Lessons learned at Enterprise 2.0

We’ve been in Boston this week at the Enterprise 2.0 conference (which got a great writeup in the Wall Street Journal today.) It’s been an interesting series of discussions about next-generation IT applications. Here are some of the lessons we’ve learned over the week.

2 dimensions of attendees

The conference felt like it was torn between audiences. I wanted to put everyone in a room and split them into IT (on the North side) and business (on the South side) and then split those groups in two by external, customer-focused apps (East) and internal, employee-focused apps (West.) Then I could have sensible conversations with a group (“IT platforms for connecting to customers” in the Northeast corner, for example.)

But when I spoke to attendees, many said they liked the overlap and lack of clear divisions, because it allowed serendipity to creep in — a marketing manager who learned about some new technology; or a community manager that learned how to turn an externally-facing community inwards.

Things to consider in deployment strategies

There’s a huge spectrum of deployment approaches out there. Cloud computing offers a bunch of them, from IaaS to PaaS to SaaS. But there are also economic models — freemium, pay-as-you-go, and so on — that change the investment needed. Ultimately, these trends mean the upfront investment is moving quickly towards zero.

So what happens to ROI when the I approaches zero? The short answer is, we experiment a lot more. But while we no longer need to justify IT investment up front, we now need to prove value after the fact. Because of this, there’s an increased focus on measuring adoption, engagement, and the actual business benefits of those experiments (and killing the ones that have the lowest value.)

Enterprises need analytics

A corollary to this is that analytics matters. While web analytics has traditionally been used to look at public-facing web apps, enterprise applications face an adoption crisis. And to maximize usage and adoption, they need to take many of the lessons learned from public web optimization, such as conversion funnels and A/B testing, and apply them to their internal apps.

Tricks for getting your users to engage

There are some good tricks for engagement, however. Some of the ones suggested by one of the panels we ran were:

  • Pre-populate profiles with bad information and people will correct it. Leave it blank and they won’t touch it. Sometimes a lousy picture is the best motivator.
  • Have an ask and a purpose on profiles. Don’t just show static information — include things like, “these are the three things Mary needs help with now.” Then build an achievement/badge model into the app to create a web of social obligations in which you can ensnare your users, just like Farmville does.
  • Cut off the old ways. If your CEO is going to respond to questions on a new platform, get him or her to refuse to respond on the old ones. And measure decline in those old ways to prove the conversion.
  • Give access to a priviledged few, rather than expecting everyone to use something. Once you’ve created a perceived scarcity (the way Google did with GMail), then consent to the demands of the resentful unincluded and you’ll get better adoption.
  • Automatically map legacy content into the new format. Convert emails people send into message threads, forcing them to go to the new app to respond.

IP concerns are an impediment to community adoption

We had a great discussion of IP issues and the dark side of cloud computing. The law is a decade behind clouds and SaaS, and it’s unclear whether you’re properly protected by the contracts you sign, particularly across international borders. As democratic, open computing becomes more popular, IT needs to reassess the balance between the upside of openness and the downside of exposure.

IT has the keys to the kingdom

Many community managers seemed at odds with IT. But IT has the keys to what happened. If you consider a profile, a useful metric isn’t “completeness” or “how many views.” What matters is “how many times did someone connect with this person and have a meaningful exchange?” or “how many times did someone find content created by this person because of the profile?” In both cases, it’s IT that has the hooks into the messaging and document management systems that complete the transaction.

Similarly, while much of the conference dealt with “active” collaboration — profiles, chat, and so on — and was therefore obsessed with encouraging engagement, there was much less focus on “passive” collaboration. Passive collaboration involves data mining, watching what URLs and documents people use, and so on. It’s built into the IT systems, and doesn’t require active participation by the end users.

I think IT — prodded by companies like Microsoft, Cisco, and Oracle — will completely bypass many of the active, business-led initiatives we see today and take over the platforms on which social business, both inside and out, is done.