Fumble! What not to do at a cloud computing conference

Ah, conferences. Attending IT events is one of those traditional fall pastimes in North America and Europe. Maybe because I spend lots of time going to them, I get very particular about how they are run, especially when it comes to what speakers and organizers deliver.

The recent flurry of cloud computing conferences got me thinking that these events have their own special challenges. Since cloud computing is at the top of Gartner’s hype curve (and top of their strategic technologies list for 2010), it’s a topic that a big chunk of the IT world is looking for information about. However, recent conferences on the topic have really fumbled what could have been a much better learning experience for everyone involved.

So, instead of spending a post complaining about SYS-CON and this week's Cloud Computing Expo in Santa Clara (or other recent cloud events for that matter), I figure I'd come up with a list for organizers, speakers, and attendees alike to keep the ball moving forward on the industry conversation about cloud computing. Here are my thoughts, along with a football-season-appropriate commentary (that'll be American football, by the way) about the impact of each of these issues:

Don’t host a panel of vendors and then lob the question “So, what’s your definition of cloud computing?” If you want to have this discussion, set up a Cloud 101 or Definitions track. But be warned: asking this question doesn’t do anything except eat up the clock. If you think you’re helping clarify the topic for the audience, think again. No one is going to agree, and the people onstage are going to try to redefine it to match their biases.

A better conversation starter? "Explain how you helped a customer take a first step toward cloud computing." Or maybe: "What underlying assumptions does your product/solution/service make that customers need to know about?" Or even: "What has been the hardest thing about helping customers move to cloud computing?" But, please, don't make us sit through another round of cloud definition roulette.

[Impact: similar annoyance factor as seeing the same Bud Light commercial for the 12th time. Or hearing the ref say the phrase, "After further review, the play stands," after 20 minutes of unnecessary deliberation.]

Don’t retread well-trodden ground, and at least point us to what's new. With cloud computing evolving so fast, it's especially worthy to note that each conference happens at a different point in time and should have something to add in to the industry understanding of where things are now and where they are going. I felt Agatha Poon of Yankee Group had a couple interesting stats about vertical adoption of cloud computing (healthcare liked it, manufacturing was worried about technology maturity levels, and financial services was very wary of the hype). In general, however, she didn't cover much new ground. I bet she lost some of the people in the audience that she hoped to engage. The EMC speaker, too, missed the opportunity to mention the Vblock, Acadia, and VCE news -- something the audience was probably a bit curious about given the story had just broken.

[This is pretty much equivalent to hearing the ref say, "Offsetting penalties, repeat first down." Do-overs can be a bit of a drag.]

Find the customers. Encourage them to come and speak. I think the whole keynote room at Cloud Computing Expo felt embarrassed by having so few hands pop up whenever a speaker asked, "How many of you are end users?" This is the most glaring item that's been missing at cloud computing events I've been to: real-world stories. At this week's event, EMC's Mike Feinberg asked, "Who has actually developed on a cloud platform?" I saw 2 hands go up. "Who has an account on a cloud computing platform?" A few more hands went up, but not many. Yikes. By the end of the conference, speakers stopped asking the question, knowing they were speaking to a room full of vendors.

newScale's Scott Hammond gave a lively pitch (as he is so good at doing) about what one of their customers was experiencing. He recounted how the customer needed to "blow up that crazy, ugly diagram of how things are done today." What would have been even better? Having the customer there to say that themselves.

Why are end users so timid to show themselves and discuss cloud computing? Maybe it's too early. Maybe they're busy. Maybe we didn't ask in a compelling way to make it worth their time. Or maybe having end users attend is not a necessary part of the conference organizer's business model. Other upcoming conferences will likely do much better on this front than SYS-CON did.

[I'd call this a long throw on third down, dropped by the star receiver, just past the first-down marker. Could have been great...but wasn't. Very unsatisfying. Time to punt.]

When you do have customer speakers, make sure they are willing to be specific about their experiences. It's almost worse to think you're going to get a bunch of useful, in-the-trenches information from a customer speaker, only to listen to a bland summary of the benefits of cloud computing. Jill Tummler Singer, the deputy CIO of the CIA, spoke at Cloud Computing Expo in nothing but generalities until the Q&A session at the end of her talk. Now maybe that's because everything that the CIA is doing with their IT infrastructure is classified, but I doubt it.

The other customer general session speaker I heard was Tony Langenstein from Iowa Health Systems. He recounted the urgent chaos around the June 2008 floods in Iowa that put an IHS data center out of commission and what he and his team did to deal with it. It was a great disaster recovery story, complete with hour-by-hour accounts, descriptions of water gushing through every nook and cranny of his data center, and before-and-after photos. The only problem? When it came to discussing the cloud computing angle (related to his use of storage) as opposed to just the data center DR issues, he wasn't specific at all. All he really said was that it "just worked." Entertaining (especially the "after" photos of fish in odd places), but not as useful as it could have been.

[This is important to get right. Not doing so is kind of like using your first-round draft pick to grab a potential franchise quarterback. And then benching him in favor of the journeyman back-up QB.]

Expert speakers can be from vendors. Just be sure to clearly delineate sales pitches from analytical, use case, or strategic content. Make sure vendors know why they are speaking -- and how to keep their selling instincts bottled up. If you're a vendor creating a pitch, it's easy to tell the difference: if you have slides with product names, it's a sales pitch. If there are slides about what customers are doing, it's probably a pretty good start.

Since a good chunk (if not all) of the presenters at Cloud Computing Expo paid for their slots, it's natural that you get a tendency toward product sales pitches. Unfortunately, that's not what the cloud computing discussion needs right now. While SYS-CON was probably able to extract some good revenue from the conference, pervasive sales pitches make it a lot harder for attendees to find the good content.

Nevertheless, there were a few very engaging, intriguing, and useful sessions that I attended (and probably several more that I missed). Yahoo!'s Dr. Raghu Ramakrishnan led what I heard Jake Kaldenbaugh (@Jakewk) call a "large-scale cloud data management master class" in his Tuesday general session. Dr. Ramakrishnan showed what putting data front and center actually means in the cloud. I'm betting Surendra Reddy's presentation (@sureddy from Yahoo!) was also pretty refreshing (thought I missed that one). I thought CA's own Stephen Elliot weaved in quite a bit of useful and relevant experiences on both virtualization and cloud. John Treadway (@cloudbzz) from Unisys and Peter Nickolov from 3Tera both did a very good job of directing business-level discussions without aggressively pitching their wares (even when pushed to do so).

[The vendor sales pitch thing reminds me of the annoyance of sitting through TV timeouts -- while you are in the stadium where the game's being played. You can't speed it up and it's exasperating to sit through.]

Vendor demos can be useful in showing what is meant. There's definitely a place at conferences for vendor-specific content, including demos. When there's someone taking a new approach or delivering a new offering, an overview of a vendor's solution is actually interesting and useful.

Cloud Computing Expo had a few like this that were worth attending. AppZero did a thrill-a-minute demo (that's sarcasm) of its Virtual Application Appliance technology that explained nicely their potential value. (To be fair, my sarcasm issn't aimed at what AppZero actually showed; it worked as advertised. It's just that, as I learned at Cassatt and BEA, infrastructure demos are tough to make exciting.) CloudSwitch explained (and quasi-demoed) a tricky bit of software indirection: their way of using a public cloud without letting your systems know that they are actually using a public cloud.

[This is kind of like a flea-flicker play. Seeing a good demo of new stuff can be pretty snazzy -- and can go for big gains. If it doesn't work, though, it does seems kind of foolish.]

Enough pontificating already. Conferences should help (somehow) capture practical advice. I heard some of the (few) customers who were attending ask a couple types of questions. First, usually of particular vendors: how does something they had been describing actually work. Sure, it's a product question, but it shows an interest in getting to specifics. The other question I heard was more general: OK, so how do I pitch these approaches we're hearing about to my organization back home? Or, (even more pointedly) assuming I buy into all this cloud computing stuff, where do I start?

[This is like the importance of having a good running game when it's fourth and goal. You want to be able to get the job done in a way that's simple, straightforward, and effective. Done right, this (and said conference) is a touchdown. Done badly, everything else is kind of pointless.]

Extracting value

These last few practical questions by end users are what can make a few days out of the office worth it. Few conferences do that. All said, and despite as negative as I sound here, the Cloud Computing Expo event was better than I expected. I even met a few Twitter followers/followees. In addition, I missed the Cloud Camp that was held the final evening -- certainly a totally different style of event that might even address some of the comments above.

In any case, extracting value from cloud computing events shouldn't be so hard. And if we're going to make progress on cloud computing, we've got to do better. Or at least find a better coach.

Note: For my "as-it-happened" tweets and commentary on Cloud Computing Expo in Santa Clara, take a look back at www.twitter.com/jayfry3 from Mon., Nov. 2 through Wed., Nov. 4.