Posts from November 2005

This is not a cliche but a great line that I heard several weeks ago that I have been using since (so it may become a cliche):

If you must forecast, do it often.

Forecasting is always a difficult proposition and loaded with risks for the person doing the forecasting.

We ask each of our companies to go through a budget process at year end (right now) and set the goals for the following year.  That is essentially a forecasting exercise (on the top line at least).

And many of our companies have incentive comp plans (equity or cash) that depend on hitting the budget or somewhere very close to plan.

So the risk of missing the forecast is real and tangible to everyone in the senior management of the company.

When a company has no revenues, this isn’t a big deal.  And when a company has $50mm in revenues and is growing at 20% per year, its not that big of a deal.

But when a company has a couple million in revenues and is trying to double or triple that (in the "launch stage" before "escape velocity"), there is a lot of risk in the budgeting process.

And you don’t want a team to miss the budget in the first half of the year and have no incentive to try for the rest of the year (like the NY Jets this year).

And then there is the issue of expense structure.  That needs to ramp in advance of revenue growth but how much should it ramp?  These are the kinds of decisions that are hard to make on an annual basis in a company in hypergrowth mode.

So a couple years ago Matt Blumberg and Jack Sinclair, CEO and CFO (now COO) of Return Path came to the Board with an interesting proposal.  They suggested that they develop an annual budget and four quarterly budgets.  And they suggested that at the end of each quarter, they develop a new quarterly budget for the next quarter and beyond, which is essentially a reforecast based on what happened in the current quarter.  The net of this was that we went to a rolling budget processs where there was a big budgeting effort at year end and a shorter one at the end of each quarter.

If you must forecast, do it often.

That approach worked great and got Return Path through several years of hyper growth, multiple acquisitions, and a changing business model and mix.  They have now moved to semi-annual rebudgeting and may get to annual this year. That’s the goal of this approach, to grow out of it.

Since then, we have tried this approach with several other portfolio companies and are encouraged by everyone’s receptiveness to it.

Forecasting is tricky business, if you must do it, do it often.

UPDATE: Matt Blumberg has posted his thoughts on this topic and his advice to entrepreneurs who want to give this approach a try is really solid.  If you are interested in how to do this well, go read Matt’s post.

David Beisel is quickly becoming one of my favorite VC bloggers because his posts are so good.

On Monday, he gave us this chart in a post he called Bipolarization of Internet Acquisitions in 2005:

I think the data is essentially correct and I agree that "undisclosed" deals are probably under $50M.

If you add undisclosed to Under $50M, you’ll get something that looks more like this:

Sorry about the lack of titles, legends,and data labels, I am lazy this morning, but they would be identical to David’s chart above.

I think there isn’t really a bi-polarization going on in the Internet acquisition market, there is something slightly different happening.

I heard some say this summer that Yahoo! "was thinking that $25M is about the right price for these web services things".  That is not a direct quote, but directionally correct.

Buyers are either picking things up before they have a business model, scale, and significant VC investment, or much later.  The middle ground (between $50M and $500M) is where the effect of VC comes into play.

If an entrepreneur chooses to raise $10M of venture capital and take significant dilution, then the price at which he can sell and make a decent return goes up.  And the price at which the VCs will want to sell goes up too.

That’s why there is this valley in the middle of these charts.

On Nov 7th, I posted about Loudspeakers, a new podcast from Raj Bala.

Some of you may have missed that post, so I am going to say it again.

Loudspeakers is a fantastic podcast. 

If you are interested in new interesting music give Loudspeakers a listen.

I listened to the fifth podcast this morning (an hour long) and it inspired this post.  You can listen to podcast#5 here.

And the opening track on podast #6, Sharon Jones doing How Long Do I Have To Wait For You? is pure soul at its best.

To play with iTunes:
1. Download iTunes (v 4.9 or higher) and install it
2. Select Advanced, Subscribe to Podcast, and enter this into the dialog:

http://feeds.feedburner.com/internmentcamp/loudspeaker

I love the "play" button that delicious automatically puts on every mp3 link in delicious.

Now you can have that feature on your blog by simply putting this line of javascript in your template:

<script type="text/javascript" src="http://del.icio.us/js/playtagger"></script>

I have done that on my blog and every mp3 link will now be "playable".

Check out Sunday’s Positively 10th Street post or Monday’s MP3 of the Week post for examples of how this works.

Thanks Dan and crew.  You rock!!!

btw – this was part of a new release delicious did last night.  Check out the new home page, the new popular page (with a "fresh only" option),and my new bookmark page.

Notice that I filed this post under VC and Technology, not Politics.

That’s because I am going to refrain from talking about the ridiculous notion that man was "designed intellgently" that many in this country seem to believe despite all the evidence to the contrary.

But this is really about startups and how they get built.

For years, I have been fascinated by the fact that many of the very best startup companies come out of "side projects". They are accidents really.  eBay, Google, and Yahoo! are all examples of this mode of starting companies.  Delicious was born this way.  One of my favorite web services, Sitemeter, started this way.  So did Vimeo.  I could go on and on, but like a Oscar speech, I need to stop. Sorry to all the "side project" startups I left off this list.

Brad and I have been seeing a lot of "one man bands" as well lately.  Companies that have been single handedly started by one person with some outsourced development.  These people see something they’d like to have and they build it.  And all of a sudden, they are in our office with a pitch deck and the need for money to turn the thing they’ve built into a company.

So I was at breakfast a couple weeks ago with Nick Denton and we got to talking about this phenomenon.  So Nick says, "it’s the evolution vs intelligent design debate".  And I about choked on my really great full english breakfast (which is not dead and is alive and well at the Coffee Shop in Union Square in NYC).

Nick is right, there are two ways to build a company.

You can design it from scratch, figuring out exactly what you want to build, getting it all down on paper, raising some money, and then building it.  And there are plenty of success stories for that way of building a company.

Or you can just find yourself doing a startup because something you started as a hobby, or to serve your own needs, just took on a life of its own and you have no choice but to evolve it into a business.

We don’t have a preference for one way or the other, but I will say that there is something particularly special about the companies that are created via the evolution approach.

They seem more "authentic", to borrow a word from David Beisel.

There is so much more that can be done with this line of thinking, but I am going to stop here. This may become a series of posts where I talk about various differences between building evolved companies versus designed companies.  Or maybe others in the VC/entrepreneur world will pick up on this "meme" and run with it, which would be awesome.

I want to thank Nick for putting this idea into my brain. I’ve enjoyed thinking about it.  I hope you do too.

In the past week, we have started to see the development of what Richard MacManus calls the "Second Coming of Content and RSS Feeds" in his excellent and "must read" post on this topic.

I have been sensing this for a while now.  It’s not one thing, it’s a bunch of things coming together to make RSS way more useful and fundamental than it was when I discovered it three years ago.

The Future of Media post, where I gave you this formula:

1 – Microchunk it – Reduce the content to its simplest form.
2 – Free it – Put it out there without walls around it or strings on it.
3 – Syndicate it – Let anyone take it and run with it.
4 – Monetize it – Put the monetization and tracking systems into the microchunk.

was the beginning of my articulation of the changing nature of RSS and the media world it is impacting.

There was another important moment for me and that was our Sessions event, where Tim O’Reilly and Brad Burnham talked about "putting a string on data".  I wrote an entire post on that concept on the Union Square Ventures blog over the past weekend that it seems that nobody has read.  I won’t try to summarize that post here, but the basic question of whether you can "put a string on data" is a really important issue relating to the future of RSS and media in general.

And then over the weekend, Ray Ozzie puts out this post talking about making RSS two-way, with something Microsoft is calling Simple Sharing Extensions (SSE). As an aside, but an important aside, Microsoft is making SSE available under a creative commons license.  Here is the spec for SSE and a FAQ in case you are interested in learning more.

I will let the techies tell me how big of a deal SSE is, but it sure appears like a big deal to me.  Making RSS two-way seems like a huge leap forward.

And Dave Winer, the father of RSS, thinks its a good thing too in his post on SSE where he says the following:

Microsoft’s new approach to synchronizing RSS
and OPML, using methods pioneered in Ozzie’s earlier work, and keeping
the "really simple" approach that’s worked so well with networked
syndication and outlining, combines the best of our two schools of
thought, and this creativity is available for everyone to use.
It’s a proud moment for me, I hope for Ray and Jack and the rest of the
people at Microsoft, and perhaps for the open development community on
the Internet.

Then yesterday afternoon, via Richard McManus’ post I linked to at the top of this post, I came across this mindblowing post by Dick Costolo, the CEO of Feedburner, in which he put all the pieces together for me.

I am not going to even attempt to paraphrase Dick’s post.  You really need to read it in its entirety, but I will say that he is introducing a concept that may well be the answer to the question about "putting a string on the data". Dick says, near the end of the post:

If we manage syndicated content at a more atomic level by attaching
“threads” to the item, we can provide tools to publishers that enable
not just the tracking of the thread, but also use the thread as a
communications line between the world of web services and the content
item. We can essentially staple rules, patterns, and meta-data to the
content in a live and “always on” way, wherever the content goes.

Thread or strings, it doesnt’ matter to me what words we use.

We are headed into a world of atomic content/data objects that are free, open, mixable, mashable, "two-way" and "always on".

And that is just huge.

View this clip on Vimeo

Buried about 7 pages back in this morning’s NY Times business section, behind stores about Woodward, Google (i guess there’s a rule that there has to be at least one google story per day), American Idol, and Lime (a new tv/web convergence play), there is a little story that is way more important than all of them combined.

Laurie Flynn writes that Tivo To Go is going to support PSPs and video iPods.

The audio iPod is way more than a listening device. It’s a place shifting device.  I have one iPod, I use to play music in the gym, on my bike, in my car, in my office, walking to work, etc.  The same will be true of the PSP and the video iPod.

All we need is a way to get content off the closed networks and on to open networks.  Tivo To Go is one way, but an important way, that this is going to happen.

And as usual, kids will lead the way.

Two weeks ago, I posted Paul Duncan’s "You Look Like An Animal" to my MP3 of the Week.

A few hours later I got an email from his record label thanking me.  I was shocked and thrilled that someone in the music business understood the value of what I am doing with this weekly feature.

We exchanged a couple emails and the guy at the record label told me:

what you guys do is absolutely wonderful  …. any label or artist who doesn’t feel that way is crazy…it’s digital  word of mouth, a short review, a kind phrase, just a simple link  sometimes…but it allows someone who might not normally hear one of  our artists to be given a little nudge in their direction…and I  can’t thank all of you guys enough 🙂

Of course, one of the first artists who figured this out was Jeff Tweedy when he put Yankee Hotel Foxtrot up on the web for free after his label wouldn’t release it.  He’s become a hero to millions of web music listeners and rightly so, because he’s an amazing musician too.

So with that intro, I am going to post a Wilco song this week.  It’s called A Magazine Called Sunset and I first heard it on the Yankee Hotel Foxtrot Demos Bootleg which I found on bit torrent last year.  This song did not make the final cut and get on the Yankee Hotel Foxtrot record.

But it did get on the Australian version of the record and is now available as the Bridge or More Like The Moon EP (six great songs) for free on Wilco’s website.

So here is A Magazine Called Sunset.

btw – I posted this one at 192kbps since its already available for free on Wilco’s website (as an ftp download or i would have simply linked to it there).