No Users, No Glory: 10 Principles of Enterprise Social Software Adoption

It would seem that running a startup and introducing an enterprise social system at a large corporation are totally different businesses. Different they certainly are, but there is one thing that makes them surprisingly close to each other. In order to be successful, both ventures have to build a devoted customer base. It simply doesn’t matter how beautiful your product is if no one is interested in using it.

Unlike traditional business software, enterprise social systems have a very strong dependency on user adoption – just like a new service offered by a fresh Web 2.0 company. There are multiple reasons behind this dependency (you can find my take on these reasons in this Yellow Brick Road post at FASTForward). Regardless of what’s causing it, the truth is simple: No users, no glory.

In this post I’d like to summarize some lessons I’ve learned (sometimes the hard way) while making a highly social internal system adopted at a large company. Enterprise Social Computing team I lead at Microsoft is responsible among other things for developing and driving adoption of CodeBox – Microsoft’s internal shared source system for collaborative software development (more on it here). An example of domain-specific enterprise social system, CodeBox had many great features when we launched it a couple of years ago. What it didn’t have was users. And getting these users has taken us much, much more than making the system available and announcing it.

We’ve found that the key is to think like a startup, treating employees like customers and fully realizing that they had been taking care of their business before your solution existed, and will move on if you go bankrupt. While the exact list will vary depending on the system you trying to get adopted and factors like organizational dynamics, here are some core principles that, in my opinion, seriously help with adoption of majority of social software.

  1. Know who your users are, what they need, and which of their pain points you can address with your solution. If you don’t get this one right, nothing else matters. Unless you understand who your target users are and know (not just assume) what they need, you will be very lucky to get any users at all. There’s no need to conduct a year-long study, but you have to do your homework.
  2. Plan and execute on your adoption strategy. In startup terms, take your customer acquisition seriously. Broad adoption very rarely happens on its own, and to make it happen you need a plan.  
  3. Run pilots and early adopter programs, targeting people who really care about the problem space. They may be harsh in their criticism, but their feedback is invaluable, and once you get them on board they become your most active promoters. Going big is typical for enterprise deployment, but resist that temptation. Getting a hundred enthusiastic users is way more important than exposing your system to 5,000 bored looks of people who will never come back.
  4. Build and nurture an active community around your system. While related to #3, this principle is about helping your users to start relying on each other. Create channels for them to communicate, expose these channels in the right places of your system, make it very easy to participate. And most of all, be involved. Translation: in the beginning be prepared to answer every single question and comment.
  5. Define a clear set of metrics and track them meticulously all the time. These may be basics in the startup world, but they are too easy to ignore when you’re launching an internal enterprise system. These are not your traditional availability metrics (though they can’t be ignored either). You have to know how many users come to your system. Where they come from. What they do while in it. And most importantly, how many of them come back. Trends are your friends.
  6. Expose your new solution in the contexts where your users currently operate. New shiny destinations can be tempting, but users don’t have time to go too often out of the tools they live in. If your users spend their day in an email client, don’t hope they’ll be happy to add a great new site to their daily routine. Find a way to expose your system in their mailboxes. 
  7. Identify the most successful customers and give them stage to tell their story to others. No pitch is more convincing then an authentic experience of a peer. Organize panels, create case studies, promote success stories.
  8. Show the way for your users. The benefits of a new system are much better understood when someone articulates them for you. Launching the system and even providing feature-level help is not enough. You need to show why anyone should care about it. Create appealing meaningful demos, explain, elaborate, engage.
  9. Walk the talk. In other words, use your solution everyday to do some (better yet, a lot of) real work. You’d be surprised how much you learn when your own deliverables depend on the system you thought was perfect for others. Plus, when you say “this how we use it in my team” this means a lot.
  10. Be flexible and adjust as you go. Watching the trends and listening to user feedback doesn’t mean much unless you act on this data. Sometimes it’s about making minor adjustments, and sometimes it’s about seriously changing your priorities. The good part is that your metrics and your users will always tell you whether your adjustment has worked.

Note that this list is about adoption; it does not cover many things you need to take into account to design and build the system.

As of today, thousands of people use CodeBox on regular basis. New projects appear daily. People connect all the time with others. And in some way it is ironic that in order to make this happen within a large corporation we had to think like a tiny startup.

The Curious Case of Enterprise 2.0

Last week FASTForwardBlog has posted my post The Curious Case of Enterprise 2.0. I’m not a big fan of crossposting (while it may be good for traffic, it sort of kills the conversation, which is one of the main things I’m looking for in a blog post). So here I’m following the movie trailer model and posting only the opening paragraphs.

People in the technology world love creating new words. In fact, they are responsible for the most of the recent English language growth. Be it telephone, internet or crowdsourcing there’s always a new technology behind the new term. The techies have even managed to introduce their trade traditions into the mass conscience. When was the last time you used a sequence of dot-separated numbers to describe a large official organization? Yet all the talk about Government 2.0 doesn’t seem to surprise anyone. The lack of surprise however doesn’t imply shared understanding. Just try asking ten people who use the term Web 2.0 what exactly it means – and most likely you will get ten different answers. 

Although 2.0 memes are everywhere, hardly any of them have generated as much controversy as Enterprise 2.0 (a.k.a. E2.0). Introduced by Andrew McAfee in 2006 and later expanded by others, the term seemed to be a slam dunk for a while. Forward-looking and fresh, it was (and still is) about application of social software and corresponding approaches in business. However, somewhere along the way something didn’t go quite right.

If you have any comments after reading the post at FFBlog please join the discussion over there. BTW, if you have any good crossposting tips, please share them here,  as I plan to keep posting on FFBlog from time to time.

Hello, human: A Dialog With Wolfram Alpha

If you like reading tech news, visit sites like TechCrunch or simply happen to have enough tech people among your Twitter contacts you must have heard about Wolfram|Alpha. In case you haven’t heard about it, here’s a 5-word summary: it’s a new search engine. Well, technically it’s a computational knowledge engine, but you wouldn’t know that by reading most of the news coverage prior to the site’s launch last weekend. Some said it will be cool, others predicted it’ll be important, yet others proclaimed it’ll be huge. And obviously, some said it’ll be a compete failure. Now as the site has launched we will undoubtedly see reviews comparing it to Google and other search engines, praising or questioning its capabilities, and wondering about its business model. Meanwhile, I think, Wolfram|Alpha has already a remarkable achievement under its belt: it has proven that even today people feel that there’s more to be found on the Web.

Think about it: you can essentially find anything these days. Between major search engines and hundreds of special ones you can find sites, books, videos, songs, pictures, products, people, articles, businesses, free stuff… just about anything. Sometimes it may take a bit of an effort, but it’s not that hard. The world’s knowledge is no longer a black box it used to be not so long ago. Sure there’s the deep Web and the promise of semantic search, but most people searching online don’t even know about these things. So why would anyone even care about a new search engine, yet alone care enough to turn it into a hot topic? One answer could be that people have been always seeking better ways to explore the world around them. Another possibility is that people want new cool toys in any area, be it mobile phones or search engines. Or it could be that the buzz about Wolfram|Alpha has more to do with its main creator than with its capabilities (though this wouldn’t explain similar buzz around Cuil back in 2008).

Whatever the right answer is, it seems that Wolfram|Alpha is ahead of the pack in one aspect that is rarely used to compare search engine: sense of humor. The only engine that could’ve competed with it in that department would’ve been Ms. Dewey, but she’s no longer with us.  What you see below is a real dialog I had a with the system (or rather with the people who equipped it with such distinctive behavior). Whether you need another search engine or not, it’s worth a test drive — if for nothing else then at least for the pleasure of asking a computer about the meaning of life — and receiving an answer.














Twitterator 3: Rise of the Machines


I’m not a big fan of writing about the same topic twice in a row, but this development is too curious to ignore. We are witnessing evolution in action. And rapid evolution at that. The artificial creatures I described in the previous post have been evolving quicker than radioactive mutants. Over the last few days Twitter spam has been getting serious attention from bloggers and media alike.  Leena Rao has written about it on TechCrunch (re-posted same day by Washington Post). Seesmic’s Loïc Le Meur outlined a potential solution. And of course this has been a major topic of discussion on Twitter itself. And what the Twitter bots (a.k.a. twones) have done? They’ve evolved.


Judge for yourself. Four days ago the bot that decided to follow me was a boring creature with an unpronounceable name, no real feed, a standard background and an absolutely dull default picture. He was even boring to write about. Four days later that poor creature is history. A new generation twone who has started following me today has a name (meet Nancy), a custom (yes, custom) background picture, a bio proclaiming love of parties, and a picture that is rather far from the default. But most importantly, it has a feed. Unlike her simpleminded ancestor, Nancy does not want to raise suspicions by jumping straight to business. Instead she starts with innocent chit-chat. She tweets about flu, about her excitement over Twitter and about some other events. It’s not until her sixth message that she gets to the point: free ringtones.


Does she have siblings? Of course, she does. Two pagesof them. While they all seem to share the same bio, name and background, and even tweet feed, they do look different. The result? Dozens of responses from people they chose to follow. After all, Nancies look way more real than the first-gen twones. The most interesting part are re-tweets. Now that Nancies are so real, their tweets seem to be worth re-tweeting and in an ironic twist, real people start helping out spammers without realizing who they are dealing with.




If Avayxrlxry was the original Terminator, these bots are more like T-Xfrom Terminator 3. And this is just the beginning. In just four days Nancies have become way more believable. Six months from now they will be almost indistinguishable from real people – at least without a serious effort. And that may turn into a killer of Twitter’s promise to become the world’s chatter channel. What’s the use in analyzing chatter of 50 million people, if half of them are spam programs? Not only this may turn businesses away from Twitter, but it may also taint Twitter’s biggest accomplishment – popularization of open data pipes (which among other things has resulted in Facebook’s today’s announcement). One of the major factors behind Twitter’s success is its openness. It’d be bad to see this idea labeled as bad for business apps. So if you’re in business of writing a Twitter client invest into spam detection and protection. You’ll help a great idea to become a bigger success, and will be among the first to offer functionality that soon will be critical for businesses. Nancy wouldn’t mind – she’ll be too busy selling ringtones.

Twitter Wars: Episode II – Attack of the Twones



Remember the first time you got spam email? This probably happened in the last century, in fact, in the last millennium. But there was that moment many moons ago, before junk filters got introduced, and before spam became a fact of life, when you looked at your mail and there it was – an automatically generated message trying to sell you something. Irritating or curious, it was signaling that the system had gone mainstream. Sharks had moved in. Same is true for any communication system. The moment it becomes successful it starts attracting predators.


This day has come for Twitter (Dharmesh Shah captured this really well in his  Mashable cartoon). Going mainstream has a price. These days many twittizens worry about stars rushing to get the first million of followers. But star attention is not a problem and neither is the fact that large companies are getting serious about Twitter. These are simply signs of growing up. The problem is this little excited follower who decided to follow me today:




The force is strong with this strangely named user. Despite a whopping single update, this young Anakin is following 905 people and moreover, has followers. His “free backberry” wisdom has already attracted 71 follower – and likely will attract more, once it finds another 900 people to follow. A simple search for that blackberry line reveals that Avayxrlxry has over twenty siblings — all busy pointing at the same site, all having a single update, all following nearly thousand people each and all having followers. Some of them are already suspended, but most are alive and well.


Now imagine thousands, if not millions of these automated creatures moving to inhabit Twitter and what they can do to the infrastructure, not to mention search and tag clouds. As companies start relying on Twitter for things like competitive intelligence, they may find themselves in a totally different Twitter galaxy a year or two from now — one that is far, far away from the relativly clean and reliable Twitter or today. Three things make Twitter so attractive for businesses: reach, relationship building and, most of all, data mining. And that last one can be affected significanly by armies of Avayxrlxrys. The old “junk in junk out” rule is fundamental for data mining, and Avayxrlxrys are great at producing junk in excessive amounts.


As Twitter matures and moves from the days of early childhood to fame to a platform used by large businesses, it would have no choice but to keep investing into building better spam filters and smarter suspicious activity recognition modules. It would have to walk the same path that email has been walking for years, only it would have to walk it faster. Twones are closing in.


Scale, Speed and Context

We live through a transformation that is as fundamental to the world, as the one that brought us all online. People around the globe are getting connected – not to the internet, but to each other. Thanks to the rapidly evolving technologies, these connections happen on a scale, with a speed, and in contexts that were hardly imaginable just five years ago. Because of these three factors—scale, speed, context—these connections are changing the world in every aspects. And while there isn’t a single area of people’s lives that hasn’t been touched by this change, everything we’ve witnessed so far is only a preview for things to come.


We all wear different hats on any given day. We’re consumers, friends, family members, readers and taxpayers. Most of us are also employees – people have to wear some variation of that hat to feed the rest of the hats. That hat is always slow to react to the fundamental changes, but once it does, it goes all the way (just think how many corporate sites, not to mention intranet sites, were around in 1996). And now the change that has been revolutionizing the consumer space starts transforming the enterprise. Make no mistake – despite thousands of major companies building presence on Facebook and Twitter, the enterprise and social technologies are still in the early dating stage. Most companies still don’t quite know yet what do with all these technologies, and most importantly, how to use them to significantly improve their business. But the process has started and it is irreversible. What we’re about to see over the next five to ten years will represent a dramatic change in the way companies function. In fact, it will have more impact on the business world than the internet. The internet has changed how companies communicate and do business.

The connections revolution will change how they organize themselves and how they function. 

Working for a major corporation (Microsoft that is) and having a job of leading major efforts to build and adopt social within it, I couldn’t help but realize that any large company  is going to deal with this sooner or later. Once you spend month after month talking to people across the company and the industry, identifying right scenarios, shaping the technologies and thinking through what would it take to adopt them, you start seeing the enormous impact social technologies will have on people at any company. Because at the end, it’s all about people. About connecting them with each other and empowering them through these connections. And Connections is the place to have a dialog about this change and—in a way—to shape it through this ongoing conversation.


The comments button is one click away… you know what to do.


Get every new post delivered to your Inbox.