Voting Lessons Learned

Now that the suggest a session for OpenWorld campaign has ended, it’s time to reflect on what we learned.

First off, thanks to everyone who submitted a session idea. There were 281 submitted before the deadline, and someone even figured out how to create a session after we closed the nomination form. Bonus points for ingenuity.

Now that it’s over, what worked?

Promote Early and Often
As I noted a week after we opened nominations, the UI wasn’t supporting a very fair distribution of ideas, since the only two views we offered were Latest and Greatest, leaving a gap of underserved ideas.

This is something we planned to address (seriously), but we got pulled off Mix onto Connect. So, the time ticked away, and we never got a good filtering UI deployed. This will be a must for next year, if the Events team decides to renew the program.

People adjusted though. Rather than relying on the flawed UI, many session owners promoted early and often through blogs, Twitter, the Oracle Community, and through the Mix social network (w00t! for network effects).

I watched all this promotion through Mix, Google Alerts and Summize, and I have to say I’m impressed. The Events team offered up something valuable, and people went after it with vigor.

I think one person deserves special recognition for his promotional efforts: Steve Stein from Perot Systems Consulting. Steve’s session idea made the top five even though he submitted it less than a month before the voting deadline.

Only one other idea submitted with less than a month lead time made the top 25. Steve waged a short, but effective, promotional war, leveraging his network of 531 people and even creating a blog specifically for his presentation topic.

Anyway, Steve’s example shows that promotion was key to making the top 25. All it took was 58 votes to get selected so every single vote counted.

Stay on the First Page, Above the Fold
I’ve heard a few comments that submitting too early was a bad idea (and yes, the UI is partially responsible), but I disagree. My session idea was one of the first because we needed to seed a few ideas before going live. I did no promotion, other than in the first blog post announcing the program.

And I still got within 7 votes of making the cut.

Being first in allowed my idea to get votes longer, but this wasn’t enough on its own. For quite some time, my session idea was at the top. What really hurt it overall was when it fell below the fold and off the first page of Greatest ideas, and this isn’t an issue that better filtering can completely solve.

I think most people either: a) voted specifically for an idea they saw promoted somewhere and/or b) voted for a few interesting ones on the first page, mainly above the fold. This is a variaton of the Google effect, i.e. if you’re not on the first page of Google results, you don’t exist. We tried to mitigate this by expanding the pages to 50 ideas, but that doesn’t help the below the fold problem.

Food for thought next year. Use promotion to keep these pole positions and you’re in good shape.

Be Controversial (Ahem) Interesting
I’m convinced another reason my session floated toward the top for as long as it did was because it had “porn” in the title.

But if that’s not enough to convince you that controversy works, what about “BEA Aqualogic versus Oracle Fusion Middleware shoot out“? This idea finished second, although I’m not exactly sure how it didn’t finish first by a landslide.

The timing was perfect for it too, since the BEA acquisition closed in late April, and all those newly-minted Oracle employees discovered Mix and the suggest a session program.

Don’t try to tell me this session didn’t garner 50% or more of its votes due to its potentially controversial content, using the term “shoot out” certainly helped.

I’ll be surprised if this isn’t one of the top sessions attended at OpenWorld. Kudos to Lonneke Dikmans who may be the star of OOW 2008.

Anyway, refer back to the drive by voters who casually voted for sessions on the first page, above the fold. Guess which sessions drew their attention?

Use Your Star Power
This goes hand-in-hand with promotion, and I don’t think it’s accidental that well-known, Oracle experts like Eddie Awad, Richard Foote, Chris Muir and Dan Norris made the list.

Disclaimer: I’m not that knowledgeable about all the Oracle experts out there, so please don’t flame me for not mentioning someone as an expert.

Moving on, these experts are popular for a reason, which I’m sure helped their sessions garner votes. If you read Eddie’s blog or you’ve seen Dan speak at a conference, you’re likely to vote up their sessions because they are known quantities.

Of course, on the flip-side, the open nature of the program allows people to see fresh faces present too, which is also a good thing.

Go with APEX
This logic may not work every year, but wow, did it ever work this year. Seven of the top 25 session ideas were APEX-related. I’m sure Carl is glowing right now.

What is it about APEX? Maybe Carl or someone else can clue me in by commenting. I wonder if this will be a trend going forward, i.e. product-based communities storming the voting to push their sessions to the top.

Nothing wrong with that, just interesting, and it says a lot about the APEX community and how involved they are. Kudos.

So, these are the nuggets I’ve seen. Overall, I’m pleased with the experiment.

It’s worth noting that OpenWorld 2008 will be the first (AFAIK) vendor-sponsored conference to include community-voted sessions. and by community, I mean anyone, not just attendees or special-interest groups, since Mix is an open community that anyone can join.

Let’s pause to thank Marketing for making this happen, and bonus points for not censoring the controversial topics.

This is also your chance to sound off about anything you liked, disliked thought was unfair, etc.

Find the comments.