For a large part of 2019, my company New Internet Labs has served as app reviewer in App Mining. Answering @muneeb’s call for feedback on the program, I wanted to share some of my thoughts. I think Muneeb did a great job highlighting high-level challenges on the recent App Mining call - I’ll comment on those later, but first let me share some other observations:
Observations & feedback
I’ve been blown away by the enthusiasm of many of our app miners. I get to try most of the apps every month and it’s been really exciting to see the progress of many of the apps, especially those that are being actively developed as side projects while still holding down a day job.
Building a business
Ultimately for our app miners and our ecosystem to be successful, they need to be able to turn their apps into profitable businesses. Ideally these should be businesses where users receiving value are paying apps developers for that value so that app devs aren’t left to rely on selling users as the product. App mining as it stands today doesn’t do much to help developers move in that direction either from a technology perspective or incentive structure.
What is a Blockstack app?
We didn’t clearly communicate at the beginning of App Mining what a Blockstack app is. Is it any software that uses any component of Blockstack technology? Is it a web app? Is it a hardware device? Is it a mobile app? Is it a consumer app? An enterprise app?
New Internet Labs has our own views as to what apps should look like: they should be web-based and ultimately distributed via Blockstack Names. That’s just our take though - it might be worth it to the ecosystem provide a mechanism for other types of apps: mobile, hardware based, etc - to compete but in a way that doesn’t also force them to conform to our view of an app.
Metrics that incentivize one type of app do so at the expense of others.
Incentive for quick one-off apps
The incentives of App Mining are such that the number of apps entered into app mining was more of a driver of success than quality of apps entered. We’d rather see people trying to get traction with one or two apps (already a HUGE task) than pumping out lots of apps no one is using.
On Muneeb’s thoughts
Muneeb shared with us what he sees as the 3 biggest challenges of App Mining.
- Objectively fair distribution - differentiating between quality apps that move the Blockstack ecosystem forward vs apps that are simply gaming the metric.
- Privacy preserving analytics - the ability to track active users and user growth.
- A healthy decentralized ecosystem that determines how App Mining works.
In general, I agree with these challenges. Some comments:
Objectively fair distribution
As an app reviewer, we felt a lot of resources were wasted reviewing apps that were not good-faith attempts at building successful apps to grow the ecosystem or becoming true decentralized apps. There are several SaaS apps, for example, that are entirely centralized and went so far as to submit slightly modified staging versions of their apps that they never intended users to use just so they could tick the boxes on app mining and win the reward.
There are two parts to this. The first are privacy preserving analytics for the use of the app developer - to help them make decisions about the design and marketing of their apps. This is something that’s clearly necessary and I believe is a solvable problem but much different from analytics as they relate to app mining. If the app developer is the only audience for analytics, they have little incentive to game them.
Analytics when used as a way to distribute app mining rewards is a much harder problem. As soon as you introduce an incentive to game analytics, you start having to deal with problems like sybil attack, reputation and separating “real” users from “fake” users. I’m not sure it’s even possible to determine a “real” user from a “fake” user since the difference between the two is a matter of intent and the structure of the person’s incentives - information that only exists in the user’s mind.
For either type of analytics, tracking needs to be consent-based and opt-in by default.
A healthy decentralized ecosystem behind App Mining
The way forward
I think we’re at a unique point in the history of Blockstack where we’re right before the launch of Stacks v2.0 and the availability of a number of new technologies in the stack. We should take this opportunity to revisit some of our assumptions and consider new approaches that better reward apps that build our ecosystem before rolling out App Mining 2.0.