AGILEVC My idle thoughts on tech startups

April 17, 2011

I thought it was time to revisit my series of “Cliff Notes” S-1s.  In 2010 I deconstructed the filings for GameFly and Quinstreet.  Quinstreet priced at $15.00/sh and debuted on 2/10/10, today they’re at $20.53 (close 4/13/11) an increase of nearly 40%.  GameFly filed in 2010 and remains in registration, though 2011 has seen a positive start for VC-backed IPOs with 14 in Q1 2011.  Last week ZipCar began trading and has a market cap of about $1 billion.

I’m pleased to now turn to Kayak’s S-1 filing.  Kayak was started here in my backyard of Boston… co-founder & CTO Paul English and the product/engineering team is based here in Concord MA.  Co-founder & CEO Steve Hafner and the business team are based in Norwalk, CT.  And Boston VC firm General Catalyst played a big hand in getting the company off the ground by bringing the co-founders together, as well as providing deep expertise in the travel industry and Kayak’s initial funding.  Now that Google’s acquisition of ITA is closed, following lenghty FTC review, it would appear Kayak is poised to proceed with their IPO in the coming months.

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

Kayak Software Corporation

Filing Date:  initial S-1 filed Nov 17, 2010, updated March 9, 2011

Founding Date:  2004

Headquarters:  Norwalk, CT & Concord, MA

What They Do:  Travel search engine

How They Do It:  Aggregate data from travel data warehouses like ITA as well as indexing travel providers websites, provide this information to consumers in a highly customizable search engine

How They Make Money:  Majority of Kayak’s revenue actually comes from advertising on their site (55%), not lead generation or referral fees to travel suppliers as you might think (more on this below).

Financial Snapshot:

  • 2010 Revenue: $170 million
  • Revenue growth: 51% YoY (2010), 1% YoY (2009), 131% YoY (2008)
  • 2010 Gross Profit: $162 million (adjusted EBITDA of $32 million)
  • Gross Margin: 94% –> i.e. for every $1 of revenue Kayak only spends $0.06 paying for travel data from ITA or others (customers acquisition spend is not included in COGS)
  • 2010 Net Income: $8 million
  • 2010 Operating Income:  $16 million

Search Statistics:

  • 2010 Queries: 634 million
  • Query Type:  85% airline, 11% hotel, 4% other (rental cars, package vacations, etc)
  • Mobile Usage:  8% of 2010 query volume (mobile app download growth of 344% YoY)
  • Revenue Per Query:  $269 RPM (revenue per 1,000 searches)

Notable Aspects of Their Business:

1) Kayak’s an Ad Business More Than a Lead Gen Business:  To some this may be semantics more than anything.  At the end of the day they offer a very popular travel search to consumers and then serve as a conduit funneling that traffic to travel providers and travel agents who then complete bookings.  But the majority (58%) of Kayak’s revenue comes from ads in and around the search results (which Kayak categorizes as “advertising” revenue) rather than users clicking on a specific flight or hotel to a travel provider or agent (which Kayak categorizes as “distribution” or “referral” revenue – aka lead gen).  Ad revenue is still mostly performance based CPC, though some sold on CPM basis.  Distribution revenue is CPC and CPA.  

At the end of the day Kayak’s playing a key role in the online travel process, but it appears more of the revenue comes from filling top of the conversion funnel rather than the middle or bottom of it. Historically more revenue came from distribution/lead-gen (57% in 2007), but this tipped in 2008 though appears to be steady from 2009 to 2010 at about 58% advertising and 42% distribution.

2) Impressive Organic Traffic: Kayak doesn’t break out it’s unique visitors by source, but they do highlight query volume by source.  An incredible 73% of their search queries are conducted by users who directly navigate to their sites… very robust organic traffic.  Only 7% of the search queries are performed by users referred by search engines, meaning Kayak’s brand is very strong and they continue to build it ($39M spent in 2010 on brand advertising, nearly $92M in total marketing spend).

3) ITA Relationship:  Much has been made of the fact that Kayak along with OTAs (online travel agents) license data from ITA Software which is now part of Google.  On the one hand Google generates significant search ad revenue from companies like Kayak, Expedia, and Orbitz but of course they may seek to compete with these companies in the future once they absorb ITA.  About 46% of airline related queries on Kayak (again vast majority of usage for flights) currently rely on data licensed from ITA and the current agreement runs out at the end of 2013, though obviously Kayak is working hard to reduce reliance on this data.

4) Revenue Concentration:  Given the dominance of companies like Orbitz and Expedia in the OTA sector, it’s not surprising that Kayak’s revenue is highly concentrated.  Kayak generates both distribution (i.e. lead gen) and advertising revenue from both of these companies.  Expedia accounted for 24.5% of Kayak’s total revenue in 2010 and Orbitz accounted for 18.2%.  AdSense and other Google advertising accounted for 7.7% of Kayak’s 2010 revenue.  So these three companies account for about half Kayak’s revenues today.

5) High Productivity:  Kayak had 148 employees at the end of 2010.  That means that Kayak generates roughly $1.15M in revenue for each employee which puts it in the same league as Google and Apple (both >$1M/head).  Obviously most of these employees are working hard primarily for equity upside compensation, but Kayak’s personnel costs are roughly $200K/head so the company is highly productive on a per employee basis.

Pre-IPO Funding History:

Kayak has raised approximately $235M in VC funding to date.  Collectively the investors own about 70% of the company, and co-founders Steve Hafner and Paul English each own close to 10%.  Interesting to note that Hafner and English own common stock but also made meaningful investments in the Series A & B rounds.  The largest holders General Catalyst (29.7%) and Sequoia Capital (17.6%) both invested in Kayak accross multiple funds.  GC Funds II, III, and V all invested and Sequoia invested both out of a core fund (Sequoia XI) and one growth fund (Sequoia Growth III).

Series A Preferred

  • $6.6M round closed in June 2004
  • Led by General Catalyst with participation by co-founders Steve Hafner & Paul English
  • 1.5x liquidation preference, 6% accumulated dividend (1)
  • Post-money valuation probably no higher than $12M (2)

Series A-1 Preferred

  • $1.65M extension round closed Nov 2004
  • General Catalyst & Sequoia participated
  • 1.5x liquidation preference, 6% accumulated dividend
  • Pre-money valuation was initially set higher but was adjusted to match the Ser B valuation

Series B Preferred

  • $7.0M round closed Feb 2005
  • Led by Sequoia with participation by General Catalyst
  • 1.5x liquidation preference, 6% accumulated dividend (1)
  • Pre-money valuation was approx. $17M (2)

Series B-1 Preferred

  • $3.0M extension round closed April 2006
  • Sequoia & General Catalyst, same terms as Ser B

Series C Preferred

  • $11.5M round closed May 2006
  • Led by Accel (London) with participation by General Catalyst & Sequoia
  • 1.5x liquidation preference, 6% accumulated dividend (1)
  • Pre-money valuation was approx. $36M (2)

Series D Preferred

  • $166M round closed Dec 2007
  • Led by Oak Investment Partners with participation by General Catalyst, Sequoia, & Accel and others
  • 1.5x liquidation preference, 6% accumulated dividend (1)
  • Pre-money valuation was at least $250M (2)
  • This was the large round to fund the acquisition of Sidestep

 

Notes:

  1. In an IPO preferred share classes are converted into common stock, and liquidation preferences and accumulated but unpaid dividends essentially go away.
  2. One can infer valuations based on per share prices of preferred stock and oustanding common shares (~5.3M as of 12/31/09).  Precise valuations are impossible to determine because of adjustments to employee option pools, possible buybacks of common stock, etc.  But these should be directionally correct.
  3. All of the above is my personal analysis and opinion, nothing here should be construed as investment advice or a recommendation to buy or sell Kayak stock.