Hello, with plug-in lab has anyone made money through freemium models. In addition, is there a way to see how many people use my plug-in? In terms of installs
We are a happy PluginLab customer for our plugin “The Bolt”, but not for monetization. We do that ourselves via Recurly, b/c we already had that baked into our backend low-code platform that we use to build our plugin’s functionality. Our model is freemium and yes we have paying customers.
Yes you can see all the people (they call “members”) who have installed your plugin. You can download that information as well, e.g. CSV export.
Best of all, you can view your plugin’s usage activity as members use it in real-time.
We’ve been very happy with PluginLab! They are always very helpful.
Thank you for the valuable info’ do users find your plug-in through external marketing or simply organically by searching through the plug-in store. I am working on plug-in that doesn’t have an existing brand or site, wondering if people install the apps without some sort of external marketing campagin
Thus far we get a trickle of new customers every day just from being in the store and being found among the 800+ other plugins.
Unfortunately the store only gives 120 characters to describe plugins and use for search matches, which is woefully not enough for our plugin that currently has 15 different tools. Nor does it provide direct links to plugins, even if we wanted to somehow drive traffic to it via external marketing.
We currently consider our plugin to be an experiment and are prepared to wait and see what happens as the store matures. Hopefully one day it will blossom into a similar experience as the apple app store for plugin providers and customers. As it stands now, it’s something of a dumpster fire.
Thank you for the insight. I agree it’s a dumpster fire, but it may make a good investment one day. My plug-in research Butler is also experimental as well. How many users per day are you getting from the organic traffic if I may ask?
We are lucky to get 10 new signups a day. But our generic 120-character store description makes it hard to find b/c we can’t specify exactly the types of functionality our plugin provides, like giving ChatGPT persistent memory, math & logic computation (sort of like code interpreter by executing code it writes to solve problems), web search/scrape, PDF read/write, etc.
These are interesting stats, dunno how accurate:
Sure we could break up our plugin into multiple individual ones, but having a single one works around the 3-plugin limit.
For now we’ll wait and see if the store catalog functionality improves.
Here is some other reported info about what’s coming soon:
So perhaps plugins will become available to the free users of ChatGPT 3.5, which would greatly expand the audience for plugins.
Having 10 signups is impressive to me. Do you know what percentage of your users sign up and retention rates? I’m thinking that if I get 100 users a day and 1-3 sign up per day, achieving a 3% signup rate would be a success for this app and a greater than 10% retention (my app has simple research functionality, it’s called research butler in the app store). I agree that the description isn’t extensive enough to explain the sharing functionality, but perhaps it’s meant to display only the core functionality. I’m considering creating a helper operation that provides users with more specific advice that my application has to offer. Your app sounds very enticing, particularly with the persistent memory. It seems like an all-in-one tool for data analysis and coding. I wasn’t aware of the three-plugin limit. Can we remove old submissions if we choose to create new plugins beyond that limit?"
The 3-plugin limit is the max number of plugins you can have active at one time in a single conversation in ChatGPT, not a developer limit on how many you can have in the store.
Our retention rate is low b/c many people install the plugin and never use it. We blame this on the too short 120-char description, so people don’t have much of a clue as to what functionality the plugin actually performs.
The onboarding of new customers when they install a plugin stinks, b/c ChatGPT doesn’t give the user any clue what to do after installing plugins. We ended up sending them an email with instructions when they do the OAuth flow. We also have a “help” operation like you suggest is needed.
My retention rate is extremely low as well it’s a research plug-in (allows user to search a site for articles, I give users a 10 search quota until paywall. only 5% of my active reach that. Therefore, no subscribers yet as of day 5. I however have released a bar chart plug-in (anybarchart) which is much more interactive. Hopefully that pans out.
My question for you(if you don’t mind) is do you only rely on traffic through search or do you use ads. Also per how many active users do you get a subscriber. I understand it’s highly variable by application but essentially I am trying to tune the plug-in lab settings until I get my first purchase. Best,
We currently only derive traffic from the store itself. No ads.
We are mostly hoping that OpenAI will improve the store one fine day and that it eventually works more like the apple app store. It’s really disappointing b/c the current store is as if a high-school intern designed and coded it. Heck, why don’t they just eat their own cooking if it’s so great and magical and let their AI design and build a better store in just a minute? I think that sort of speaks to the real limitations of AI technology.
As a whole we don’t see very sticky usage patterns even though we provide 15 general-purpose tools. Maybe that’s just b/c people are playing around with lots of different plugins to find ones that suit them best. We could send our surveys to our users to find out more if we wanted, but for now it’s ok just being a mystery.
Yeah, I guess the take away is that it’s very difficult to measure what they users want and what they’re responding to due to lack of infrastructure. For now, I am just trying to get my users to make more calls because that’s the only true metric of how interested they are for the time being