are in the process of building out the AI tools for Zapier and Hootsuite to begin with and test how people, and your businesses, use them. Zapier allows you to integrate with so many other applications, and it’s very popular in the no code community, so that makes building a Metaranx AI tool for Zapier a no-brainer. Now, you can take your data that integrates with Zapier already, such as through Google Sheets, Slack, Facebook, Airtable, Shopify and so much more - and run artificial intelligence tasks on it to analyze and make predictions. The best thing is that you can later use this data to build your own AI tools, too!
Next, we’re focusing on Hootsuite. A lot of marketers use Hootsuite for their social media channels to monitor and improve their communications. Now, you can apply artificial intelligence tasks directly to your social streams (kind of like what the extension would accomplish!).
Building Metaranx artificial intelligence tools with your data
Next, we want you to be able to build AI capabilities into your applications and websites. To enable this, we need to build components for no code marketplaces, such as Adalo. When you build an app and want AI in it, you’ll be able to add our component from their marketplaces. This is a really exciting step to encourage everyone to build or include AI in what they’re already creating in the no code space!
What Metaranx will look like going forward: Metaranx 2.0
Going forward, when you land on the Metaranx console, you will be greeted by our suite of applications available on different platforms, whether it’s the extension or Hoosuite’s marketplace application. When you use an AI on our platform, you’ll either directly click a tool you want to use - such as our more popular Tweet Analyzer or navigate to our tool to use on another platform. You won’t choose between the different types of AI tasks you wish to accomplish, such as Language versus Image, because we’ve realized that’s just extra steps. If you want to build an AI tool on Metaranx, we have are simplifying the process even further in terms of UI/UX and will put more focus into this section once we’ve found a better product market fit with the using AI component. You will still be able to see all your data with the goal that you can download into a .CSV and build an AI tool of your own with it.
Our mission remains the same: we strive to democratize artificial intelligence and empower everyone to use, understand, and build tools that will help them achieve their goals with their data - whether personally or professionally. To include more people in the process of AI, we need to build a platform that allows everyone to feel confident in what they’re doing and harness the power of AI for their projects, not just the big corporations or well-off countries. Everyone deserves to be informed and in control of their data, the machine learning models that process it, and the predictions AI makes on behaviors relevant to the end user.
Where’s my invite? I signed up a million years ago!
We had planned to have the extension launched by October 16th but as you have read above, that did not go according to plan and we aren’t certain of how long the Chrome Store approval process takes. But we’re moving forward! This build Sprint to add some tools to marketplaces, revamp the Metaranx console, and run testing will take approximately three weeks. Based on our roadmap, we should be able to invite all closed beta sign ups into the platform by November 20th, in staggered groups. I send invites in batches so I can send email requests for feedback and we can implement some iterations throughout the process. We plan to be in open beta in December, depending how onboarding, churn, and feedback looks like with our closed beta sign ups.
If you received an invite from us already you will be receiving a new one to Metaranx 2.0! The platform will look a bit different, we will have our subscription model launched, and you will be able to access all new tools and capabilities. We can’t wait!
I hope this has helped you better understand everything happening behind the scenes at Metaranx. We are working hard to bring artificial intelligence to everyone and that continues to be our mission. We can only democratize AI with your participation and thankfully, the no code community has been incredible at taking part - even when we haven’t built something that solves your needs just yet. We look forward to launching Metaranx 2.0 to all our closed beta sign ups and seeing your reaction to our different tools, extension, and changes!