T. Rowe goes 2.0

These days T. Rowe Price vice president Kirk Kness is busy implementing relatively low-cost, Web 2.0-based collaboration tools including wikis and blogs.

As vice president of T. Rowe Price Associates, Kirk Kness is responsible for using technology innovatively for the Baltimore-based investment company. In past positions with T. Rowe he was involved in multimillion-dollar infrastructure projects, but these days he’s busy implementing relatively low-cost, Web 2.0-based collaboration tools including wikis and blogs. 

The company’s goal in using Web 2.0 is to gather intelligence more effectively companywide as it’s created, and make it easy to find, share and use — perhaps even in ways the company has yet to envision.

What specific Web 2.0 applications do you have in place?

We have the Confluence wiki platform [from Atlassian Software] in place in our call center. We have [Jive Software’s] Jive Forums [discussion forum] platform in place. We have pockets of instant messaging rolled out. We’re starting to look at another layer of instant messaging called persistent instant messaging, which is a product called MindAlign that Microsoft just bought.

But it’s not really about the tools; it’s more about what you get from the tools. The real value that we’re trying to get — and what I’m trying to drive towards — is in context with these tenets that describe what Web 2.0 is supposed to bring to an enterprise: harnessing collective intelligence, discoverability, low-barrier tools, emergent structures, collaboration, unintended use and community. They are the factors that we’re really starting to try to figure out. How can we drive to those using these different tools? What can they bring us?

Let’s start by explaining that first one, harnessing collective intelligence.

That’s the ability to actually gather the tacit or the implicit knowledge and bring it together. So, in the context of collaboration, how can we create tools, how can we put in place models that allow people to collaborate who maybe naturally wouldn’t collaborate? I don’t believe that if you put a tool out there, and you say, ‘Everybody collaborate now, because it’s for the better of the company’ that it’s actually going to happen. You collaborate around a community that forms naturally. And that’s the thinking that we’re trying to bring in to the enterprise. How can we better enable what I call “smart information” or making content smarter so that we can make our employees smarter? By making our employees smarter, we then ultimately make our clients more informed.

OK, how about the other tenets?

Discoverability is basically the ability to get the content into forms or formats so it can be discovered. That’s the linkable model that’s used on the Web, so everything is discoverable. If information is locked in an e-mail somewhere, it’s not discoverable. If it’s not discoverable, then it cannot create community.

Low-barrier tools are really important. We cannot create friction for our associates. With traditional Web content tools, it tends to be difficult to enter content. Wikis are public-authorship-type tools. It’s “what you see is what you get”. You may even have the ability to author content in Word, and then push a button and publish it out. That’s why blogs exploded — because [publishing] became easy. If you think about Web 1.0, it was about creating Web pages and knowing HTML. How many people do you know who have a blog also know HTML?

Emergent structure is what emerges from the collision of this information. That’s where tagging and folksonomies and discoverability come in. I can tag a piece of content and not know that Brian down the hall had another piece of content that he tagged similarly, and all of a sudden there’s a relationship between those two pieces of content.

What do you mean by folksonomy?

Thomas Vander Wal coined the term folksonomy in 2005 to mean an ordered set of categories that emerges from how people tag items. This is directly out of a book called Everything is Miscellaneous. That’s as opposed to a taxonomy, which is when information is tagged upfront, and then you try to follow that order. Folksonomy is when the order emerges by the way it’s tagged. Aren’t you glad I have this book [Everything Is Miscellaneous: The Power of the New Digital Disorder, by David Weinberger, 2007] in front of me?

Yes, I am.

So, that’s emergent structures. The next one is collaboration, which is an outgrowth of how these things fit together.

Unintended use was the next one, which gets to facilitating the ability for unintended use. You may have a situation where information that was created for one context is viable in a second context. A good example is when Google launched Google Maps and then somebody built an application that basically did a mash-up between Craigslist and Google Maps. It showed on a Google Map exactly where apartments for rent in Chicago were located. Google didn’t intend for Google Maps to be used that way. But it spawned a whole other way of looking at something.

Community is basically allowing the emergence of community, allowing people to innovate by being able to share and communicate with each other.

Can you give an example of how all this plays out in practice?

Sure. Say I had a call that came into our Colorado call center, and the topic may have been about a particular policy, say a new IRS policy around the Katrina disaster, and how they’ve eased up on a tax deadline. The rep doesn’t know anything about it so he gets a hold of a liaison, and the liaison finds the relevant information and sends it back to the rep, who informs the client about what the ruling is. 

Now, what if another person calls into our Tampa call center with the exact same question? The ability to know that that information exists can make our reps more informed and more intelligent, which makes our clients more informed and more intelligent. Instead of saying, ‘I have to get back to you on that,’ the rep can more intelligently find this information.

Web 2.0 is really about information coming to you and to your space, where Web 1.0 was you go out and search for everything.

The call center is where you’re using the Confluence wiki tool, right?

That’s exactly right. In that particular case, we’re using forums and wikis to document transactions. We’re using tagging so that people can more easily find information, and we’ve got RSS feeds attached to that information so it gets published and is instantly available to anyone who has subscribed to a given topic.

How does the RSS part work?

That’s part of the platform. If you are watching a particular space, you have certain topics that you’re paying attention to. As soon as something gets published on one of those topics, it comes back into you as an alert, and you can see it instantly.

I would imagine you’ve got lots of different topics to deal with. How do you make sense of them all?

That’s a really valid point. Where we started with this is tax and IRA information, and we’re starting to branch out from there. So, we’ve organized our topics at the highest levels around things like contribution, moving assets, redemptions and taxes. So, instead of [defining] a brokerage contribution and a 529 contribution and a mutual-fund contribution or annuity contribution, it’s organized around contribution itself. Underneath that you can find how it applies to brokerage or 529s or whatever.

If I’m a call center rep and I get a call about a tax question, how do I find out whether some information exists that will help me?

A couple of different ways. We built this collaboration platform that they can go into that has a base-level navigation built in based on the topics I just described — contribution, redemption or whatever. We’ve also got a Google search engine in place where they can search on that topic across much more information [beyond what’s in the collaboration platform]. We’ve got RSS feeds set up so they can look at the recently updated information, if they’re just browsing for information.

And we’ve got tagging in place. We do use a taxonomy where they can tag content the way they see it. So, ultimately it starts to create emergent structures around the organization of content.

The second model, which we’re starting to move into now, is making the things more contextual so we’ll be able to map the content itself, based on the taxonomy, to the process. So it could be somebody is calling in and they’re interested in retirement planning. Now we can drive content across these different systems based on that topic, making it much more contextual, and drive it right into the [call center] application. That’s where we’re thinking of going now.

What tools do you use to do that?

We’re looking at the same tools. And one of the tools that I didn’t mention to you was an open source social bookmarking tool called Scuttle that allows you to bookmark a particular URL or a link and then tag that link. The tagging is what drives the relevancy to the topic.

What effect have the tools had in your call centers?

We’re now able to much, much more quickly get content available. We’ve been able to remove blocks around content flow and around content authorship. We’re basically making it easier to fix content, as opposed to hard to enter content.

We’re really still at the tip of the iceberg on this, but it’s starting to raise some awareness. Lots of different parts of the organization are starting to say, ‘Hey, we want to do this, too,’ because they’re starting to understand what this will bring. We’re starting to shift some of the thinking and potentially even some of the culture on how to look at this flow of information. 

I would imagine in ’08, it’s really going to start to take off even more as we start to really imbed and mature some of these strategies. So, the place that it’s the most prevalent, and the place where we’ve gotten the most traction is in the retail call-centers right now. But we’re looking very strongly at other areas in the company.

Have you been able to determine an ROI from implementing the collaboration tools?

We believe that it’s improving our call-center time, but we don’t know for sure. It potentially could be that calls are the same length, but customers are getting better and more information. We do know we’re radically improving the ability to capture and publish information, but we’ve never actually crunched the numbers to determine an ROI.

If you can’t determine the ROI, doesn’t that affect your ability to get funding for projects?

No. Because for this particular effort, [we know the tools are] working. And because, when we started this, we didn’t go after a big multimillion-dollar infrastructure-type project. We went after low-barrier tools, which were relatively inexpensive. We were able to do this at a fraction of the cost of traditional business tools. And we were able to have real value, even if it’s word-of-mouth value, pretty quickly. So, it kind of gives you some credence.

How do you get people to start using these tools when you initially roll them out?

You’ve got to create low-barrier tools, and you’ve got to reduce friction. If you don’t reduce friction — meaning, if you don’t present them in a way that makes it easy for them to use the tools, they will not use the tools. It’s got to be natural. If it’s just another tool that they have in their arsenal, it won’t matter. We completely replaced [older systems] where this content and information was stored with this new world.

What did you replace?

An old, first-generation Internet site that was all HTML-based in JavaScript and IFrames and difficult navigation — it made for a resistant process around content authorship.

What was the process around authorship?

It was first-generation content tools, and you sent information off to an area that knows how to create HTML. You put it into the pipeline and waited a month or sometimes longer because they got backlogged.

And now it’s pretty much instantaneous?

Some of it is instantaneous; some of it isn’t. But it’s much, much smoother. Some of it is within hours or days vs. weeks or months.

In terms of getting people to use the tools, once the tools are out there, do you tell people they have to use them or just encourage them to?

It depends on the area. It’s a little different in the call centers where it’s actually an enabling tool for the reps because the information that they need is in there. And when we turned off the old content, the news tools were pretty much all they had. Some of the other areas where we’re looking to use the same model, it’s going to be much more difficult because it’s not as explicit. It will be more implicit. If we’re trying to use it for, like best practices around sales processes and things like that, we’ve got to make it easy, and we’re going to work through that. And come back to me in six months, and I’ll let you know how that’s going.

Isn’t there a certain level of marketing that has to go on, too?

Yes. There’s an absolute level of what I call rah-rah. You got to find your enthusiast, if you will. And we did. Some of our tax liaisons were very enthusiastic about this. It made their job easier, and they were the ones that the reps were talking to all the time. So, they were the ones who started getting excited about it. And that definitely makes a difference.

What would you say are some of the biggest challenges you had in rolling these tools out?

Culture and understanding. Here’s the thing. I’ve been on the phone with you for the better part of 45 minutes. And I’ve given you a dose of a year’s worth of thinking, if not more. And there’s so much, much more to this on how this stuff fits together. So, when you get in front of a senior manager and you’ve got five minutes, how do you convince them this is what we really need to do? So, you’ve really got to do your homework. You got to be really prepared and ready to do small, little things to kind of chip away at the bigger vision.

What would you say has surprised you the most about your deployment so far?

Two things. One, the adoption rate surprised me. It’s actually catching on like wildfire. And I don’t know why that would have surprised me, but it did to a degree. The second thing is the fact that the reps have the ability to comment and tag content, yet they don’t do it because they don’t think they can. We have to tell people, ‘Hey, you’re allowed to comment and you’re allowed to tag that content.’ And that’s a cultural thing. But we’re playing through that.

Any big disappointments?

Probably not at this point. I will tell you, we definitely have some churning that’s going on. And here’s an example. I met with one of the head guys yesterday, and he made this analogy, which I thought was really good. He says, ‘Kirk, this is what I see. It’s kind of like you’re in a motorboat, flying across this perfectly pristine lake, going a million miles an hour. It’s smooth sailing for you, but I’m in the back of the boat, looking behind, watching you churn the water. You’re digging up stuff all over the place. You’ve got to turn around every now and then and look because I’m the one catching this stuff.’

His point was that you have to recognize these tools can present big changes, and they do make waves. You just have to be cognizant of that and look back once in a while.

Desmond is events editor for Network World and president of PDEdit, an IT publishing company in Southborough, Mass. Reach him at paul@pdedit.com.