Pattern Finder

Using Twitter to Do "Real" Work

Image via CrunchBase

While the media continues to talk about Twitter as a social phenomenon, I'm struck with how people are starting to use it as a replacement for longtime standbys. For example:

• Product Support: Back in June, I talked to an IT manager at a small financial services firm in the UK. They'd moved to Google Apps and were quite happy with it. As he noted, "It's refreshing to have relatively happy users." When I asked him about support, he said, "Nothing has been broken. When you're looking for information it's easy to find; if you're unsure, one Tweet will get you the info." It's been my experience that enterprises are less than thrilled with Google Apps support--it's difficult to get Google on the phone and you spend a lot of time trolling through forums. This IT manager didn't seem to have that problem, I suspect because he had a relatively small installation (eight employees). However, he also was completely sidestepping the normal support procedure. He'd moved beyond dealing with the vendor; he was quizzing the users via Twitter.
• Product Intelligence: SharePoint 2010 is currently in beta test, and enterprises are pounding it and poking it to understand it. If you go to Microsoft and ask how the beta is going, they'll give you the party line that things are going well (which I think is accurate, as far as it goes). However, that's the 50,000 foot view. What you really want to know is, "What are people doing with it, what are they trying out, where are they running into trouble?" With Twitter, that's easy to find out. Just do a search for "#SP2010." (On Google, do a search for "#sp2010  site:twitter.com"; on TweetDeck, add a column and search for "#sp2010". That tag gives you a realtime pulse on what's happening with SharePoint 2010.

A Reminder of How Bad Things Can Be in IT--And How They're Improving

Image via Wikipedia

On Tuesday, Martha Coakley, the current Massachusetts Attorney General and a college classmate of mine (Williams '75), won the Democratic nomination for U.S. Senator from Massachusetts. (For you trivia buffs, here's what she looked like--and I looked like--freshman year in college.) Given Massachusetts' strong Democratic leanings, she'll probably win the election in January, taking Ted Kennedy's place.

Her win reminded me of one of the most depressing conversations about IT I've ever had. I don't see Martha that often--maybe once every couple of years--but back around 2002 she called me up to give her advice on what to do about her IT department. At the time, Martha was Middlesex County DA, and she was trying to figure out what her options were for improving IT support for her office. She sent her car around--nice touch, I thought--and we spent an hour going over her current systems, what she was trying to do, and her budget.

While the specifics escape me now, I remember being horrified at (1) what she needed to do and (2) what she could do, given her budget. She had an IT department of one. He was doing yeoman's duty, but was seriously underpaid. It was just a disaster. I helped her with some terminology and gave her some good questions to ask--Martha would be the first to admit she's not high tech savvy--but I really couldn't help her that much. All the money was going to keep the current systems running, and there was no money left over for improving things. It was the classic case of being on a treadmill and not being able to get off.

If we had the same conversation today I think it would go a bit better. Some of the money that she had to spend for infrastructure could be re-routed to a SaaS service. Given the confidential nature of the work in a District Attorney office, I'm not sure everything could be SaaS-based, but perhaps a few things could be offloaded. Her big issue was helping her attorneys find and track information. Search technology has certainly improved, and the cost of collaboration software has gone down. For example, SharePoint is less expensive than some of the KM systems of the time.

So while we're all still struggling with how to use SaaS effectively and what IT's role should be in today's world--should it continue running the IT engine room or should it become more of a coordinator?--I think things aren't as bleak as they were back then. At the very least, there are more options. And that's a good thing.

Opening Keynote at Gilbane Conference

Following are my notes from the opening keynote at the Gilbane Conference yesterday. The panelists were: 

• Susan Parker, Director, Mass.gov, Commonwealth of Massachusetts
• Michael Edson, Director, Web and New Media Strategy, Office of the CIO, Smithsonian Institution
• Luuk de Jager, Senior Director, B2C Organizational Empowerment, Central Marketing Office Online, Philips Consumer Lifestyle

Parker: Mass.gov

Before Mass.gov, a lot of different web sites. Mass.gov came along in 2001.

Goals of Mass.gov:

• Brand: Simple, recognizable brand name
• User-Centric: Content organized according to users, not government structure
• Consistency: Same look and feel
• Consolidation: From hundreds of individual web sites to a handful of sub-portals

Mass.gov came out of an e-government task force chaired by Governor Celluci. Vision was high level. Over time has evolved into “a single face of government.”

Did an analysis of a consolidated web site; reduced the number of people needed from 20-30 to 5-6 people.

We want to deliver info to where people live, such as on Facebook. We have a long way to go to do that. We've just started on a Massachusetts Open Data initiative.

We have had a problem with being dependent on expensive consultants to keep our systems running. For products, we use Interwoven Team Site and Vignette.

Statistics:

• Mass.gov gets 2.5 million unique visitors a month; there are 200 content authors.
• 104 individual web sites in the Executive Branch before Mass.gov. In December 2009, 36 total sites.
• Before Mass.gov, 157 separate sites in all Mass government. Dec 2009 is 89; goal is 30 in June 2012.

Edson: Smithsonian

Used a process of "workshops to wiki." Time was the enemy. Used a public-facing wiki over the course of 6 weeks. To find information about the project, search for “Smithsonian Web Strategy.” Recorded the workshops on the wiki.

We’re a red ink business. We want to lose less money but involve a lot more people.

Build your APIs first; we want to have people reuse our content.

De Jager: Philips

Statistics:

• Philips has $30 billion annual revenue.
  
• 15,000 active SKUs, 52 countries

Four pillars:

• .com: The marketing site
• .shop: Used for consumer intelligence
• .live: User-generated content
• .org: Global scale, local flavor

Information Intelligence: Watching How Decisions Are Made

Image via Wikipedia

My colleague Mike Gotta recently put up a blog post entitled, “Social Analytics: Another Front For Business Intelligence.” In it he talks about Socialcast’s announcement of its Social Business Intelligence platform, and the fact that “the need for comprehensive analytics when it comes to social activities is very real.” I agree.

However, I think it gets a lot more interesting when it’s seen as part of a larger capability that I call Information Intelligence. The idea is that tracking digital content—where it’s created, where it goes, who consumes it, what it’s about—supplies enterprises with insight into how they’re creating and consuming information. We’ve done this for years with physical goods--ERP systems monitor production plans, inventory levels, shop floor status, and distribution pipelines. By understanding how physical goods move through the factory, we’re able to optimize the process. It’s about time that we did the same thing for information. After all, we are inclined to call this the “information economy.”

We’ve created islands of analysis for digital content; now we need to start building the bridges between them. We need to integrate mechanisms such as:

• Web Analytics: monitoring user behavior on web sites
• Blog Analytics: tracking posts and linkbacks on blogs
• Search Analytics: watching what users search for and the results they get
• E-mail Monitoring: analyzing e-mail flow (e.g., Gist and Xobni)
• Social Network Analysis: documenting which users link to and correspond with other users

It’s only when we start connecting these dots that we’ll start to understand how ideas flow, where the bottlenecks are, and where the experts reside. Put another way, we’ll be able to watch how a specific enterprise makes its decisions—the first step to optimizing the process. We’re a long way from having an integrated system—but we should at least start thinking about how to create one.

Adobe Buys Omniture: The Long View

Image via Wikipedia

On September 15, Adobe announced that it was buying web analytics vendor Omniture. This solves some short-term issues for both companies. Adobe’s revenue is down, and Omniture’s annual revenue of $300+ million will bolster that; Omniture wants to call higher and leverage Adobe’s many offices as a way to increase sales.

The financial analysts were all aflutter over the short term implications. Some weren’t quite sure where the synergy was (“OK, explain to me how a company that sells Dreamweaver needs web analytics.”); others felt Adobe paid too much (I’m inclined to agree, but knowing Josh James, I’m not surprised he got a good price); and others said Microsoft, Oracle, and others might top Adobe’s offer.

However, assuming the deal goes through—and I think it will—what does this mean in the long term? It means that we’re finally starting to see some market movement towards content analytics: that is, creating a feedback loop that monitors content creation and consumption as a way to help enterprises create more relevant content at a lower cost.

I discussed this concept in a Burton Group report three years ago: Content Analytics: Assessing the Value of Corporate Content (Burton Group subscription required). In the analyst call announcing the planned acquisition, Adobe said their customers were telling them that while they could monitor what parts of a web site or Flash application users were clicking on, they needed this analytics capability to be better integrated with the content creation process. Customers, Adobe, and Omniture typically discuss this integration work within the context of web site content—not surprising, since both Adobe and Omniture focus on this area.

However, there’s nothing to say that this integration work couldn’t expand to other areas, such as Adobe’s SaaS solution (Acrobat.com), digital images, documents. And in fact, in a call I had last week with Adobe and Omniture last week on the acquisition, they said as much to me. Both companies have a much broader vision of where this will ultimately go than was apparent in last week’s news coverage. This is the real long term value of this deal—expanding the creation/analytics integration work so that all digital content types are instrumented and monitored as a way to deliver the most relevant documents at the lowest cost. Put another way, in the long run, Adobe could use real data to tell you whether it’s worth spending three hours honing this project report—or whether you should spend only  20 minutes on it so you can move on to other work that will have a better payoff. That’s the path that Adobe is starting down, and that’s a good thing for enterprises.

Free Google Apps Goes Away: Why This Is Good for Enterprises

Image via CrunchBase

Google made a splash yesterday by burying the sign-up hyperlink to Google Apps Standard Edition (free) and pointing people instead to the $50/user/year version: Google Apps Premier Edition. The title of TechCrunch's article about the move ("What The Hell Happened to the Free Version of Google Apps?") hints at the general reaction. Several commenters said they felt that Google was violating its "Don't be evil" mantra with this move.

At a superficial level, this is about Google's plowing ahead with a different pricing strategy, leaving some miffed users and prospects in its wake. However, at a deeper level, this announcement signals a major organizational change at Google.

Think about it. Google has been following this dual pricing strategy for 2.5 years--it could have easily kept on a steady course. Why did it change? My guess is that upper management told the Enterprise Division that it would have to start paying its own way--it could no longer live off of ad subsidies.

In other words, this is Google saying, "Let's figure out if we have a viable business here: let's stop treating Google Apps like a lark and get serious." (A hint of the "get serious" attitude is that yesterday Google also removed the beta status from Google Apps.) If the supposition that the Enterprise Division is being told it has to stand on its own two feet is correct, that implies something else: that the Enterprise Division will be able to call its own shots. With its own money coming in, it will be able to develop the apps it needs, rather than making do with hand-me-downs (such as Google Apps) from the consumer side of the house.

In my view, the Enterprise Division has been sort of hamstrung by being a hobby at Google. (Not completely hamstrung--the Google Search Appliance, derived from Google's web search expertise, has been a big hit and a money-maker. However, Google Apps hasn't taken off with enterprises because Google hasn't made the necessary feature changes that enterprises need.) Living or dying based on paying customers will concentrate the mind of the Enterprise Division wonderfully, and that will be good for enterprises. 

The $64,000 Question: Google Docs, Acrobat.com, or Other?

A week ago, ReadWriteWeb published an interesting poll, entitled, "Poll: Which Web Office Suite Would You Pay For? Adobe or Google?" At the moment, out of 519 votes cast, 22% (114 votes) were for Adobe, 38% (199 votes) were for Google, and the greatest percentage, 40% (206 votes) were for neither. A number of commenters took the article to task for failing to point out ThinkFree and Zoho, two startups that offer productivity suites. ReadWriteWeb replied, "...we intentionally left them out so as not to dilute the vote between the two companies that are arguably industry giants as opposed to (awesome!) startups."

The problem with such open polls is you never really know the survey universe--it could include some Adobe or Google employees, hoping to tilt the results; it could include many respondents from SMBs, rather than large enterprises; and so on. Probably the more interesting part of the article is the comments, as they highlight many of the viewpoints I hear from clients. Here is a sampling:

A week ago, ReadWriteWeb published an interesting poll, entitled, "Poll: Which Web Office Suite Would You Pay For? Adobe or Google?" At the moment, out of 519 votes cast, 22% (114 votes) were for Adobe, 38% (199 votes) were for Google, and the greatest percentage, 40% (206 votes) were for neither. A number of commenters took the article to task for failing to point out ThinkFree and Zoho, two startups that offer productivity suites. ReadWriteWeb replied, "...we intentionally left them out so as not to dilute the vote between the two companies that are arguably industry giants as opposed to (awesome!) startups."

The problem with such open polls is you never really know the survey universe--it could include some Adobe or Google employees, hoping to tilt the results; it could include many respondents from SMBs, rather than large enterprises; and so on. Probably the more interesting part of the article is the comments, as they highlight many of the viewpoints I hear from clients. Here is a sampling:

The Shifting Sands of Product Archetypes

It's taken me a long time to figure this out, but I recently realized that clients often view and describe products via archetypes. Rather than recounting a long laundry list of desired features, they'll say, "What I want is something like SharePoint, but with an additional capability." Initially, I classified this common way of talking as a time-saver, but I've come to realize its also a very useful market analysis tool--because the archetypes shift over time.

Let me give some examples. Way back when--40 years ago--if you wanted a really deluxe typewriter, you dreamed of an IBM Selectric. It looked cool, it had that nifty rotating ball, and made beautiful copies. When I entered college, my mind's eye was on a Selectric--but I made do with a Smith Corona. By the late 70's, people's mental image of the Selectric was replaced with a Wang Word Processor; which in turn eventually got replaced by a PC running Microsoft Word. Another example in terms of portable entertainment is the Sony Walkman; it got usurped by the Apple iPod. If your product is the archetype, you control the market conversation--your product is the baseline by which other products are described and evaluated.

The Tsunami of Google Wave

Image via CrunchBase

Well, Google certainly made a big splash with Wave at its I/O Conference. (Backstory: Google sent me an invitation. However, since Google Apps Premier Edition has been such a wet squib, I decided not to go. Oh, well, live and learn.)

The media has gone into over-hype drive, asking the wrong questions such as, "Is this the end of e-mail?" and "Is this the downfall of SharePoint?" (Uh, the short answer is "No" and "No.") These are the wrong questions because they look at this announcement through a technical lens: will this technology replace that technology? The better question to ask is, "How will this affect business viewpoints and processes?" And from that vantage point, Google Wave is a significant announcement.

It's significant because Google Wave makes physical a groundswell towards contextual collaboration that has been building for some time. There are a bunch of communications technologies (e.g., e-mail, IM), collaboration technologies (e.g., social networking, workspaces), and content technologies (e.g., productivity suites, document sharing) that the market has always perceived as separate. Google Wave takes them, munges them together, and allow workers to perform both synchronous and asynchronous collaboration within a work context. Rather than making the worker open an e-mail application and a word processing application and an IM application, the worker just works. The three Cs come to him, rather than him going to them. The impact of Wave is that it's an instantiation of a vision that many people--including Burton Group--have been talking about for a long time.

In the real world, what that means is enterprises will start to say to vendors, "I want a Wave-like product." Google has given them a new product category to ask for. So give Google an A+ for vision and market impact. However, they'll probably earn a much lower grade for delivering on the promise for large enterprises. There's a bunch of dynamics at work here.

A Good Idea: Paying for What You Use

"Paying for what you use." Yes, I know, it's a radical concept in the software world, but not in other areas. Electricity: the utility company measures your electricity consumption, and bills you for it. If you consume a lot of electricity at peak hours, then you pay a lot more than someone who uses little electricity at off-peak hours. Gasoline: if you drive a lot, you buy a lot of gasoline; if you drive a little, you buy little gasoline.

In the software world, equitable payment for what you consume is much more elusive. The definition of consumption is all over the place. If you're talking servers, the measurement might include the server model number ("size"), the number of CPUs, the number of potential users, the number of concurrent users, etc. If you're talking PCs, the measurement is usually a user, but with little regard to their usage of the software. An expert who uses every feature in the software is often billed the same as someone who buys the software but never uses it. True, there is some rudimentary granularity in the market--for example, the business intelligence market has often offered two types of client software: an expensive rich client package for the analyst who drills into the numbers and creates reports, and a less expensive client (usually web browser-based) for the casual user who reads canned reports from time to time.

However, in this recession, that rudimentary granularity is no longer enough. I'm getting a lot of calls from clients who feel they're paying too much for Microsoft Office and want to explore other options. It isn't that they feel they're being overcharged across the board--they think Office is an excellent suite for their expert users. For example, Microsoft Excel is incredibly feature-rich, and their spreadsheet jockeys can do incredible things with it. At the same time, these same companies have given Excel to workers who just use it as an application for generating simple tables. The most spreadsheet-specific thing they do is sometimes add up a column--and you can do that in Word.

So when someone like Google or IBM or Sun comes along and says, "We can give you a pretty good productivity suite for a much lower cost [including free]," these budget-strapped companies take notice. Now due to a variety of issues--the need to support expert users, some file translation issues, etc.--it's not easy to do a wholesale replacement of Office. However, surgical replacement can be done.

In my view, both sides of this argument are wrong: Microsoft saying that you need to overbuy because you never know when a worker might need a certain feature (true, but not as often as Microsoft claims); Google, IBM, and Sun saying that you don't need all that functionality (actually, sometimes you do). In other words, both sides assume that rudimentary granularity is here to stay.

The way out of this conumdrum is to say, "No, we're going to charge customers for what they use." With a SaaS-based solution, that's quite easy to do--the web server tracks feature usage and charges back in a much more granular fashion. For example, one user is typically a casual user, but because she's been assigned to a three-month project requiring heavy usage of a spreadsheet, she's charged for heavy formula usage. After that, after she goes back to using the spreadsheet as a glorified table creator, she's charged the lower rate again.

Even with software, this is possible. Microsoft has been monitoring usage stats of Office for years, as a way to understand feature usage. However, it's part of the development process, and has never been used for licensing.

So the next time you sit down with your solution provider to talk about licensing, start the conversation by saying, "You know, we'd actually like to pay for what we use." If the vendors hear that enough from customers, they may start rethinking how they do licensing. Electricity and gas have been sold in an equitable fashion for the past 80 years; it would be nice if software caught up. 

Ernie Roberts--A Blogger Before There Were Blogs

Cover via Amazon

An uncle of mine, Ernie Roberts, died on Monday. It's not a name that means much to most people--unless you're a sports aficionado in the Boston area. Then it means a lot. He worked at The Boston Globe newspaper from 1947 until 1983, working his way up in the Sports department hierarchy--from evening sports editor to sports editor to executive sports editor. He was instrumental in making The Boston Globe's Sports section the powerhouse that it was in the 1970's and 1980's. The story in the Globe notes, "He's probably more responsible for making the Globe sports section great than any other individual," said Globe sports columnist Dan Shaughnessy.

Most readers of the Globe didn't know him as an architect of the Sports section's writing dynasty. Instead, to them he was just another Globe columnist--but with a gimmick. Because Ernie always started his column talking about breakfast--what he had that morning, or what some sport celebrity's favorite breakfast was. After I moved to the Boston area in 1975 I'd read his column from time to time and I can still remember being struck at how casual it all was. He'd start with the breakfast thing, and then talk about this, and then talk about that. It was very casual and conversational--back in a more formal time when most business people wore suits and ties to work and the term "business casual" hadn't been invented.

Now that I think back on it, Ernie was ahead of his time--he was writing a blog before blogs were invented. He instinctively realized that people hungered for a quick, casual conversation--even while he labored at a newspaper, the classic broadcast medium--and he gave it to them. In other words, Ernie and others popularized a writing style that finally found an outlet when blogs came along--a quirky, quick commentary on events of the day.

I'll close with an Ernie story that's probably more of a commentary on me, rather than my uncle. Ernie called me one afternoon in 1978 when I was working at John Hancock in Boston. Excitement was high at the time because the Red Sox were in a tight race with the New York Yankees for the American League East. Ernie had some extra tickets for a Boston/New York game at Fenway and had called to ask if I wanted them. (At the time I lived in Salem, MA and took the commuter rail back and forth to work.) I did a quick calculation, figured that the game would get out late and that I'd be taking a midnight train back to Salem, and declined them. When I casually mentioned to my underwriting colleagues that the Executive Sports Editor of The Boston Globe had offered me free tickets on the third base line to a Red Sox/NY Yankees game--and that I'd declined them--I was almost lynched on the spot.

Saving Money in IT: Granularity Is Key

Saving money in IT is certainly on everyone's mind these days, and one of the keys to doing so is using "granularity."

Huh? The idea is that for a long time, IT has not been very granular in its approach to things. Purchasing over capacity has been a way of life. The mainframe is a perfect example of this: this was a huge box, much of which sat idle until the enterprise had written a bunch of programs for it. Things got a bit better once minicomputers and PCs came along: you could size the box to its tasks. But even that wasn't quite right: if you bought a new application, it usually required a new server. So you had a bunch of computers, all of which had some excess capacity within them. If you could add all of that excess capacity together--and use it--then you probably could coast for a couple of years without having to buy a new computer.

The thing that has added granularity to hardware is virtualization. Now you can split a physical server into multiple virtual machines--and expand or contract the size of those machines as needs dictate. This ability to make a machine just the right size helps explain virtualization's current popularity. It comes down to granularity.

When it comes to software, we've once again typically taken the "big box" approach--or, to be more accurate, the "big list of features" approach. Somebody needs a productivity suite? Give them Microsoft Office. It does everything and then some. In many cases, it does more than the employee needs; but there's no easy way to get Microsoft Office lite from Microsoft. Some enterprises fix that by not buying Microsoft Office for casual users.

However, that is starting to change. Google, OpenOffice.org, IBM Lotus, Sun, and others are offering productivity suites with less functionality than Office--but also making the argument that that's OK because they cost less. Their wins are typically not in displacing Office--but rather in giving productivity suites to users who had none. Once again, it comes down to granularity.

And that makes sense--one of the ways you save money is by buying exactly what you need and nothing more. In the past, due to technology constraints, IT has had to overbuy--but those days are waning. State-of-the-art IT shops are focusing on granularity--whether it's in the form of virtualization or segmenting their employees--and are saving money while at the same time supporting the business.

Note: cross-posted in the Burton Group CCS blog