As you can see from the date of my last post, I’ve been a bit tied up. Not to worry – new content coming soon. Including some more thoughts leading up to our Glue conference. In the meantime here’s a repost of some thoughts from Eric Norlin as we near 3 weeks out from Glue.

Yes, we are “rounding the bend” — 23 days until Gluecon. I feel like I’ve droned on endlessly about how great it’s going to be (I probably have), so let me just take a different approach by highlighting some things I’m looking forward to — starting at the END of Day 2 and working backwards a bit.

David Linthicum’s closing keynote: I’ve never met David, but he has the reputation of being super smart — having been around SOA architecture stuff, and now cloud stuff for quite some time. David told me he’s more comfortable speaking to a technical crowd, and I said, “perfect.” I’m looking forward to his thoughts where we go from here.

The  ”Hacking Identity” track  – which highlights user managed access (Eve Maler), federated provisioning (Nishant Kaushik), XAuth (Chris Messina), and Webfinger (Brad Fitzpatrick) and follows it up with a discussion moderated by Ian Glazer (of Burton Group, now Gartner).

“Integrating Drizzle” with Eric Day from Rackspace. Rackspace brought most of the Drizzle guys on board when the Sun-Oracle merger happened. I’m anxious to learn more.

“On Hadoop” with Todd Lipcon from Cloudera. Hadoop is about as dominant as it gets at the moment, and I profess to knowing far less about it than I should. You?

Three sessions on scalability and the cloud stack — from Bradford Stephens (Hadoop, HBase, Zookeeper), Oren Teich (of Heroku, on scaling apps), and Sebastian Stadil (Scalr).

That’s just one possible “track” of sessions that runs from Lunch to the end of Day 2. That’s an extra-large helping of information. In one afternoon. And you can make your own choices.

I’m really proud of what we’ve got going, and when I look around at what else is out there – be it 1 day unconferences (that cost the same as gluecon), or large tradeshows (that cost 2-3x gluecon), and exceedingly happy about the value that we’re offering developers. So, get your butt to Gluecon!

[Updated with the new video link/embed]

You may not have realized this, but Niel Robertson is not only building a unique SEM marketplace, but he’s also a passionate knitter. The video below shows off both of these passions and is well worth watching.

On a more serious note, I’m a big fan of companies producing these sorts of fun but informative videos. People respond so much better to highly visual, straightforward media such as this.

Recently Dan Frommer over at the Silicon Alley Insider wrote about the reemergence of “stealth mode”. While I’d argue that it never really went away in the first place, it seems that more people are seeing the virtues of staying under the radar as they build the basics of their business (despite TechCrunch’s declaration that stealth was stupid…). After some hazing from my Foundry partners about several of the companies that I work with remaining silent about their operations (and in some cases off of our website so as not to attract attention to the fact that they’ve been funded) I wrote a post on the topic here.

With one of our portfolio companies, Trada, recently emerging from 18 months of quiet operations, we’re taking the subject up once again (very timely if Frommer’s post is any indication). In the case of Trada, staying quiet about the expert SEM marketplace they were building allowed the company the room to develop their platform in relative quiet and without tipping off any actual or potential competitor to what they were doing. To be clear, we weren’t completely secret about it. We launched with a large number of customers and an active marketplace – we were just selective about who we told what we were up to. This allowed us to both work out the kinks in the system before we put it under the load of the public eye and to be able to launch and talk about not just what we were planning on doing, but what we were actually doing (and had been for some time) for many advertisers who had been users of the company’s platform for months.

Niel Robertson, the CEO of Trada, wrote one of the best articles that I think I’ve read on the subject of stealth on the Trada blog that further describes our thought process around operating the business stealthfully for a number of months. I’d strongly suggest you take a look at it.

And with all this discussion of “stealth” I think Niel and I may have won over at least one of my partners. Brad put up a post yesterday talking about how a number of companies in the Foundry Portfolio have benefitted from keeping all or parts of what they were working on under wraps and then announcing with a bang what they were up to (AdMeld, Oblong, Zynga as well as at least one more that still hasn’t announced it’s existence).

Perhaps all this conversation leads to the conclusion that stealth isn’t the problem – it’s how companies use it.

Someone at a meeting I was in a few weeks ago made a statement to the effect that he valued management (sales management in particular) following this mantra.

I couldn’t disagree more.

While it makes for a great VC cliché it seems to me that it’s not a good plan to set an expectation with companies that you work with that you want them to essentially lie to you about the results they expect. Following this down the management line – from board to CEO to VP of Sales to Sales Manager to Salespeople – and you’ll completely cloud your view of what’s really happening in a business (where at every step of the way each person tries to set up an expectation that is lower than what they actually believe they deliver).

Instead, how about setting a realistic expectation for performance, lay out the factors that influence success and then try like hell to beat the number.

That sounds like a better plan to me.

Earlier this week, Medialets made a significant announcement: the launch of their universal mobile ad SDK. This is an initiative that the company has been working on for months and is truly groundbreaking in the industry. The Medialets universal ad SDK allows app developers to have complete flexibility in how they manage the ad inventory available through their applications. Publishers can serve Medialets rich media ads, ads from other mobile ad networks as well as ads from their own first party ad server – all from one SDK. The Medialets SDK is already up and running with major publishers (Washington Post, NPR, Variety, Fandango, etc.).

This represents not just an achievement for Medialets, but a step forward in the evolution of mobile advertising (which is much too fragmented). We’re excited that Medialets is in the middle of the mobile ad revolution.

I can’t help myself at this time of year but to remind you that you probably haven’t registered for Glue yet and that you need to get on it. Glue is one of three conferences that Foundry helps facilitate with Eric Norlin (the others are Defrag and Blur, the latter of which is still in development but coming this summer). Glue is an in depth discussion about the web as a platform and the future in a world where most (all?) apps live in the cloud (either public or private). Once again, Eric has put together a fantastic agenda (you can see the full list of topics here). This year’s speakers include:

• Michael Barrett, CISO, PayPal
• Professor Eric Brewer, creator of the CAP Theorem
• Chris Hoff, Dir. of Cloud Solutions, Cisco
• Ryan Sarver, Dir. of Platform, Twitter
• Jonathan Ellis, Lead on the Cassandra project

In addition, you’ll hear about:

• The Cassandra Database: Inside Twitter’s Choice
• What’s up with OAuth/WRAP?
• 5 Things I Hate about your API Terms of Service
• The Apache Cloud Stack (Hadoop, HBase, Zookeeper, etc)
• App PaaS vs. Enhanced Cloud System Infrastructure
• Inside MongoDB: the internals of a NoSQL database
• Understanding User-managed Access

Glue runs May 26th and 27th and is being held this year at the Omni Interlocken (in Broomfield, CO).  If you register now you still qualify for the early bird rate of $525 (regular price is $695). In addition, this year Glue is sponsoring CloudCamp the day before the conference. CloudCamp is free, but space is limited (you can sign up here).

Note to CO readers: We’re really pushing hard this year for greater local attendance. We’ve chosen to run all of our conferences here in Colorado in an effort to push the region as a leader in innovative thinking around emerging technology trends. This is a great opportunity to participate in a national conversation about key themes in technology right here in your back yard. Show up and represent!

You already know the about the state of the venture capital industry in 2009: venture investing down (32%), exits down (14%; slowest exit year for VC backed companies since 1995), fundraising down (56%), IPO’s almost non-existent (8 venture backed IPOs in 2009). It’s a bleak picture for the industry overall, even if there’s a group of us that continue to believe this is a great market in which to be investing (and it clearly is). These stats got me thinking about the future of the venture industry and I thought I’d offer up some thoughts on where we might be headed.

First, let me frame the conversation by stating that I agree with Fred Wilson’s assumption that somewhere around $15Bn is the right “steady state” investment pace for the venture industry as an asset class. At this investment level the return profile of the industry maps to a reasonable expectation of inputs and outputs (the money invested in start-ups as compared to the exit activity). By that measure, we actually still have a ways to go to reach that equilibrium in the venture markets.

According to VentureSource, $21Bn was invested by the venture capital asset class in 2009, and this amount was the lowest investment total in the 10 years of data that I had access to. The system is still a little bit out of equilibrium, however, as this is a far greater total than the amount of capital raised by venture firms in 2009 ($12Bn). In fact looking back at the past five years $14Bn more has been invested by firms than has been raised. While presumably this will lead (eventually) to fewer dollars invested, the VC fundraising average for the past 5 years has been almost $25Bn, suggesting that we still have a ways to go to get to Fred’s $15Bn bogy. 

What’s even more interesting to me is to consider the nature of this fundraising and the ramifications it has on the industry as a whole. I believe what we’re going to see in the venture industry is a bifurcation of fundraising– basically a barbell on the graph of fund sizes. Large, well known, multi-sector and multi-stage “mega-funds” will be able to raise $750MM or greater at one end of the scale, and smaller, more focused funds will raise $250MM or less on the other end – with a relatively small number of funds in the middle.

Looking at the 2009 fundraising data shows that this trend is already taking shape, three well known funds in the former category closed on over $3Bn in commitments

– NEA ($1.24Bn), Norwest ($1.2Bn) and Khosla Ventures ($800MM). At the bottom end of the scale there were numerous funds that raised money in the $25MM – $250MM range).  And while there were certainly a few funds raised in the middle (notably Greylock, Matrix, DCM, CRV and Andressen Horowitz) my hypothesis is that fundraising in this size range will diminish over time as LPs move their money either to a smaller number of diversified, extremely large funds or the larger number of smaller, more focused funds (Foundry is clearly in the latter category).

Thoughts?

While search marketing is already a huge business, more and more companies each day are discovering the advantages of advertising directly to customers through search engines. Companies like that they can directly measure the impact of their spending – from the clicks they are generating all the way through the products they are selling as a result of those site visits and that they can quickly and easily scale up their spending on what’s working in their search campaigns. With different ad groups, ad copy and landing pages, search marketers can customize their campaigns to fit their business needs.

If there’s a downside to search, however, it’s that effectively managing search campaigns is extremely difficult. Even if you confine your efforts solely to Google the complexities of creating ad groups, generating keywords, pricing each keyword, creating deep links into your product catalogue, managing spend variants by day, figuring out broad match vs. phrase match vs exact match vs negative match, etc are daunting.

One of the ironies of search is that while technology of search itself is in many ways disaggregating the relationship between marketers and consumers (and bringing them directly together), the business of search itself isn’t something that can easily be disaggregated by technology in the same way.  Search is simply not something that lends itself well to machine automation. And while there are a few software platforms available for managing search campaigns (mostly focused on the high end of the market spending > $100k/month on search) these packages are primarily designed for people who are already search experts. It’s almost impossible to take search knowledge and put it into an algorithm. As a result, companies that lack this expertise are at a huge disadvantage in the search game (this is true of many agencies as well who use search marketing as a lead-in to offer other more lucrative services).

Today we’re launching Trada. And fundamentally changing the game in search marketing.

Trada has been working in stealth mode for the last 18 months to build a system that harnesses the power of a “crowd” of search experts to work on behalf of advertisers. The Trada system easily allows advertisers to upload campaign information and connect with hundreds of search experts. It’s not a referral site – the Trada experts work together, through the Trada platform, to create the broadest possible campaign for each advertiser. These experts get paid only for generating clicks and/or conversions for Trada advertisers (depending on whether a campaign is in pay per click or pay per action mode). We work in the middle to enable these campaigns and make our margin based on our search experts’ ability to beat your pre-determined CPC or CPA rates.

The company opened its system to a small group of advertisers in January 2009 as it worked out the specifics of the platform. Trada has served over 70 customers in that time period. The average campaign in the Trada system has over 100 ads (most proposed by Trada optimizers), 6,200 keywords and an average of more than 20 optimizers working on behalf of each advertiser. If you’re working in search marketing, these numbers blow you away. Advertisers can currently run campaigns – through a single Trada interface – on Google, Yahoo and Bing.

This company is near and dear to my heart, as I’ve been with CEO Niel Robertson and the rest of the Trada team from the very start of the business (and together with them am a co-founder of the company; read Niel’s post launching the company on the Trada blog). I’ve known Niel for almost 10 years now and one of our goals with Trada has been to step away from the traditional VC/CEO relationship. We’ve done that over the last 18 months of the business and developed an unusually close partnership – the initial result of which you see today. There’s a ton more to come with Trada. Stay tuned!

Learn more about the search revolution at www.trada.com.

I have to admit that I never really loved my old blog format (either of them, actually). I never got the colors right (way too much orange, strange brown) and it was far too cluttered looking.

Thanks to the great guys over a Slice of Lime (and our IT guy Ross Carlson) I now have a great new look and feel for VC Adventure which I’m really excited about. You’ll notice a much better twitter feed (with a clever little bird “tweeting” what I have to say), a short bio on the front page, a significantly better main body look and a much easier ability to tweet, comment, etc. Welcome to 2010…

One note – during the transition to this new format (and underlying CSS – we’re now on WordPress) comments to my posts will be down. You can still comment, but the hundreds of old comments to previously written posts will take a few more days to transfer over (we need to switch over my Intense Debate

account, which Ross insists will take most of the rest of this week). Sorry about that. That also means that the “Popular Posts” feature won’t be live for a bit either (it polls posts for comments to determine the most popular articles).

John Maggio, CEO of VCIR presenting company Clementine Art, used the following slide to present his company’s financial opportunity (after I told him in a practice session that his original slide was too complicated). Created by his 6-year old daughter, it was by far my favorite slide of the conference!