Solving the Notifications Dilema
Paul Croubalian
Indie apps, Android, iOS, MacOS, Windows, and Web. I like finding the pain points and taking them away. Full-Stack Dev
This post is not the post I intended to write.
My original thought was to write about three ideas for apps. We would discuss them, weigh the pros and cons of each, and establish the features they should have. Finally, we could decide which I would build.
It was a nice idea. There are a heckuva lot of smart, insightful people on this platform. We could have an ongoing discussion and regular updates on the state of our "group-built" app.
It sounded like a good idea.
Our group-think app would likely have been a useful and enjoyable thing.
Yes, I blush as I write this.
Four days after the initial post, there are 83 impressions. These are not 83 interactions, they are only 83 impressions.
That means the post only appeared on a feed... 83 times.
There was only one interaction.
That interaction was with my sister.
Oy, ve.
Some may think, “But, Paul, you gotta build up your presence first. Build up your follower count. Get some connections! Impressions will rise once you build an audience. Your engagement will follow!”
Uh-huh, sure it will.
Oh, did I mention I have 3,257 followers and 500+ connections? Respectable, albeit not at Susan Rooks' level.
Yet, my update appeared on 83 feeds out of a possible 3,257. Either 3,174 connections haven’t logged in since I posted all those days ago, or this is a broken system.
Come to think of it, either way, this is a broken system.?
I thought we could have an ongoing discussion and regular updates on the state of our group-built app.
Now, I'll go another route. Let's fix the broken Notifications system.
I doubt it comes as a surprise to anyone. This is nothing new. I wrote "How To Solve LinkedIn Notifications And Groups.", way back in January 2018. If anything, the problem is worse today.
Good thing for us all, technology has advanced to make the fix easier. At least the potential is there for an easier fix.
React Native lets me write a single codebase for Android, iOS, Windows, and Web. SQL is still tempting for structured data, but MongoDB and Express can also kick serious butt. Even AI-driven voice-processing tools are now pretty simple to build. They were, for all intents and purposes, impossible in 2018.
I guess this "problem" has been bugging us all long enough. It's time to slay this particular dragon.
(Yes, in this case, "dragon" is spelled, "b-i-t-c-h"!)
First thoughts are rarely the best thoughts
My first thought was both simple and simplistic. The problem is that we must have a say in what we see. Right now we don’t. Let’s force the issue.
Let’s build a solution that allows us to decide what we see and when.
We can create a master Web App linked to IOS and Android apps. While we're at it, why not a companion Chrome Extension? The Web app will track accounts. It will find new content on its own. When it does, it'll let us know through the apps and/or extensions!?
Notifications would link back to the original content. That original content could be anything. It can be a LinkedIn article, a YouTube video, a website blog post, -- whatever.
领英推荐
We would never again be at the mercy of some algorithm that decides what we should see and when. We will decide!
Easy peasy lemon squeezy! Or is it?
I mentioned it above. First thoughts are rarely the best thoughts. Heck, they are hardly ever even good thoughts!
Imagine this. What if all my 3,258 people posted on the same day??
Imagine 3,258 notifications! Even if evenly distributed throughout a workday, that would be a notification every 9 seconds or so! But, we all know they would concentrate around midday rather than spread out. So, say a notification every couple of seconds. Max!
Even that neat distribution pattern is impossible. The reality is closer to them shooting off at random.
The freaking phone would sound like a machine gun. I would be the first to uninstall my app! I may even toss my phone against the nearest brick wall, thus silencing it forevermore.?
Neither outcome is desirable.
A multi-tiered approach??
Let’s be honest. I don’t need a notification for every post by everyone, every time.
Do you think splitting notifications into many tiers is unfair? I don't think so. I suppose I could test it, but I don't want to. The very thought of a notification per second annoys me.
I'll go with my initial assumption since I am too chicken to test it.
The bottom line is that all authors are not created equal to all readers. All posts are not created equal, either. Some posts by some authors are more interesting to me. I want to know about those posts as soon as possible. A near-immediate notification is the way to go.
Other posts are less important to me.? A daily summary of all such posts as a group is good enough. It's still a far cry from where we are now.
We don't all have the same tastes. We don't all have the same desire to see the same posts. Would this dilemma even exist if we did?
What might our options be?
Let's say we shoot off an immediate notification for your faves. We then do a daily summary for the second batch. Everyone else goes weekly.
Everyone, regardless of their category, is added to the feed, That would be in simple chronological order.
A summary could be short. List the writer’s name and the post title and I'm happy. Make the title double as the link to the original, and I’m thrilled. Add two or three lines from the post and I'm in Heaven.
Posts not set with a priority would go on a weekly summary. That summary could follow the same structure as the daily summary.
Finally, we could have a Twitter-like feed for everything. As the Web app finds content, it adds it to the feed. Depending on activity, that feed could be slow, or crazy active. I?doubt it will ever reach Twitter/X activity levels.
How it might work
The system handles the rest.
Ideas? Thoughts? Concerns? List them in the comments. Maybe we can get our discussion after all.
Paul
I'll eventually get around to creating a new sign-off.