Posts from August 2011

Don't Let Good Friends Hear About It On TechCrunch

A few weeks ago, we were getting ready to close on an investment. I asked one of my partners when the news of the financing would be announced. He said, "pretty much right away." And I said, "we had better let our friends who run XYZ company know." My partner said, "absolutely, can you compose an email we can send together?"

This has been our mantra for a while now. We have a lot of friends in the startup world, entrepreneurs, VCs, management team members, and it is quite often that we are involved in a transaction, a financing, a sale, a merger, that might impact them and their interests. We don't want them to read about it on TechCrunch. We want them to hear about it from us first.

You'd be surprised how much goodwill this practice creates. Nobody is happy hearing that you just funded a competitor of theirs, or something similar, but when they get an advanced warning from you along with an explanation of your thinking behind the move, it goes a long way. They aren't caught off guard and they can process the information calmly. It doesn't mean they will be happy about the news. But it does mean that your relationship with them will be preserved (in most cases).

Doing this is tricky. Investors have an obligation to keep information they have confidential until the subject company decides to disclose/announce. So you have to time the tipoff of your friends carefully. We typically wait until the day of the announcement or at the earliest the evening before. In many cases we will clear the tipoff with the subject company as well. A lot depends on the people and relationships involved on all sides.

Relationships and reputation are the currency of business. I do not believe you can be successful in business without both. So preserving them is critical. And one good way to do that is to make sure good friends don't hear about things that are important to them on TechCrunch.

#VC & Technology

Something's Happening Here

In the past week, the daily visitors at AVC have gone from an average of 10k per day to over 20k per day.

AVC visitors

While this influx of new visitors is welcome (welcome everyone!), it is also unnerving. Where is it coming from?

Well the answer to that is pretty obvious – StumbleUpon. Here are the top traffic sources to AVC for the past week.

AVC traffic sources past seven days

Typically direct traffic brings 40k visits per week. Readers clicking through from the RSS feed make up about 10k visits per week. Google search and Twitter each bring about 7,500 visits per week. StumbleUpon had been next on the list, bringing around 5k visits per week over the past few months. I have noticed StumbleUpon traffic growing pretty consistently over the course of this year.

But I've never seen a firehose of traffic and visitors like I've seen this past week from StumbleUpon. I'm not entirely clear what is happening here. But I hope it continues. More readers means more conversations and opinions and more friends. Welcome Stumblers.


Pricing A Follow-On Venture Investment

Today on MBA Mondays, I am going to walk you through some math that our team does when looking at a venture investment in a company that is starting to scale its business.

Let's assume we have a portfolio company. I will call it It is a marketplace for fitness experiences. We invested in it last year as it was getting ready to launch. A year later the business is scaling nicely and needs more expansion capital. The founders don't really want to go out and do a fundraising process. So they have asked the existing investors to make them an offer for an internal round. They believe they need $3mm of expansion capital to get them to cash flow breakeven.

So now the VC firm (us) needs to figure out what is a fair price. So we pull out Google Docs and run some numbers. For those who didn't click on the link and see the spreadsheet, here are the numbers:

– is on track to generate $10mm in gross transactions in 2011

– they operate on an all-in "take rate" of 9% so their net revenue in 2011 will be $900k

– they will have operating costs in 2011 of $1.5mm and they will lose $600k this year

– they plan to triple gross transactions in 2012 to $30mm and grow to $150mm in gross transactions by 2014

– they plan to do this while ramping operating costs to $3mm in 2012 and to $7mm in 2014

We lay all of those number out in a spreadsheet and then look for some multiples to apply to them to get to a sense of value. The two multiples I like to use for marketplace businesses are enterprise value/gross marketplace transactions and enterprise value/EBITDA. And the multiples I like to use are 1x gross marketplace transactions and 20x EBTIDA. These are for internet marketplaces that are growing fast and are category leaders.

I've observed these multiples over a long time, going back to eBay and Mercado Libre a decade or more ago. We keep a spreadsheet of all Internet marketplace financing transactions in our portfolio and also include transactions we are very familiar with. That spreadsheet validates these multiples again and again.

When using multiples, one question that comes up is "do we apply these multiples to the current year results (which are almost in the bag), the current run rate (current month X 12), or next year's forecast. My answer to this quesion is "yes." I like to apply these multiples to all three and then triangulate from there. The reason being that when markets are frothy, investors will often give a company valuation credit for the next year's forecast (meaning a forward multiple). But when markets are tough, the multiple will be on the last twelve months (meaning a trailing multiple). You don't know what kind of market you will find yourself in so you should look at the multiples in a number of ways and triangulate to get to a comfort zone.

We did this in our spreadsheet (just for the current year and the next year) for our two multiples (1xGross and 20xEBITDA) and we got to a range of valuations for 2011, 2012, 2013, and 2014. They are:

2011: $10mm to $30mm (midpoint $20mm)

2012: $30mm to $75mm(midpoint $52.5mm)

2013: $35mm to $150mm (midpoint $92.5mm)

2014: $130mm to $150mm (midpoint $140mm)

So that's how we think about valuation. The spreadsheet says that if the Company hits plan, it will grow from a valuation of $20mm now to a valuation of $140mm in three years. And if we invest at that valuation of $20mm, we stand to make 7x on our investment in three years if the Company hits plan. If we pay at the top of the valuation range right now ($30mm) and get out at the top of the valuation range in 2014 ($150mm), we stand to make 5x.

I believe 5x to 7x is a good return objective on a follow-on investment in a venture stage company that is scaling nicely. We look to make 10x on our initial investment but cut our return objectives back as the risk comes out of the investment. There is still a tremendous amount of risk in a follow-on investment of this stage (mostly related to executing, hitting plan, etc) and a big multiple is still appropriate.

So that's pretty much all that is involved. We talk this over with our entire team and decide what to offer and what our walk away price is. Based on this analysis, I believe our offer would be around $25mm pre-money, $28mm post-money. We might go up a couple million to get the round done but I think $30mm post-money would be as high as we would go. At that point, we would encourage the founders to go out and find new investors to price and lead the round.

#MBA Mondays

Google+ Is An Identity Service

Andy Carvin blogged some interesting comments by Eric Schmidt at the Edinburgh TV Festival yesterday:

He (Eric) replied by saying that G+ was build primarily as an identity service, so fundamentally, it depends on people using their real names if they're going to build future products that leverage that information. 

I blogged about identity the other day and outlined my view that every social service is hosting a part of your identity online. I listed almost ten social services that host a piece of my online identity in that post.

So I agree with Eric that Google+ is in part an identity service. But "primarily an identity service?" That's notable. Particularly in the context of what he said after that. "if they're going to build future products that leverage that information."

It begs the question of whom Google built this service for? You or them. And the answer to why you need to use your real name in the service is because they need you to. Well at least we got that out there and can deal with it.


The Amazon Tablet

I saw this picture at BGR yesterday and immediately thought "I want that."

We go back and forth in our family about Kindle vs iPad for reading. The Kindle is lighter and smaller. The iPad has a browser and other services for enhanced reading/researching as I blogged about recently.

What we all want is a hybrid of the two – a kindle that is a full blown tablet computer with a browser, apps, and an OS. It looks like Amazon is going to bring that to market this fall. I'm getting one for myself and one for the Gotham Gal. And I'm pretty sure my mom and dad are getting them too. It looks like a killer product.


Jig: What Do You Need?

There are entrepreneurs we love to work with. Joshua Schachter is on that list. When he told us last year he was starting a new company, Tasty Labs, we said "we're in." It wasn't exactly clear what Joshua wanted to build, but we knew it was in the general area of a marketplace for things people need.

Joshua put together a killer team and they started coding stuff up. A product emerged. We've been using it in alpha and beta for a while. It's evolved. And sometime in the past few days, Joshua took the covers off. It is called Jig and it is indeed a marketplace for things people need.

Go take a look. Tell us what you need. Get help with those needs. Pretty simple.

I've got a need posted on Jig. Maybe you can help me with it.

Or maybe you've got a need. Post it and maybe you'll get that need resolved.

That's Jig. I like it. I hope you do too.

#VC & Technology#Web/Tech

Users First, Brands Second

I like simple ways to think about things. Like mobile first, web second. These kinds of constructs work for me. One that I've been using a lot lately to describe what we like and don't like as much is User First, Brands Second.

When you are building your product and thinking about your go to market strategy, you need to decide who your first users will be and how they will take your product into the market. You can focus on getting everyday internet users first. Or you can focus on getting brands first and working with them to get users. This decision is critical and will impact almost everything about your business going forward. So don't make this decision lightly.

There are some great businesses that have gone with the Brands First, Users Second approach. Two that come to mind are Groupon and Buddy Media. They are very different businesses but both go first to brands, work with them to craft offerings for users, and then work to get those offerings in the hands of users.

Contrast that with Foursquare. The Foursquare app launched without any brands on it and went after user adoption directly. Today they have over 10mm registered users. Millions of users engaged on the Foursquare service attracts brands. And the brands come to Foursquare without much effort on Foursquare's part. This is the User First, Brand Second approach.

Facebook, Twitter, and Tumblr are all User First, Brands Second services. The brands are all over these services now. But for the most part, these services didn't do much to bring them. The engaged users did.

Our firm vastly prefers the Users First, Brands Second approach. I don't want to say definitively that we would not invest in a Brands First, Users Second business, but it would have to be something very interesting to get us to do it.

We do have several portfolio companies that took some version of a Brands First, Users Second approach to market. WorkMarket brings large employers and staffing firms onto their platform who put work orders into their system. That brings workers who pick up the work and get paid via the Workmarket platform. SoundCloud initially targeted audio content creators who put their work on the SoundCloud platform and according to Quantcast, that audio content brings over 14mm people a month to the service. But in both cases, these services launched day one with a valuable offering to both brands and users and they were not crafted specifically for the brand's benefit. That is a key point to be focused on.

The biggest problem with a Brands First, Users Second approach is you can get caught up in product development efforts to satisfy the brands and as a result you can't put enough energy into satisfying the users. And if that happens too much, you end up servicing the needs of the brands over the needs of the users and then you are a service business not a platform.

So when thinking about whether or not your company is a good fit for investment by our firm, think about whether your business is User First, Brand Second, or the other way around. That will be one of the first things we focus on when we evaluate it.

#VC & Technology#Web/Tech

Fifty For Fifty!

Last night we passed our goal of raising $50,000 for teacher's projects that bring families closer to the classroom. The Gotham Gal and I are delighted. This has been an amazing gift that all of you have given us. Almost 10,000 students and their families will benefit from your generosity.

This is the fifth monthly campaign that this blog community has done via Donors Choose. In past years, we've raised about $20,000 each year from about 200 donors. This year, we raised more than $50,000 from 314 donors (and counting). That's a big step up!

We will leave the Giving Page and blog widget up for the remainder of the month. Hopefully we can raise a bit more for classrooms. At the end of the month, we will wrap this up, I will write a final blog post with the stats on the campaign, and I will list all the donors and link to their twitter handles if we have them.

On November 9th, from 6-8pm in the USV Event Space, the Gotham Gal and I will host a meetup for all the people who donated in this campaign plus some of the wonderful people from Donors Choose who make it possible to do this. We will get a Meetup page up shortly so people can RSVP for that event. It will be limited to those who gave in this campaign.

I talked to Charles Best, the founder and CEO of Donors Choose, about this campaign last week. He said that there are only a few other communities in the world that can rally around a Donors Choose campaign like this. This is a special place and you are special people. I am so gratified to be the "bartender" of this joint.

#hacking education

Identity, Authentication, and Provisioning Them Online

Christina jotted down some thoughts on indentity on a flight to SF and I read them this morning. In her post, she references Ev's excellent post on the same topic from a while back. So I went on a bike ride as the sun rose over the east end of long island and thought a bit about all of this.

Before going on, I'd like to emphasize that these thoughts are mine and mine only. Nobody has seen this post before publishing other than me, including my partners and our portfolio companies. It does not represent the opinions of any company I and/or our firm are involved in.

I don't have a single online identity. I have many. They are rich, representative, and different from each other.

@fredwilson (tumblr)

And many many more.

I apologize to all the services out there (in and out of our portfolio) that I left off this list.

I believe that OpenID is on the right track. In the OpenID scheme:

The term OpenID may also refer to an ID as specified in the OpenID standard; these IDs take the form of a unique URL, and are managed by some 'OpenID provider' that handles authentication.

OpenID has two important concepts in it. The first is identity. The second is authentication. The two are totally different but they have become comingled on the web because the leading third party authentication services, Facebook, Twitter, and Google, are combining the two in interesting ways.

When you build a web and/or mobile app and you want to make it easy for the user to share data between your app and one of these big three web services, you provide for one button authentication to them. Everyone who uses the web and mobile apps is now familiar with "login with Facebook", "login with Twitter" or "login with Google". We use them all the time. They make things easier on us.

These authentication services provide some notion of identity as well. But only your identity in their service. Not your entire identity.

So back to OpenID for a minute. I really like the idea that a URL can be an ID. But I don't like the idea that one URL is your ID. I like the idea that a list of URLs makes up your ID. I started my list at the beginning of this post. It is not complete by any means, but it is a good start.

So what I want is a layer that sits on top of all these services, aggregates up all of my URLs (identities), and then provides authentication in the same way that Facebook, Twitter, and Google do today.

And I'd like this layer to be able to provision to web services exactly the same data that you can get (and give) by authenticating directly with the social platforms. And, of course, I'd like to control what data gets provisioned to what apps.

Many have taken a stab at this over the past few years. It is a big opportunity and a big problem. But none (including OpenID) have gotten the kind of traction that Facebook, Twitter, and Google have. I believe there are several reasons for that. First, you need a brand that users recognize (and trust??) to be able to do this. Second, the authentication experience needs to be simple, easy, and not geeky in the least. And third, you need the cooperation of Facebook, Twitter, and Google to do this well and it is in their interests to be the providers of authentication and identity on the Internet so getting that cooperation has been tough.

The good news is it is becoming increasingly clear that no one web service will control our identity online. The success of Google+, Tumblr, Foursquare, Instagram, etc, etc in the past year has shown that users want more social platforms in their lives, not less. Or at least that some users want different social platforms than the ones that have been leading the way for the past decade.

So maybe the big three can get together and cooperate on building this authentication layer on top of their services and promoting is as an indepedent way to authenticate and provision identity and related data to web and mobile services. I'd love to see that happen and I suspect the Internet would be a better place because of it.


Financing Options: Working Capital Financing

We are coming to the end of the Financing Options series. This is the final post in the series. Today we are going to talk about working capital financing.

For those of you not steeped in finance and accounting matters, I suggest you go back and read the Balance Sheet post before reading on. Working Capital Financing relies on a company's balance sheet to support the loan so understanding how a balance sheet works is important to understanding working capital financing.

As a company grows, it starts to consume a lot of cash in the day to day operations of the business that has nothing to do with its profits or losses. This type of cash consumption is called working capital. In accounting terms, working capital is equal to current assets minus current liabilities. In layman's terms, working capital is what your customers owe you plus any inventory you have built up minus what you owe your suppliers and employees. Working capital also includes any cash you have in the bank.

One of the many awesome things about a software business is that it rarely has any inventory. But for the purposes of this post, we need to think about a business that has inventory because inventory buildup is a big reason that companies consume working capital.

Let's think about a company that makes iPad stands like this one (I have it, it's awesome). Let's say it costs $25 to manufacture one iPad stand. Let's say you have orders for 10,000 of them at a wholesale price of $40. So you need to  come up with $250,000 to produce the inventory to meet the demand. Then you ship the iPad stands to Amazon or some other retailer. And then you wait 60 to 90 days to get paid the $400,000 by that retailer.

On paper, your business looks great. You have revenues of $400,000 and costs of $250,000. You have profits of $150,000. But you cash situation is horrible. You are out $250,000 and you are going to wait 60 to 90 days to get the $400,000 from retail. And you've got another order but this time it is for 20,000 units. You need to come up with $500,000 to meet demand.

This is known as a working capital issue. The business is making plenty of money on paper but can't manage its cash needs. And the faster it grows, the worse it gets.

This is exactly the situation working capital financing was designed to deal with. Banks and finance companies will loan companies, particularly profitable companies, the money they need to purchase inventory and wait to get paid by their customers. Banks will rely on the purchase orders on hand and the actual value of the inventory that the company has in stock to backup the loan. They will also take into account the money the company owes its suppliers and employees in determining exactly how much capital to loan the company.

Most working capital financing has built in cushions. Banks will not loan 100 cents on the dollar of working capital. They might loan 75% or 50%. But as working capital grows, they will increase the size of the loans they make. These are all short term loans because the inventory eventually gets sold and the customers eventually pay. A typical way these loans are structured are lines of credit and revolvers meaning that as the money comes back into the business, the loans get repaid, but the total amount available under the loan stays the same so the company can just borrow it back when it needs the money again.

For companies that are particularly shaky, there is a technique known as "factoring" where the bank actually takes the amounts of money due from the customers as collateral and gets paid directly by the customers and then remits the extra amounts to the company. The bank essentially becomes the accounts recievable department of the company. Back in the dark days in the aftermath of the crash of the internet bubble, I got a bank to do this for one of our portfolio companies and it was the only way we got through a major financial crisis.

Even a software based business can build up a lot of working capital. It ususally results from the company having to pay its obligations much faster than its customers are paying the company. If you have customers that pay in 90 days and you are growing revenues quickly, then you can find yourself in a major cash squeeze. Working capital financing is a great way to manage that kind of cash squeeze.

#MBA Mondays