In 2010, I tried building a self-prioritizing RSS reader. Three friends and I built a prototype and called it Deliverss. We thought the two s’s on the end were clever. The idea was that posts could be crawled for their key topics or themes using an API called Alchemy and that the user would train the system on what they actually read through article opens, time reading an article, thumbs up/down, clicks to other articles, etc.
Within the sites that you subscribed to we wanted to discover and present what you would actually read. I still think this is a good idea.
I’ve been fascinated and frustrated with RSS since 2007. I like how easily I can follow a site or feed. But I’m frustrated and eventually abandon every RSS reader I’ve ever used.
Twice a year I log back in to Reeder, filter out sites I don’t read anymore, add some new ones. Then I check it for three days, less so each day, until I’m back to not using RSS for another few months.
Using RSS over the years, I started creating and managing three folders: high, medium, and low-volume publishers. High-volume publishers publish 20+ times per week. Medium-volume publish 10–20 times per week. Low-volume publish zero to nine times per week.
Over time, I unsubscribed from all of the high-volume publishers because it’s just too much to keep up with. The New York Times, ESPN, Esquire, or TechCrunch are like Twitter: I don’t need to see it all. I’m happy dipping in and out of the stream as needed. The most important items or the ones relevant to me specifically will find me somehow, usually through email or virality.
I end up loosely following the medium-volume publishers. They’re probably the ones that sink my RSS use because some are just enough to keep up with, some are way too much. It depends on the source’s publishing frequency, length, and content density. Brain Pickings is a good example of a prototypical medium-volume publisher for me. Amazing, deep, thought-provoking articles but way too much to read and keep up with regularly.
Low-volume publishers are usually pretty easy to keep up with but can occasionally be overwhelming for the same reason as the medium-volumers. Fewer posts but longer posts, more links, a lot of interesting digressions. For the most part, the low-volume publishers are easy enough to keep up with.
At the end of the three day experiment I end up with a loosely curated list of sites I either like a lot, think I’ll like given some more time, or the ones that my aspirational-self wants to like.
And everything I found in my RSS experiment that I like but don’t have time for right now, I add it to Pocket or Instapaper.
Time-shifting is one of the most powerful tools we have to manage content online. Mailbox and Google’s Inbox have proven how necessary the ability is for managing our inbox.
In the case of news and articles, time-shifting is the Facebook-event “Maybe.” It’s a non-committal push to some vague point in the future where our mental desperation for something to do may override our laziness and impulse to keep deferring the the decision.
I want read-it-later apps to work so badly for me but they don’t. They’re blackholes for articles I’ll never get back to, like RSS readers end up being for incoming content.
I jokingly tweeted this earlier this year, which people seemed to enjoy.
I believe strongly in time-shifting but it can’t function on its own. It has to be coupled with its opposing force — Decay.
Think about some of the friends you had in college. Think about how tight you guys were at the time. Maybe you were roommates, or they lived on your floor. Maybe you had class together, played video games, went to football games, dinners, and parties. After college you got a job, they moved, and you went your separate ways. Even though you can probably pick up right where you left off, the bond just isn’t as strong as it was when you were 19. That’s decay.
In social psychology, after about nine months of little or no contact with someone, your bond weakens. Unconsciously in your brain they move from the center green circle below, to one layer out. In that outer layer, you associate more loosely and less frequently with that person. Without nurturing that relationship, you can decay further to the outermost circle or out of the concentric circles altogether.
This is some of the psychology that Path is based on. It’s also how Facebook’s EdgeRank work.
Instead of thinking of the sum of your Instapaper library as content, think of it as a cognitive load.
Imagine 50 articles that you added to Instapaper with good intention to read. Forty five of the articles you haven’t read, even after opening and closing the app a few times. You continue to push them to some undefined future point but more importantly, you start thinking about them more and whether or not to read them. They start to weigh on you because it’s an unresolved mass.
With a decay mechanic, an article that’s repeatedly pushed into the future begins to approach a cliff where it will be automatically deleted or archived since it wasn’t important enough to read weeks or months after you added it to Instapaper.
In designing a system with decay, we also have to design for loss prevention, the over-assigning of importance to something because we simply don’t want it to go away and feel the potential resulting pain.
Couple RSS and read-it-later apps together and you have a horrible set of tools for the avid online reader. That’s where we are today. Reading online is really difficult. I don’t need to go into detail about how thinly stretched and fickle our attention spans are.
As I write this, I have 14 tabs open, some updating with counts of unread items on that service. I have seven links I opened from a Sunday-morning email newsletter/digest of a blog that I enjoy. I have a Github gist with code I want to incorporate into StopCoin, a Quip doc and a Google doc open with ideas I’m fleshing out with friends, my phone is buzzing with texts, and I’m weighing, for the seventh or eighth time, whether buying an iPad will solve all of my reading wants and woes. I’ve been distracted from writing this very paragraph a few different times as things have flashed on my screen, falsely demanding attention.
Enough with the problems, let’s talk about solutions.
RSS and read-it-later aren’t fully developed yet. Much like foursquare and Gowalla popularized checking in and brought location sharing to many other apps, Instapaper and Pocket introduced the world to time-shifting. Many apps have a bookmarking or read-it-later feature and that’s usually where it ends.
It’s easy to say read-it-later services are poorly designed. It more correct to say that they’re not yet finished being designed; we haven’t discovered the mental processes, triggers, and workflows needed to create the ideal reading experience online.
Here’s what I’d build:
- RSS should be rebranded as Subscribe on every website that offers it. Everyone is familiar with what a subscription is. Many people look at the RSS icon and click it, wondering what the heck that is.
- The RSS reader and read-it-later app should be the same. A one-stop-shop for all of your digital reading.
- The service itself should be opinionated. For example, you shouldn’t be allowed to subscribe to high-volume publishers. The correlation with abandonment is too high.
- Articles should show keywords about themselves and users should be able to veto them. Vetoing a keyword will pre-filter articles that include the keyword from appearing in your feed in the future.
- On the read-it-later side articles should be filtered by reading time. Articles should also be ordered by least decay at the top to most decay on the bottom in an attempt to prevent people from trying to restore and save articles that they’ve never gotten to and most likely never will.
- The web and mobile experiences need to be drastically different. We interact so differently with our phones versus our iPads versus our laptops.
- Users should be able to use either Subscribe or read-it-later without needing to use both.
- The mobile experience should probably only show quick-to-consume articles. Or it should ask the user how much time they have and present articles that fit in that timeframe.
- Both the Subscribe and the read-it-later sides of the app should have an optional feature for use when you’re finished reading to note why it was important to you. Two reasons for this: 1) So you spend some time reflecting on an article to make more sense of it and 2) so that when you search for it later, you have some context around why you enjoyed it in the first place.
- Unread counts should never exist.
- The design should center around removing cognitive stress, not adding to it.
- Mailbox-like delay controls are important to control moving articles around but need to be carefully design so the read-it-later functionality doesn’t feel like an inbox.
- Subscribe/read-it-later should be a library of sorts. Something you visit for pleasure for a pleasurable outcome. It shouldn’t be another list of things to get through or feel like homework.
- It should be a freemium service. The premium version could give you more control over the underlying decay levers, for example.
I think this could really work; I may be totally wrong. In any case, that’s how I’d fix reading online.
If you enjoyed this post and want to get more like it, along with early betas, exclusive posts and more, sign up for my email list.
Zack Shapiro is the creator of Built in Public, a project studio and year-long experiment in creativity and transparency. StopCoin is the first project to come out of Built in Public. Zack also founded Luna, a nighttime delivery company based in San Francisco, which was acquired in early 2014.