The "Patzer Problem": Outcomes & Expectations

There's some buzz going round right now precipitated by a guest post on Techcrunch this weekend by Tod Sacerdoti of Brightroll. Sacerdoti chose to coin a phrase of "Patzer Problem" to to described the sentiment that Mint's ultimate acquisition by Intuit may not have fulfilled the objective of a "big" exit for some of its VC investors. Various other folks weighed in yesterday including Rob Hayes of First Round Capital (an investor in Mint) and Mark Suster of GRP Partners. I think Suster's points about backing entrepreneurs (vs markets, products, etc) and the fact that the VC business is one built upon long term relationships both ring particularly true.

I find it rather unnecessary to belittle Aaron Patzer personally or Mint in the way the original post described. I for one take my hat off to Patzer and everyone involved with the company for taking on an incumbent like Quicken, innovating around a new revenue model for personal finance software, and rapidly building a company from nothing to $170M acquisition in about three years. No matter what your expectations or relative perceptions may be, that's an awesome accomplishment.

I have literally no idea what dialogue Mint and its investors had at various stages of development, so I certainly won't try to opine on what did or might have occurred in their specific case. But the underlying issue of goals and expectations is one worth examining, because a broad alignment here is one of the key factors in successful relationships between entrepreneurs and investors. Ask anyone who's had parents or other family members help pay for their wedding... when you take other people's money, for better or worse their expectations come with it.

I believe in a few simple precepts which both entrepreneurs and startup investors should strive for:

1. There's a broad range of potential startup investors with different "bite size", strategy, and outcome expectations. There's no particular "right" model for VC, but entrepreneurs should familiarize themselves with different categories of investor prior to seeking funding.
2. Before committing to a funding path, entrepreneurs and investors should have a frank conversation about what they believe the company could accomplish and what each group hopes to achieve.
3. Entrepreneurs & investors who have meaningfully divergent expectations should part friends but not work together.
4. Where broadly aligned and proceeding with a funding round, entrepreneurs & investors should think about structuring the investment to reflect shared goals (size of round, terms, protective provisions, etc).
5. As companies continue to develop, entrepreneurs & investors should have a candid ongoing dialog to evaluate changing expectations and what the implications are.

For example, when we raised our Series A round for LinkedIn back in 2003 our CEO Reid Hoffman (my boss) had a conversation like #2 with Mark Kvamme who was planning to lead the investment for Sequoia Capital. We all believed we had a shot at building a large company, so the founding team of LinkedIn was prepared to agree to a structure which would essentially give Sequoia "extra" returns if we sold the company for a comparatively small amount (n.b. - basically a tiered liquidation preference which was high at low exit values and went away at higher values). Reid and all the LinkedIn investors have had a healthy dialogue over the years about whether or not to pursue additional funding, possible exit opportunities, and the like.

Even adhering to these precepts, it's certainly possible for disagreements to occur whether between founders and investors, among different founders, or among different investors. But clearly both investors and entrepreneurs can use these guidelines to try to avoid situations where startups pursue the wrong funding path or an incompatible investment partner.

"Cliff Notes" S-1s (Part II): Gamefly

Here's the second installment in my "Cliff Notes" versions of S-1 filings by internet & digital media companies that have recently filed to go public. The first one on Quinstreet is here, they priced their offering at $15.00/sh and started trading publicly on February 12th.

======================================

GameFly, Inc.

Filing Date: S-1 filed Feb 10, 2010

Founding Date: 2002

Headquarters: Los Angeles, CA

What They Do: Netflix for video games (console - Xbox, PS3, Wii, etc)

How They Do It: Gamefly website used to acquire & manage customer subscriptions, physical distribution centers handle game inventory which is fulfilled by US Mail

How They Make Money: Vast majority of revenue (nearly 90%) from consumers paying a monthly subscription fee ($15.95 - 36.95), small amount of revenue from sale of previously rented games and advertising

Financial Snapshot:

• 2009 Revenue: $93 million (1)
• Revenue Growth: 28% YoY (2009 FY), 44% YoY (2008 FY)
• 2009 Gross Profit: $45 million (1)
• Gross Margin: 48% --> i.e. for every $1 of revenue of GameFly spends $0.52 on buying game inventory, postage, and other costs of service (customer acquisition spend is not included in COGS)
• 2009 Net Income: $1.6M (1)
• 2009 Operating Cash Flow: $4.6M (1)

Subscription Metrics:

• 334K paid subscribers at 9/30/09
• 8.2% monthly churn
• $20.94 monthly avg revenue per user (ARPU) --> this appears to have held remarkably consistent from 2006 to today
• $25.90 cost per acquired subscriber (CPA) --> appears to have dropped from ~$30 in recent years
• $169.34 implied lifetime value (LTV) (2) --> i.e. for every $1 of marketing spend GameFly generates ~$6 of revenue or ~$3 of gross profit

Notable Aspects of Their Business:

• "Back" Catalog Usage: Over 50% of customer rentals are of video games more than 6 months old, which is likely an important driver of GameFly's profit margins. If most rentals were of newly released titles, their cost of sales would probably be higher for two reasons. First, GameFly's inventory would be "turned" far fewer times if most rentals were of new games... either stocking fewer games and having more customers churn (cancel their subscription) as they can't get the game they want, or having to keep a larger inventory of new titles (turned fewer times) but happier customers. Secondly, game publishers typically sell older games for lower prices than brand new ones so GameFly can presumably reduce their inventory acquisition cost (even wholesale).
• Delivery Beyond Mail Order: It's not huge surprise, but GameFly recognizes the threat/opportunity to expand into new delivery channels. This includes online gameplay and download which IMO will be difficult for GameFly simply because there are strong incumbents in online casual games (Zynga, EA/Playfish, et al), the challenges posed by online play of "hardcore" games (e.g. OnLive), and the threat of online distribution through console platforms themselves (Xbox Live, etc). But GameFly is piloting kiosk based game rental & sales with two national retailers (unnamed), so if the company began as "Netflix for games" perhaps they'll soon also be "Redbox for games".

Pre-IPO Funding History (3):

In total GameFly took approximately $20M in VC funding, which is comparatively modest for a startup with nearly $100M in revenue and poised to go public. Interestingly GameFly's chief backer, Sequoia Capital, owns a slight majority of the company (51.6%) prior to IPO. While it's not unusual for a VC syndicate to collectively own a majority of a startup by the time it reaches exit, it's fairly rare that a single firm has majority ownership. Sequoia was initially a minority investor but participated in every round of funding from Series A thru Series D.

Another interesting wrinkle (well, maybe not that interesting... except to perhaps us fellow VCs) is that Sequoia invested in GameFly out of two separate funds. In and of itself this isn't that uncommon, but the sequencing of the investments is somewhat unusual in that Sequoia X led the initial investment in GameFly's Series A and participated in every subsequent round. Sequoia IX, i.e. an older fund, invested first in GameFly's Series C and then participated in the Series D (final VC round prior to IPO). There's nothing inherently "wrong" with this and certainly a firm like Sequoia with a long and successful track record has broad latitude with LPs to pursue its investment strategy. But to the extent a VC firm is investing in a company out of two different funds, it's more common that the two funds initially invest at the same round or perhaps the older fund invests before the newer one.

Series A Preferred

• $2.0M round, closed 2004
• Major investor was Sequoia Capital [$1.75M]
• Pre-money valuation probably no more than $4.7M (4)
• 8% dividend, 1x non-participating liquidation preference (5)

Series B Preferred

• $4.6M round, close date unclear (sometime prior to 2008)
• Major investors were Sequoia Capital [$4.0M] and Peter Thiel [$0.25M]
• Pre-money valuation probably no more than $8.3M (4)
• 8% dividend, 1x non-participating liquidation preference (5)

Series C Preferred

• $6.2M round, close date unclear (sometime prior to 2008)
• Major investors were Sequoia Capital [$5.5M] and Peter Thiel [$0.25M]
• Pre-money valuation probably no more than $32.0M (4)
• 8% dividend, 1x non-participating liquidation preference (5)

Series D Preferred

• $8.5M round, close date unclear (sometime prior to 2008)
• Major investors were Sequoia Capital [$4.1M] and Tenaya Capital [$4.0M] (6)
• Pre-money valuation probably no more than $46.5M (4)
• 8% dividend, 1x participating liquidation preference w/ 3x cap (5)

Notes:

(1) Actually for 12mo ended 9/30/09, GameFly reports on a fiscal year ending Mar 31

(2) You can calculate an implied LTV for any subscription business where you know both ARPU and churn rate. Email me if you want to understand how to do this, but basically it's a decay formula to calculate the # of periods an average customer remains a subscriber and then multiplying that by the ARPU.

(3) Gleaned from S-1 itself and the exhibits (investor rights agrmt, certificate of incorporation, etc)

(4) S-1's typically provide info about number of shares, share price, and sometimes date of issuance for each round of preferred stock financing but not how many common shares were outstanding at the time. By 2008 there were approximately 4.7M common shares outstanding and 5.4M by 2009. It's conceivable there were fewer common shares at the times of the Ser A-D rounds though there were likely unissued shares for the option pool, so the fully-diluted share count at the time of the round could have been higher (implying a higher post-money valuation).

(5) The preferred stock classes will all be converted to common at the IPO, so liquidation preference and other rights will go away

(6) Tenaya Capital is the former venture capital arm of Lehman Brothers, which now operates as an independent firm

Disclaimer:

This post should in no way be construed as a recommendation to purchase Gamefly stock or any other security, these are simply my own personal observations.

"Cliff Notes" S-1s (Part I): Quinstreet

Venturesome off to a slow start here in 2010, but I'm working on a series of posts that will hopefully be interesting / useful to readers of this blog.

Since I previously described 2010 as a potentially strong year for IPOs, I thought I'd follow-up with some analysis of recent VC-backed internet startups which are in the process of going public. In the last several months there have been a number of S-1 filings including QuinStreet, Ancestry.com, ReachLocal, and Everyday Health (fka Waterfront Media).

S-1's are chock full of interesting information, but it takes a long time to comb through them and you have to know where to look to find the data. So I thought I'd do a series of "cliff notes" versions of these filings for reasonably quick and easy reading. In many cases I know folks who are involved with these companies as execs or investors, but I will be relying solely on public info from the S-1's and not anything communicated privately. Also my descriptions of what these companies do and how they do it aren't precisely how they might position themselves, but simply a rough distillation of their business for the interested observer. First up will be Quinstreet...

====================================================

QuinStreet, Inc.

Filing Date: S-1 filed Nov 19, 2009 (subsequently updated)

Founding Date: 1999

Headquarters: Silicon Valley (Foster City, CA)

What They Do: Online direct marketing for consumer services companies, primarily in for-profit education and financial services

How They Do It: Operate websites (or partner w/ existing sites) w/ content relevant to their clients' services, e.g. WorldWideLearn.com for education, CardRatings.com and Insure.com for fin'l svcs

How They Make Money: Essentially arbitrage paid search advertising and SEO to their network of sites and convert some portion of these visitors to qualified leads (name, email, address, etc) or clicks, which their clients then pay higher amounts for

Financial Snapshot:

• 2009 Revenue: $293 million (1)
• Revenue Growth: 36% YoY (2009 FY), 15% YoY (2008 FY)
• 2009 Gross Profit: $88 million (1)
• Gross Margin: 30% --> i.e. for every $1 of revenue, QuinStreet spends $0.70 on paid search, rev share (traffic acquisition costs - TAC) to partner sites, and other direct costs
• 2009 Net Income: $21 million (1)

Notable Aspects of Their Business:

• High Revenue Concentration: QuinStreet's depended on the education vertical for most of its revenue, though it's been diversifying in recent years primarily into financial services. Education accounted for 78% of revenue in FY 2007, 74% in FY 2008, and 58% in FY 2009. For-profit education firm DeVry is their biggest customer, singlehandedly accounting for 22% of revenue in FY 2007, 23% in FY 2008, and 19% in FY 2009. QuinStreet appears to be concerned about "additional challenge with regard to DeVry" which has "recently retained an advertising agency" and has reduced it's business with QuinStreet.
• Aquisitions of Partner Sites: QuinStreet has been highly acquisitive, spending nearly $190M in recent years to buy sites. This strategy gives them greater control over the content network and has enabled the company to both strengthen existing verticals and expand into new ones. Interestingly QuinStreet has used cash (both from operations and a credit facility) to complete most of these deals, as opposed to equity.

Pre-IPO Funding History: (2)

Series A Preferred

• $9.1M round, closed July 1999
• Major investors were Split Rock Partners [$3.6M] (3), Sutter Hill Ventures [$2.8M], and Rosewood Capital [$1.0M]
• Post-money valuation may be as high as $36M though probably lower than that (4)
• 8% dividend, 1x liquidation preference (5)

Series B Preferred

• $29M round, closed in Dec 2000
• Major new investors were J&W Seligman [$9.5M] (6) and Catterton Partners [$6.0M]
• Post-money valuation may be as high as $88M though probably lower than that (4)
• 8% dividend, 1.75x liquidation preference (5)

Series C Preferred

• There was a small ($500K) Series C round which appears to be related to a company Quinstreet acquired in the UK

Notes:

(1) 2009 calendar year results (1/1/09 - 12/31/09), QuinStreet reports on fiscal year ending June 30

(2) Gleaned from both S-1 itself and exhibits (preferred stock investor rights agrmt, certificate of incorporation, etc)

(3) Split Rock was formerly St. Paul Venture Capital, the VC arm of a large insurance company

(4) S-1's typically provide info about number of shares, share price, and date of issuance for each round of preferred stock financing but not how many common shares were outstanding at the time. By 2007 there were approximately 15M common shares fully-diluted outstanding, it's likely there were fewer common shares at the time of the Ser A (1999) or Ser B (2000).

(5) The preferred stock classes will all be converted to common at the IPO, so liquidation preference and other rights will go away

(6) J&W Seligman exited the private equity business and their Quinstreet shares were acquired by GGV Capital and W Capital in the secondary market

Disclaimer:

This post should in no way be construed as a recommendation to purchase QuinStreet stock or any other security, these are simply my own personal observations

2010: Year of the IPO

So the last week of December is often a time for formulating New Years Resolutions, predictions for next year, and the like. My prediction is hardly a bold one... that 2010 will be the year of the IPO, with many venture backed companies (and LBO's, spin-offs, etc) going public.

Back in early May, I identified 15 companies in the IT / internet / digital media space that I thought would be IPO candidates when public equity markets normalized. I thought it would be interesting to revisit that list to see what's happened with these companies. To be clear, my list certainly wasn't an exhaustive one. Others in this broad software/internet sector have also filed (e.g. ReachLocal) and companies in cleantech, semiconductors, and biotech have gone public or filed here in the second half of the year (A123 Systems, Fortinet, Solyndra, Codexis, et al).

But back to my list of 15. Two of these have actually filed an S-1 at this point to go public: Quinstreet (in November) and GAIN Capital (in August). Two have raised IPO-like private investment rounds, Facebook and Zynga, both involving liquidity to existing shareholders and both reportedly closed at multi-billion dollar valuations. A further two have been involved in acquisition processes with Amazon, Zappos actually closed at nearly $1 billion earlier in the year and Vente Privee rumored in deep negotiations to be acquired for $2-4B (USD) here at the end of 2009.

That leaves nine companies from my original list: LinkedIn, Vantage Media, Demand Media, Xoom, Yandex, Kayak, Endeca, Big Fish Games, and TheLadders. In addition to Facebook and Zynga which remain private despite the large investment rounds, my prediction is that at least five of these companies (i.e. nearly half) will file to go public sometime in 2010.

Happy New Year and a prosperous 2010 to all.

Avoiding the "I'm Here Now What?" Problem

As an investor in internet companies, I have the good fortune of getting to play around with lots of new consumer services. One of the biggest frustrations I sometimes have is when I first start using a product, only to get the feeling of "I'm here, now what?" shortly after engaging.

It's almost a cliche, but the first time a user tries your product is usually the best opportunity to convert them from a passer-by to a customer (however you might define that within your business model). Even if you've mastered some means of gaining consumer exposure to your product (PR, advertising, viral mechanism, word of mouth, etc), if the first time experience doesn't enable a user to quickly determine A) what the product is and B) why they might want to use it on an ongoing basis, then from a customer acquisition basis you've still failed.

To be clear, I think this can be a really hard problem to solve. It's especially difficult when you're trying to pioneer a product that's introducing a new paradigm in user behavior. And engaging the first 50K users is a different challenge than then next 500K or 5M users (i.e. early adopter behavior vs mass mkt). But if you don't explicitly think about this initial user experience when designing and building your product, you unfortunately might miss out on whatever opportunity you had to create something really exciting.

I'm no product guru but here are a couple of the high level ways I've seen companies successfully deal with this problem. I've tried to be generic here just to make this widely applicable, at least in theory.

Tightly Control the First-Time User Experience

As a particular user becomes more familiar with a particular product, I'm generally of the belief that consumer web services should give as much control to users as feasible. But I've seen companies be very successful at growing and retaining a customer base by actually constraining the first user experience in some way. It might mean limiting certain ancillary functionality, or directing a user through a particular product flow with few/no alternative paths.

Deliver a Nugget of Value Quickly (even small one)

The value of your product or service may accrue primarily over longer periods of time. But if a brand new user can at least realize a small portion of value immediately, it goes without saying they're more likely to stick around to create / receive the long run value. When we first launched the LinkedIn website in May 2003, the first few thousand users in the door honestly couldn't do much w/ the product (limited social graph, limited functionality, etc). But we tried lots of tactics to deliver little bits of value like resume-builder type profile creation, ability to browse your connections, statistics on what sorts of people where in your network. Perhaps much of this was simply "novelty" value at the time, but enough folks stuck around that as LinkedIn's network effects kicked in and functionality expanded, the "utility" value increased significantly.

Be Explicit, Take Nothing for Granted

The people who create new products typically live and breathe them long before they launch, and far more intensely post-launch than any normal consumer does. If you've been living this new product non-stop for a long time, it might be "obvious" to you why a new user should use it and what they should do in their first experience. But by not taking anything for granted, explaining things in a simple fashion, and explicitly communicating to a new user what the product does and why they should use it you stand a far better chance of captivating that new user. None of this has to be rocket science... it might be as simple as a 1st time user text / video, a big "Start Here" button, or whatever.

Actually Try Using the Product As Target Consumer Probably Will

This probably means a initial session of 2 mins or less, not 1 hr+. If you're building a mass mkt product, it means actually working properly in IE not just in Chrome or Safari with lots of fancy browser extensions installed.

Seek to Incorporate New User Feedback

The best consumer-facing companies usually solicit feedback from customers and try to incorporate the most relevant feedback into the design of their products. But a lot of the traditional methods for doing this (feedback forms, forums/communities, user testing) ultimately yield feedback from power users, or people who have something to complain or rave about. Getting feedback from new users requires actively and explicitly seeking it, and choosing to prioritize those product enhancements in addition to the things your most loyal customers might be clamoring for (typically the stuff that gets built more quickly). There's distributed services like uTest which make doing this a lot easier and cheaper than it was even just a few yrs ago.

There are probably plenty of other approaches... if you have other general purpose ideas or specific product examples in mind, I'd love to hear about them.

How long will the Facebook ad arbitrage last?

Over the course of 2009, I've encountered a number of companies that are advertising on Facebook's self-serve ad platform to great effect.

For those who may not be familiar with it, Facebook's ad platform allows advertisers a fairly broad range of targeting options derived from a consumer's profile data (geo-location, age, profile keywords, relationship status, etc) as well as both CPM and CPC based pricing models. The ads can appear in a variety of contexts including alongside profiles, applications, or other pages on the site. Among advertisers on Facebook, perhaps most widely known are social gaming companies like Zynga, Playfish, and Playdom. Zynga is reportedly spending as much as $50M annually on FB ads to acquire new users.

But I've seen examples of other advertisers, both large and small, increasing their spend on this platform. The main reason? The low cost of inventory coupled with useful targeting options produces can in many cases very good campaign ROI. Most of these advertisers don't buy traditional display ads, though some also spend on paid search. In fairness, most of these advertisers have a direct response model with online fulfillment (i.e. join a website, install a FB app, or complete an e-commerce transaction).

What's striking to me are the similarities between this platform today and Google in 2001-2003. I'm not necessarily predicting that FB is going to grow to tens of billions of revenue or hundreds of billions in mkt cap in the next five years. But back then search advertising was not only a new paradigm, but more importantly produced good ROI for early adopters (also direct response, by and large) because it was so cheap.

Yes, it wasn't low cost alone. The measurability and targeting enabled by Google's model of sponsored keywords sold on a CPC basis made it easy for direct marketers to determine ROI and continuously improve it. Back in 2002-2003 the cost per click for most keywords was pennies, not dimes, quarters, or dollars. But over time, as the search advertising industry matured, the magnitude of the initial opportunity provided by this cheap advertising has diminished.

These early adopter advertisers captured some of the value initially. Google has captured (or recaptured, depending on your POV) a large portion of this value simply by algorithmically improving yield through a combination of higher per click pricing, reordering ad display based on click-thru rate, and all the other innovative tricks they've developed.

In addition, a number of derivative companies emerged with an explicit arbitrage model of taking this low cost advertising and making it more valuable. I don't just mean the countless mom & pop SEM firms around the world, many of whom basically are in the business of making search advertising understandable to companies who previously never advertised or never did so online. But companies like Quinstreet (whom I identified as a potential public company back in May and which filed to go public last wk), Vantage Media, Efficient Frontier, ReachLocal, and others have all built very large (Quinstreet's >$300M revenue) businesses based on some form of search advertising arbitrage. In some cases this is simply aggregating demand from many small advertisers and in others it's buying CPC ads and turning them into higher value leads or acquired customers.

As Google's become more effective at extracting the maximum dollars from search advertisers, I've seen companies conscientiously stop the growth of paid search spend in favor of other channels. Overall Google's still growing by leaps and bounds of course, but some advertisers are now reaching that point where an additional $1 of paid search is only marginally profitable. Clearly the days of cheap search ad inventory, at least for most keyword categories, are behind us.

When will Facebook reach that point? It's probably years away, but is it 3yrs, 5yrs, beyond? Will companies like Buddy Media, SocialMedia Networks, et al build big businesses on top of the Facebook marketing ecosystem, in the way several have in paid search? FB obviously has massive amounts of inventory and still only a small number of marketers (compared to Google) spending on their ad platform. And there's also lots of other ways marketers can engage on Facebook (groups, fan pages / profiles, apps) besides just ad buys. But it will be interesting to see how the ad platform grows, and whether today's very low cost remains an arbitrage opportunity for very long.

"No Brainer" Startup Acquisitions

Monday's announcement of Google's acquisition of AdMob for $750M seems to be a good outcome for both companies. AdMob has established itself as the leading mobile ad network in the US, with a focus on contextual text ads on mobile web sites analogous to Google's AdSense network on the web. Google's efforts to extend their dominance in search and web contextual ads have been mixed at best, so the opportunity to buy the leader in this space makes a ton of sense.

To me, this deal falls into a category of tech startup acquisitions that I'd describe as "no brainer" deals. These are more than just "tuck in" acquisitions of a similar business to achieve greater scale or a far away leap into a new market.

Of course integration of people, technology, and business practices is never easy and doesn't always work out as anticipated. But for the startup there's a strategic fit with the larger company, an opportunity to extend reach or grow faster than as a standalone business, and typically a great outcome in terms of a premium acquisition. For the acquirer there's a chance to own a market leader, an ability to forgo further expenditure on internal efforts to compete with the startup, and hopefully a chance to accelerate the growth of the startup's business by bringing the resources and capabilities of the new parent to bear.

In addition to AdMob / Google, some other acquisitions that I think fall into this no brainer category:

1) Intuit's acquisition of Mint for $170M (Sept-09) --> from a user adoption standpoint, Mint was killing Quicken Online. Plus Mint's revenue model of targeted offers from financial services is a business model innovation from Intuit's traditional software license or subscription approach.

2) Amazon's acquisition of Zappos for ~$900M (July-09) --> Shared core competence in zealous consumer satisfaction, Zappos category leadership in shoes, easily merged physical operations.

3) eBay's acquisition of PayPal for $1.5B (Oct-02) --> At the time PayPal went public in Feb-02, roughly 2/3rds of our payment volume was directly attributable to eBay transactions. On the other side, eBay's own payment platform (Billpoint) never got traction from sellers despite deep integration and tens of millions of investment. The proof on this one has been in the pudding... as a division of eBay, PayPal did ~$2B in revenue last year on >$50B in payment volume. It accounts for more than 25% of eBay's total revenue and is the fastest growing component of the company.

What others am I missing here? My point isn't simply to highly the "no brainer" nature of these deals. Put simply, these acquisitions are classic 1 + 1 = 3 opportunities... and the ability to spot these opportunities for partnership, investment, or acquisition are the key for success in collaboration between startups and big companies.