Byrne's Blog

LinkedIn’s IPO Filing: Analyzing the S-1

Note to readers: my new startup research boutique, Digital Due Diligence, performs in-depth research on dozens of companies, public and private, including LinkedIn and its key competitors. Visit our equity research page for more information.

Last month, LinkedIn filed a prospectus with the SEC. It’s a great case study: LinkedIn is one of the largest social networks, and it may be the most mature of the major social networking businesses. Like Facebook, LinkedIn has turned a profit; unlike Facebook, LinkedIn’s profit is dependent on several known business models.

The filing itself was a better read than most. Among other risks, it cited the possibility that:

[O]ur initial public offering could create disparities of wealth among our employees, which could adversely impact relations among employees and our culture in general.

I guess that’s what you get when your Chairman is an avowed “free-market socialist.”

The hurdle LinkedIn’s IPO faces is that the easy comparisons are completely wrong. LinkedIn shouldn’t be valued like a job board: for a variety of reasons, it’s a much more defensible model; LinkedIn is likely to stomp all over the traditional job boards (not so much traditional recruiters). But it’s not a “social network” that can be valued like Facebook and Twitter. LinkedIn’s business model is already in place, and its potential is a matter of execution, not innovation.

LinkedIn’s Competitive Advantage

Two of the strongest monopolies online are search engines and social networks. There’s rarely a good reason to use a less popular social network, and search engines have notoriously unforgiving economies of scale.

LinkedIn has solved both problems: for many users, it’s a second social network and a second search engine.

In both cases, LinkedIn isn’t competing with the dominant companies in the industry, because it’s targeting a very narrow use case. In search terms, LinkedIn is the ideal tool for finding job candidates with a very specific set of skills—project managers who know Perl and live within commuting distance of San Diego, or Ivy-Educated distressed debt analysts with at least a year of experience, for example. There’s no good way for Google or Bing to parse that query, but LinkedIn has that information at its fingertips. For recruiters, whose main “overhead” involves looking at résumés that turn out to be irrelevant, this is key.

As a social network, LinkedIn has positioned itself as the place for things you’re willing to brag about, which are not fun. This is brilliant. Most social networking tools are designed around the human need to show off. That can easily be divided into two basic categories: being happier than other people (partying, making great jokes, having kids, being close to family members, discovering funny videos), and giving up more happiness than other people (working hard, taking business trips, publishing a complex academic paper, dealing with a job search).

That’s a great niche, since things that aren’t fun are often the most impressive to potential employers. LinkedIn has clearly pressed this advantage by adding a powerpoint presentation-sharing app, an itinerary app, and a book-tracking app to user profiles. In my experience, very few people use the book-reading app to talk about fiction—they’re usually reading Too Big To Fail. The travel app isn’t for reporting on vacations; it’s for reporting on business trips.

This gives LinkedIn a purpose as a social network. They aren’t competing head-to-head with Facebook or Twitter. Instead, they’re absorbing the content that’s too boring for either. At the same time, Facebook and Twitter provide an outlet for anything less professional. It’s a symbiotic relationship that allows each network to move more content online.

(Yes, LinkedIn has Twitter integration. And some people do use Twitter to post mostly professional information. Twitter is a medium less dependent on message than LinkedIn or Facebook, so it makes sense that it could be cloned or integrated by each.)

LinkedIn also has another advantage: they have consciously tried to own people’s names as an SEO strategy. This is probably responsible for a huge fraction of their traffic. They don’t find people whose search queries imply intent: they don’t show up for terms like “tech jobs” or “find jobs,” and their ranking for “professional networking” is due to their Q&A page. LinkedIn is trying to find people who are looking for other people in a professional context, and they’re trying to make LinkedIn the way to do this.

LinkedIn actually has a great relationship with search engines; they even have “social sitelinks”:

That’s a tough policy for Google to reverse, so LinkedIn’s SEO risk is overstated.

LinkedIn’s competitive advantages rely on network effects. There is simply no better place to find out about someone’s professional skills and reputation. And there’s no better place to upload your résumé. LinkedIn has supplanted the personal site (will a hiring manager find it?) and the job board (what about when you’re not actively looking for a job?).

That’s actually quite impressive. LinkedIn is a well-executed version of two of the big business models of the 21st century: search and social.

LinkedIn’s Financials

Anyone can look at LinkedIn’s recent performance and see the salient numbers: quarterly revenues have risen from $23mm in Q2 2009 to $61mm in Q3 2010, an incease if 17.7% per quarter. Gross margins have gradually increased (75.9% to 80.9%) and variable cost as a percentage of revenue has dropped (63.8% to 57.9%) during the same period.

But the numbers get more interesting when we look at scalability:

• LinkedIn’s quarterly “hiring solutions” revenue per user has risen from about $.17/user to $.33/user. This is a reflection of the site’s network effects: with higher and higher saturation, it’s increasingly easy for LinkedIn to offer the very best applicant for a given set of criteria.

• LinkedIn’s quarterly marketing revenue per user has grown from $.16/user to $.24/user. This is a good sign that they’ve gotten better at price discrimination, and scaled their sales force up to be appropriate for the number of pageviews they handle.

• LinkedIn offers a “Premium” account for recruiters and job-seekers, offering more access to profiles, and more chances to contact users. “Premium” revenue per user has declined in the last six quarters, from $.29/user to $.19/user.

The numbers paint a pretty simple portrait. LinkedIn is three major businesses:

• Premium Subscriptions were an early cash cow, and still bring in $15.7mm/quarter. But they’re growing slowly. LinkedIn appealed to hyper-networkers early, but there aren’t that many of them out there. And LinkedIn needs to keep these prices high, to ensure that these networkers don’t bother other users.

• Hiring Solutions is LinkedIn’s most effective vehicle for monopolistic pricing. This is the revenue stream to look at when considering LinkedIn as a job board (versus LinkedIn as a social network or LinkedIn as a source of high-quality pageviews). LinkedIn is growing their hiring solutions revenue rapidly, in terms of revenue per user (12.4% quarterly growth), revenue per employee (10.5% quarterly growth), and revenue per “corporate user” (7.4% quarterly growth). This product makes LinkedIn most comparable to a job board in terms of monetization strategy. But LinkedIn’s competitive advantage, discussed above, put it well ahead of other job boards.

• Marketing. LinkedIn’s marketing revenue is icing on the cake. As a large site with a capitive audience and great SEO, LinkedIn generates pageviews. Lots of pageviews. And it sells them. LinkedIn’s marketing revenue per user has risen 6.1% per quarter in the most recent six quarters. Combine that with their user growth of 13.7% per quarter, and the result is a massive ad platoform.

“Marketing” and “Hiring Solutions” are both high gross margin businesses. LinkedIn has been able to continuously raise revenue per customer even as they go after more customers, which is a good sign that they haven’t nearly hit the point at which this market is mature.

Overall, LinkedIn’s marginal revenue per user is in the $.50 to $1.25 range, depending on the quarter. This is probably driven by two forces: on the downside, the quality of their users is declining over time as the extend their reach to people who a) have lower-prestige jobs, b) aren’t very active online, or c) are less likely to seek out new jobs. On the other hand, every new user raises the value of the site’s marketing and hiring solutions.

LinkedIn may end up using their “Premium” accounts as a way to reduce, but not eliminate, annoying interactions. They can gradually strengthen the site’s privacy settings (as they have already done), while loosening them for paying customers. That means LinkedIn captures lots of the economic gain from invading people’s privacy or sending them unsolicited messages. And since LinkedIn benefits in the long term if those interruptions are rare enough that users don’t flee the site, it’s an optimal arrangement.

The best-case scenario for LinkedIn is that they find a new stream of revenue to join this triumvirate, adding more high-margin profits to their existing revenue. But even if they don’t, they’re in a great position. All of their major revenue sources are growing in a way that implies that additional users won’t seriously disrupt things, and that additional sales staff will offer a useful contribution.

Thus, while LinkedIn is profitable, and has a few conventional business models, it’s still worth considering the possibility that it will have a sudden, sharp acceleration in growth.

One note regarding LinkedIn’s financials: they have some “variable” costs that vary with revenue (e.g. hosting costs, marketing expenses), and some that vary on a discretionary basis (e.g. investment in new products). LinkedIn has cut back their product development as a percentage of revenue, from 36.8% in early 2009 to 27.8% in mid 2010. This accounts for 138% of their increase in operating earnings. In other words, if they invested in R&D like they used to, they’d be losing money.

This is not, strictly speaking, a bad sign. LinkedIn doesn’t have a comparative advantage in this kind of spending—Google and Facebook will always be able to outbid them for the best technical talent. But LinkedIn is still a technology company, so it’s important to note that their profitability is a result of being less technical and more of a sales-and-marketing company.

There is a decent chance that this is an abberation, and that they’ll raise their product spending later on. This would account for their projection that they won’t be profitable their first year as a public company. This isn’t a useful long-term concern, but anyone buying on the IPO and looking at the next quarter should beware.

The LinkedIn Management Team

LinkedIn is traditionally considered an offshoot of the “Paypal Mafia.” Their founder, Reid Hoffman, certainly got some great experience—and a healthy grubstake— from Paypal’s IPO.

But LinkedIn’s management also hails from Yahoo!, not LinkedIn. Their CEO, Jeffrey Weiner, worked there, as did their SVP of operations and engineering. They’ve also recruited from TiVo (CFO Cteven Sordello) and Google, (Dipchand Nishar, SVP of product and user experience).

Oddly enough, many of LinkedIn’s senior executives were promoted to their current roles just in time for the IPO. Of their top six executives, four were promoted in January. This is probably a sign of the usual pre-IPO reshuffling (in a company of that size, it’s important to identify which VPs are worth talking about!) but it may also have something to do with LinkedIn’s options grants.

LinkedIn’s Competitors and Risks

LinkedIn is not a job board. It shouldn’t be compared to them. However, the market will probably look at job boards as the most similar companies to LinkedIn. A few of their competitors:

• Monster Worldwide trades at 2.5X sales. It’s fairly mature: the big determinant of Monster’s revenues is the market for jobs in general.

• TheLadders, like LinkedIn, targets high-income professionals. TheLadders has, by all accounts, done a great job of extracting more revenue from top-tier job applicants. But it’s a site for finding a new job, not curating a career.

• Dice Holdings is a more interesting case. They have vertical-specific job boards (Dice.com for technology, eFinancialCareers for finance, ClearanceJobs.com for government jobs, etc.). Their operating margins approach 25%, and they’ve shown some recent growth. Dice trades at 8X sales.

Most of the publicly traded companies in the employment space are recruiters, not job boards. Recruiters have been early adopters of LinkedIn (hence that high but declining “premium” revenue). LinkedIn may gradually erod the economics of that business, but in the short term recruiters are happy to embrace it.

There are a few companies that are actively eroding LinkedIn’s competitive strengths:

Hashable is trying to own the “introduction” part fo the LinkedIn relationship. While that’s critical in the short term, LinkedIn monetizes the metadata—where someone worked, who recommendeded them, and who they know—not just information on who met whom. Hashable is a good acquisition target for LinkedIn, and might be a long-term threat, but even if they own the personal introduction market, LinkedIn still has a lock on the long-term employment data market.

Branchout is a new Facebook app that offers career information, hosted on Facebook. This could be a threat, since there are more Facebook profiles than LinkedIn profiles. But connecting Facebook activity is a bad move (the big career risks: for people under 30: pictures of parties! For people over 30: pictures of kids!). The new Facebook profile has a similar risk/reward profile: yes, it offers some of what LinkedIn offers, and does a better job. But it’s easier to create a LinkedIn account than to massage privacy settings to the point that Facebook is a useable tool for interacting with friends and bosses.

Quora is destroying LinkedIn answers. It is what LinkedIn answers should be: a site where the right expert answers the right question. Quora launched with a good audience; LinkedIn’s good audience had gotten less active by the time they launched Answers.

Applicant Tracking Systems are a broad competitor to LinkedIn. Once the hiring process gets started, these content management systems get used a lot. The problem with them is that they vary in quality: from being awful but useable to being awful, unuseable, and ubiquitous.

LinkedIn could easily squash Taleo, one of the largest applicant tracking systems, by creating a one-click application system, that ordered applicants based on an algorithm that looked at their experience and their degree of connection to the company. This wouldn’t just give LinkedIn a new revenue source: it would help LinkedIn replace company “Careers” pages the way it’s so thoroughly replaced personal résumé pages.

Ten Questions for LinkedIn’s Management

• Why did discretionary expenses drop so much ahead of the IPO?

• What will you do to address Hashable?

• What will you do about M&A replacing HR? What if hiring is obsolete and the best employees get found because they start compelling companies? This doesn’t post big short-term risks to the LinkedIn model. But anecdotally, programmers are overrepresented among LinkedIn user (or at least LinkedIn ads). If LinkedIn is the place to hire not-the-best programmers, how does that affect the site’s pricing power?

LinkedIn can use price discrimination to deal with part of this problem. Instead of charging full price for all companies, LinkedIn can offer discounts to funded startups with a small number of employees. If LinkedIn is a great tool for hiring the first ten people, it may be a cost-effective too for hiring the next hundred.

• Field sales are growing faster than online sales. These are higher-margin deals, with bigger lead times and more generous payment terms. What will the next big hiring slowdown look like, from a balance sheet perspective?

• Are you going to remain active in the Q&A space?

• How wil you deal with out-of-date profiles? As LinkedIn matures, the average age of profiles will go up. This will reduce the utility of profiles, which is the main driver of non-ad revenue. (And since ad revenue is driven by pageviews, it’s indirectly affected by how often people log in). LinkedIn could partially mitigate this problem by sending more emails. Whenever an employer lays people off (as announced in the news or as noticed by LinkedIn), LinkedIn could email that compay’s employees and suggest that they update their profile and ask friends for a recommendation.

• How will you deal with Facebook? Facebook is the elephant in the room. If they perfect their privacy settings, LinkedIn’s only reason for existing is its relationships with recruiters, and its database of recommendations. Facebook could clone both within a year or two, if necessary. What is LinkedIn doing now to make sure that’s not an economically viable decision for Facebook?

• Do you have another business model in the works? Marketing, hiring solutions, and premium accounts are great. But are they everything?

• Will you ever partner with other job boards? LinkedIn invades the job board niche, but many job boards still have valuabe résumés in their databases.

• Will you partner with a recruiting agency? Recruiters with niche dominance could use LinkedIn especially well. They may be able to find recruiters who dominate a particular market, and are willing to pay up for special rights.

A Few Post-IPO Predictions

(In the spirit of Byron Wien’s “Ten Surprises,” these are not things with high odds of happening, but events whose odds are higher than one would normally think.)

• LinkedIn has a few soft quarters early on, but shows solid growth in FY2012.

• LinkedIn buys an Applicant Tracking System.

• LinkedIn buys Quora for $500mm. This is widely derided, but immediately raises their revenue from every business segment, and gets them some great technical and design talent. Facebook will have trouble buying Quora for personal reasons; Google because it already owns Aardvark; Twitter because it doesn’t buy other product. Yahoo, Microsoft, and AOL are the only other suitors. I doubt that either company is willing to pay so much for a pre-revenue company.

• LinkedIn subsumes the company “Careers” page the way they’ve subsumed the personal “résumés” page. They do this through a combination of SEO and purchasing an applicant tracking system.

• LinkedIn does not buy or get bought by a job board.

• LinkedIn’s stock, at IPO, is a good deal. At some point, it will probably close 20-30% below the first day’s closing price, but it will be considered a good investment five years down the line.

Full Disclosure: don’t consider this financial advice. Consult with a broker before making financial decisions, even poor ones. Assume the worst, and act on it.

In addition: I am an avid LinkedIn user. I will not be buying at the IPO, but may trade LinkedIn stock or related derivatives in the future. I will not necessarily trade in the direction implied by the above.