I cannot believe that we’re already halfway through 2023. I made a post about my plans for 2023 back in January.
I thought it would be a good time to give an update on the progress of one of those projects, my AI WordPress plugin.
So keep reading if you want to see how this project is going.
Table of Contents
My AI WordPress Plugin
At the end of last year, ChatGPT was blowing up. Everyone on the internet seemed to be talking about it and generative AI. I had some experience working with AI at IBM on the Watson project and at Moveworks on an employee service chatbot solution.
I saw many tools coming out that were capitalizing on the OpenAI APIs. I thought I could take my experience and spin up one of these tools.
These tools were making tens of thousands to millions of dollars. Check out this interview with a creator of an AI content generation tool.
So I created a WordPress plugin called WritersBlok. It uses OpenAI’s GPT-3 and GPT-4 models to help authors create content directly inside WordPress.
It was interesting for me to learn how to develop a WordPress plugin, and I took a more modern approach by using react.js instead of PHP, which traditional plugins use. This made it a bit more difficult to Google for answers.
Current Status of WritersBlok.ai
The plugin is functional with several free features and a few paid features, including full blog posts and listicle writing.
It requires that you sign up for an OpenAI API token, and OpenAI will bill you for usage, but it’s pretty cheap.
I thought I could get people to sign up for the free version, then get them to upgrade to some paid features once they use it.
Here are the current stats:
- Earnings to date: $68.97
- MRR (Monthly Recurring Revenue): $5.58.
- Current Subscribers: 1
- Free Installs: 359
- Deactivations: 133
- Advertising Spend: $1,444.15
- Profit: -$1,375.18
As you can see, this project has been an objective failure so far. I spent quite a bit advertising on Facebook, which led to many free sign-ups, but I could not convert those people to paid users.
When users install the free plugin, they get added to a MailChimp email list where I give them tips on using it and then pitch them the paid plugin.
Identifying Problems and Pain Points
Since so many other AI tools are very successful, I’m attempting to analyze where I am going wrong. Here are a few ideas that I have.
#1 Sign Up Friction
My plugin is free to install, but it is a “bring your own key” model. This means that people must sign up for an API key at OpenAI.
When users uninstall the plugin, they get 1 final email asking why they uninstalled it. I don’t get many replies, but I think I had at least 2 people tell me that they could not figure out how to get the API key.
This is surprising because I made a page and video about how to get an OpenAI API key.
The issue is that API keys are a concept that developers are familiar with (like me), but regular users are confused by this process.
This problem is very high in my funnel, filtering out many potential customers.
I think requiring an API key is causing most people to uninstall the plugin before even trying it.
#2 No Easy Trial of Premium Features
I do have a 7-day trial of the premium features. It requires the user to enter a credit card. They can cancel anytime before the 7 days are up. Only 1 person has tried it, and they decided to cancel.
The competitors I see are offering some free usage of their full products. Many don’t even require a user to sign up for the full experience.
Not having a quick and easy demo of the full functionality limits my users at the top of the funnel.
#3 Value of Premium
Comparing my tool to others, I think my offering is on par with theirs.
For the few users who get past the signup and trial friction, my email campaigns are not converting them to paid users.
I think that I am not selling the value of the tool. The users don’t see how the paid tool can benefit them and how my tool is better than the alternatives.
I think fixing the top-of-funnel issues is more important now. If users are not even entering the funnel, I cannot convert them.
#4 Being WordPress Native
When I developed the tool, I initially wanted it to be a WordPress plugin.
I made this decision because I think it’s a much better user experience to use the tool inside WordPress from the editor’s sidebar.
I have mixed feelings about this decision.
Most of the competitors that I see decided to go with a Website instead of a plugin. This doesn’t limit them to WordPress only.
The benefit of having a WordPress plugin is that you can offer a free version that can get user installs from the WordPress plugin directory.
A better approach would have been to start with a website and then incorporate a WordPress plugin through a custom API.
I could be marketing the product much better. I did some ads on Facebook, and they resulted in free users.
I also reached out to a consultant to help me market the tool. One gave me a free consultation and was very candid about the costs of marketing a tool like this.
I believe it was over $60k in the first year. I can see how that might be required, but I didn’t want to spend anywhere near that kind of money on the plugin, especially since I have not optimized it and proven it can be profitable.
How to Resolve These Issues (or If I Even Will)
I have some ideas for addressing the problems, but I’m unsure if I want to spend additional time on this project. This is the first paid software product that I tried to launch.
While I’m still deciding if I want to try to course correct, here is what I think I would do if I decided to move forward.
Going Web First
Many competitors are web-only, and they do fine. This would allow me to do development quicker since I am more familiar with websites than WordPress plugins.
Remove the API Key Requirement
This is the biggest friction point, and I would need to remove it. Instead of having the user supply an API key, I would need to use my API key. This would also mean that I take on financial responsibility for every call made to OpenAI.
Create a Free Demo of All Features
I need to allow users to experience the full tool without paying. The issue is that every free interaction I allow will cost me money on API calls. I would need to include this free usage in my marketing budget. Every user that does not convert will cost me money.
Change to a Usage-Based Pricing Model
If I don’t have the user supply an API key, I need to charge the user more than I pay for these API calls. This seems like a pretty good model because I can set a margin on top of the fixed price OpenAI charges and not lose money.
The issue here is that I need to figure out how to integrate with a payment provider that also keeps track of how many credits the user has left.
Create an API for the WordPress Plugin
I am undecided on this, but once I have a website, I can integrate the WordPress plugin with my API. So I can still offer an experience inside the WordPress editor and the website.
Big Marketing Push
Once I confirm the changes are working, I need to put a big push into marketing the plugin. I noticed that affiliates heavily promote many of the other plugins, but I’m not familiar with how to attract affiliates to promote my tool.
I would look into that, revamping the email campaign, doing more content marketing, and running additional Facebook ads.
I have posted about my past success and failures, and now, this project is a clear failure.
I believe I could turn this tool around if I invest in all of the above. My ultimate goal would be to make at least $5,000 per month. I know these other tools are making money, but the space is super crowded because the barrier to entry is low.
I have to decide if it’s worth giving it one more shot. Other people are making money doing this, but I don’t know if I’m confident I can do it.
I also see that the hype of ChatGPT is dying down a bit, and there are so many tools in this space. I’m unsure if an “also ran” AI tool is worth it.
However, the fact that there are so many tools seems to validate that there is a market here.
I know that I am being indecisive on this project.
I am investing in other projects, mainly the niche sites I am working on. They are not exactly raking in the dough, either.
I’m struggling with all my side projects right now, and seeing so many success stories online has me questioning my abilities.
If you’ve made it this far, I’m sure you have some opinion about where I went wrong. I’d love to hear your feedback and ideas in the comments below.