Who Spiked the Enterprise Activity Stream?

vodka-shot

vodka-shotAs someone who leads products and even before I did, I’ve never understood the obsession with feeds and activity streams in the context of enterprise social software applications or for that matter, enterprise software applications. The bedrock of the fabled “Facebook for the Enterprise” meme was the feed and enterprise software applications rushed to re-produce an experience that was predicated on the feed and activity stream.

But herein lies the rub: I value those I follow and friend on Twitter and Facebook a lot. But to be honest – if I miss one of your updates, my life will go on or if it’s important enough, your update will likely find me on some other channel. In the enterprise setting though, that won’t cut it. You can hardly tell your boss that you missed an important update from him cause you didn’t happen to be watching the stream. And more striking, there isn’t a single reliable piece of market research that suggests that enterprise users have been clamoring for a feed to gate every application they use. Go ahead and poll your users and ask them if they visit applications only to sit in front of a stream of data that may or may not be applicable to them. Yet, many enterprise software pundits and insiders are enamored with feeds that show endless amounts of business data being pumped into it. And those who I converse with regularly know where I’ve always stood on this topic.

Finally, the Atlantic has now declared 2013 to be the year when “the stream has crested”.  I love this explanation of why we gravitated to the feed:

“There are great reasons for why The Stream triumphed. In a world of infinite variety, it’s difficult to categorize or even find, especially before a thing has been linked. So time, newness, began to stand in for many other things. And now the Internet’s media landscape is like a never-ending store, where everything is free. No matter how hard you sprint for the horizon, it keeps receding. There is always something more. “

And an astute analogy:

“I am not joking when I say: it is easier to read Ulysses than it is to read the Internet. Because at least Ulysses has an end, an edge. Ulysses can be finished. The Internet is never finished. “

So in other words, to make up for our technical inability to tame the information tsunami otherwise known as the social web, we placed a reverse chronological fire hose in front of users in the hope that she might just capture the few droplets that matter? Exactly.

Francine Hardaway, in a conversation with Steve Rubel, nailed it:

@steverubel I don’t even think it’s a very good piece. The very idea of a stream is that you don’t HAVE to finish it. You dip into it.

— francine hardaway (@hardaway) December 15, 2013

The feed was never designed to drive closure. I can live with this in my personal social networking working world but sorry Jose, no way that’s OK when it comes to my world of work.

There is a place for a feed in enterprise as part of a larger tapestry of interaction models. It’s an excellent way to ambiently learn and get wind of many things that I can contribute to. But the primary design metaphor for enterprise products cannot be serendipity-driven in the hopes that the right information from a colleague or a software application that I care about will show up. The world of work demands a significantly more decisive design– to facilitate closure of the repeatable tasks that both my manager and I know I have to take care of, yet with a facility that that helps me manage exceptions that will undoubtedly show up, unannounced. And like Ulysses, the world of work is predicated on a rapid fire of short bursts of closure, on our way to the big finale.

It amazes me how the geek inside all of us building enterprise applications got so inebriated by activity streams that we forgot the basic tenets of how work gets done. Business applications that will ultimately resonate won’t be about transactions or about social feeds but understanding the interplay between data, people, applications and content to get stuff done.

Alexis Madrigal’s conclusion is spot on:

“I think people will want structure and endings again, eventually. Edges and balance are valuable.”

Being drunk on the feed was a ton of fun. Time to sober up.

Enter your email address to subscribe to this blog and receive notifications of new posts by email.


Tags:
17 Comments
Post Details
17 comments
daverowley
daverowley

A stream is a lossy queue.  For some use cases, a lossy queue is fine--you are sampling and guaranteed delivery of any given item isn't essential.  For other use cases, the loss of data--"dropped packets" if you will--is unacceptable.  I think the business scenarios you have in mind are of the latter kind.

Capstera
Capstera

Good post. I think not every activity is worth noting or notifying. Also, in all this noise of the activities, often the signal gets lost.

bhc3
bhc3

Sameer - I was actually mindful of your previous post on activity streams when I was designing it for an enterprise innovation management platform (a few months ago). Two thing about them that I deemed to be important.



First, I didn't want the firehose treatment. Too many enterprise activity streams are a blast of all activities. So right off the bat, the notion of cascading relevance was built into the stream. "My stuff", "Following" and "Everything". Updates on My Stuff and Following are slower paced (just a mathematical reality), and if pressed for time those are what I want to see.


Second, I lookmed at activity streams as analogous to a tree. In a typical site, there's all sorts of activity happening at the end points. People are posting ideas. Ideas get comments, votes, evaluations, etc. These actions and content were the "leaves" at the end of each branch. The thing is, it's hard to know what's happening at the end of each branch. I considered it a form of "engagement friction" to have to climb out these various branches, check out the leaves, climb back down, go up another branch, etc.


Activity streams bring all these leaves down into a central trunk where it's much easier to catch up on, and interact with, what's happening. Less chance of missing the stuff I care about.

Bitvore
Bitvore

Great post.  I think you are right on.  Unless your job is to watch the streams (and I hope that's no ones fate), the business value lies in small signals hidden in lots of noise.   We think the right approach is to let the computers do the work of watching the streams and separating the signals that are important to you from all the noise.  Most business value comes from giving the business people the digested nuggets in a familiar format.


That is, unless your a marketing person trying to create more of the noise or echo themes for your own posterity.

marie_wallace
marie_wallace

I would definitely agree that while the Social Stream (be it the Twitter Firehose or the Enterprise Activity Stream) is great for a nice refreshing swim, it's not something the average person wants to drink from, at least not without some serious filtering. However as a data scientist, I absolutely LOVE the activity stream because unlike humans a computer is more than able to drink directly from the firehose. It can suck down as much as the stream can throw at it and start to synthesize, correlate, extrapolate, infer, and derive insights. So from the perspective of bigdata analytics, I see a bright future for the Activity Stream or at least for the underlying event bus. It's the fuel that drives the engine of social analytics insight :-)

SameerPatel
SameerPatel moderator


@moehlert great rebuttal. If something changes that warrants this suggested org design and change management we've been seeking for over 7 years, I'll be very happy. :)

SameerPatel
SameerPatel moderator

@daverowley nicely put. I use the feed and find great stuff on it like the next guy. But this post is about the latter as you point out. 

SameerPatel
SameerPatel moderator

@bhc3 Hutch I love the logic and in many cases theres absolutely room for this. But let me ask: should we be questioning the need for a feed (look, it rhymes!) altogether or are you in the mode of making it as useful as it can be now that we in the tech community have been fiddling with it for so long? 


SameerPatel
SameerPatel moderator

@Bitvore Sounds great. I think we underestimate the work that machines and algorithms can do. And with the right context, machines and algorithms could nail it.

SameerPatel
SameerPatel moderator

@marie_wallace Awesome, Marie. I can totally see why a data scientist would like to do this. No question that theres big value in data analytics here but I still ask : are we fixing a headache that we created and one that the end user never asked for? :)

bhc3
bhc3

@SameerPatel I believe feeds serve a valid job-to-be-done: I want to know what has happened with the content and people that interest me. But one thing I always aim to do is separate the JTBD from the way the job is satisfied. So from that perspective, are there other ways to satisfy that job? Sure. Algorithmic recommendations are an alternative way to do this. The system determines - based on explicit and implicit signals of interest - what I would want to see.


Feeds are simpler to implement. Algos are tough. But I could see them satisfying the JTBD well with some iterative effort (and patient customers).

marie_wallace
marie_wallace

l totally get where you are coming from and would agree... however because I would argue with my toenails ;-) I would ask does the end-user ever explicitly ask for any new invention? Did anyone ask for the Smartphone or Twitter or even the Internet? Again coming from my analytics perspective, perhaps the Activity Stream is just a means to an end. Everyone wants a great transport system (planes, trains, and automobiles) however don't care about what infrastructure is used to facilitate that (bridges, tunnels, ferries,...). The activity stream could become the info highway that facilitates the analytics needed to realize cognitive computing!

marie_wallace
marie_wallace

I'll give you that Sameer :-) But taking my other analogy, I doubt most people would jump up and down about the combustion engine or reinforced concrete, but we need both for modern travel. I guess where we are is that the activity stream is less interesting to the end user than it is to the folks looking to use it for building other capabilities for the end user. It's a means to an end.

SameerPatel
SameerPatel moderator

@marie_wallace Ive been waiting for someone to play the "did anyone ask for a smart phone". card :) No they didnt. And we found out that we cant live without it when it was invented. Don't think you can say the same about the feed if you polled the same percentage of users as you would a smartphone. 


Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Follow

Get every new post on this blog delivered to your Inbox.

Join other followers: