Forget the debt ceiling. As far as Internet investors are concerned, Apple is the macro situation—it’s the central bank, the legislature, and the source of exogenous unpredictability in the web.

  • Apple owns a growing number of screens, eyeballs, and minutes. However you measure it, Apple products are a conduit for more and more attention.
  • Apple creates and dominates categories. MP3 players. Smartphones. Tablets. Every time Apple reinvents a slice of the industry, they’re a little more web-focused.
  • Apple products are purchased by wealthier, more tasteful consumers. A tasteful consumer base means that what Apple sells to 1% of the market now will be what 10% of the market wants next year. And as consumer electronics prices drop from year to year, they’ll be able to afford it.
  • Apple creates the products that other people use to reinvent their industries. Visit any startup. You’ll see the founder taking a call via iPhone while coding on a Macbook Air, and checking how the site looks on an iPad. Apple’s market share among startups has to be north of 50%. All this Apple product usage subtly shapes the sites people create (look at how boxy and linear sites were ten years ago—make them on a Dell, and you’ll get something Dell could have designed).

Apple owns the markets that define the mainstream.

Why not Google? A few years ago, Google would have been in the same position. And, in fact, Google’s unilateral decisions do have an immediate effect on businesses. But Google moves in a predictable direction: anything that raises the number of searches performed by the average Internet user, or reduces the scalability of non-AdWords, non-DoubleClick online marketing, is something they’ll be in favor of. Google is a company with a monopolistic position they need to preserve. Apple is a company that’s creating new industries (and new monopolies) out of thin air.

Google’s the one to worry about, but Apple’s the one to watch.

How Google Local Bootstrapped Out of an Antitrust Crisis

Earlier issues of Digital DD Weekly have noted that search engines use their semi-monopolistic status in order to make all-or-nothing offers to site owners. Local search is a great case study here. Essentially, Google gave sites with reviews two options: to let Google use those reviews as if they’d been written for Google itself, or to be banned from Google entirely. Given how much review sites rely on Google for traffic, it was an obvious, albeit painful, decision.

Of course, this policy opens up a thorny anti-trust offering. If a firm is in a position to make an all-or-nothing offer like that, it’s a bad sign as far as regulators are concerned. Fortunately for Google, there’s a way around this: they used third party reviews to make their Places offering popular enough to earn its own reviews—now they’re no longer using third-party reviews. Ironically, this is quite similar to the way that some sites use paid links to game search results, and then use their gamed search results to garner enough legitimate links to rank naturally. It works for black-hats, and it works for Google.

(Meanwhile, Google has removed some of the information that local marketers used in their competitive research. SEOMoz has a post on getting that data back.)

Google+: No Pseudonyms, Weak Privacy

Google+ has been lauded for its ideological focus on customized privacy. As has been pointed out in DDD Weekly before, this is not entirely fair: Facebook lists and Google+ circles are the same functionality, but Facebook has learned that mainstream users don’t like to use them. They’re clearly committed—they just bought frid.ge, the last of a clutch of startups (CollegeOnly and Fabulis were others) built entirely on the premise that Facebook’s privacy settings were too confusing for regular users.

Now, Google+ is facing two bugs. One is a customer relations bug: they’re disabling accounts for using pseudonyms, including an account started by someone who worked on the Google+ team. There are some signs that a name is fake (i.e. unusual punctuation marks), but there are legitimate edge cases, too.

The other is a series of technical bugs: tagging images automatically shares them and Google +1 buttons appear to report mouse movements. These bugs all lean in the same direction: automatically sharing data that users intend to keep private. That’s very useful for Google, but it weakens their brand. At this rate, their 2% conversion rate to Google+ may not see much improvement. As BetaBeat points out, email is a huge component of this.

Google’s Secret International Q&A Site

Google Confucius is a Q&A site available in dozens of countries outside of the US. The research Google has published on this project indicates that it’s being used as a search fallback. Right now, there’s not a specific UI element that turns a failed search into a Q&A. The closest equivalents are on the user side (someone asking a question on Yahoo Answers or Quora when they can’t find good results on Google) or through companies like Demand Media using ISP logs of searches to determine which articles to write. If Google can link search and Q&A, they’ll be far ahead of both Quora and Demand.

Google Begins Vendor Financing

This is an interesting move. Google is offering AdWords users an AdWords-only credit card. Since gaming FICO is a surprisingly common hobby in the search engine marketing world (once you’ve outsmarted Google’s algorithm, getting around other algorithms is a snap), these aren’t great credit card customers. At a large enough scale, it’s sensible for someone to provide financing for AdWords campaigns—if a user’s lifetime value is significantly higher than the cost of acquisition, that user is essentially collateral for the loan. But vendor financing has ruined more than a few companies. It’s not a good sign if Google is the only organization willing to provide this at scale.

Amazon’s Picks and Shovels

Amazon is reluctant to tell investors how much of their revenue comes from cloud computing, but it’s surely an impressive number. It’s also hard to calculate that unit’s profits based on conventional accounting: the original model was to use Amazon’s spare computing capacity, which had a minimal marginal cost. Right now, they use a complex combination of guaranteed and spot pricing to earn the maximum revenue possible from their computing assets. Exact revenue numbers aren’t available, but the street is excited about it nonetheless.

ComScore Reports the Top Sites of 1996

This report on the most popular sites of 15 years ago is worth contemplating. Experts’ understanding of the Internet is based on their experience, but their experience skews towards the early years of the web. The people who know the most about the Internet remember a time when the typical Internet user wasn’t just college educated, but was a college student (or, more likely, a grad student). As web content shifts from what younger males want to what middle-aged women want, “Internet” expertise will matter less than media and consumer expertise.

One other point from this list: people have always spent a lot of time leaning on third parties to sort and search the web. And those businesses are more durable than the content businesses that they promote. AOL and Yahoo remain top web properties; Penthousemag.com isn’t even a top performer in its market.

Betting on Fresh IPOs is Now Easier

UBS has launched two new ETNs tracking recent tech IPOs, including LinkedIn, Homeaway, Yandex, and Rackspace. It’s fine for them to meet the demand here, but this doesn’t speak well of the average investor. There is no coherent investment thesis that involves buying these stocks as a group. Hopefully, this ETN will be used as an easy hedge against speculative froth in the IPO market.

LinkedIn and Weak Links

Rick Bookstaber (quant and commentator responsible for one of those books that risk managers only had time to read once they were laid off in 2008) has a fascinating essay on LinkedIn and the power of weak ties. It’s worth reading.

SAC (Managers) Long Facebook

Business Insider reported that SAC Capital is buying into Facebook. They claim that the NYT confirms the story. More accurately, the NYT clarifies the story: SAC managers are buying Facebook, possibly for their own accounts. This is somewhat contrary to SAC’s reputation as a rapid-fire trading shop, but is entirely in line with their reputation for being able to spot a good deal.

The Cost of Not Having Spam

An Etsy employee has a superb blog post on the difficulty of creating scalable spam solutions in social media. The upshot of this is that perhaps the only way for there to be effective spam filtering in social media is for the dominant sites to be so dominant in their niche that there is simply no alternative, even for people victimized by false positives. For once, Google is not acting like a benevolent monopolist: by creating the most plausible direct competitor to Facebook, they’re making it more difficult for anyone to solve the social spam issue.

Infographic: Google’s Most Lucrative Ad Categories

Wordstream has a nice graphic highlighting Google’s most expensive keyword categories in AdWords. (Try “structured settlement securitization” at $151/click, for example.) But it’s a good summary of where Google generally gets their money: from complicated products on which there’s lots of data, few good brand names, and a big information gap between buyers and sellers. No wonder they’re touting brands and display in a big way: that’s a whole lot more fun than making money off of debt consolidation, online degrees, and rehab.

Contently Tries to Disaggregate Demand Media

Contently is creating a way for sites to source content of a given quality level, from good freelance journalists. They’ve just raised a round. Their business model is actually similar to about a third of Demand Media’s model. Contently is basically outsourcing content ideas and content marketing to other people. Here’s the problem: this part of the platform is really tough to run—someone else has to base their business on Contently in order to use it effectively, and that’s a big bet. What Contently may have to do is what Demand did: raise a bigger round, buy out an established site, and use that as a proof of concept.

There’s a very common tendency for “platform” companies to fail, but for their proofs of concept to work just fine. (Notably, Pixar made their first animations in order to show off the computers they were trying to sell.)

The Mismeasure of Twitter

The bit.ly blog has a new post measuring how much traffic really comes from Twitter. Their conclusion: “If the initial link was shared on twitter, and you see 60-70% direct clicks, it’s possible to conclude that the twitter traffic is undercounted by a factor of ~5.”

The Reddit blog has a similar post on the life cycle of a link.

SocialBios: An Odd Acquisition

Realtor.com has acquired SocialBios, but that was not the right move. SocialBios is a good feature for their site, but it’s valid in plenty of other demographics—pretty much anything in the professional services category (doctors, lawyers, accountants) would work just as well. By tying this product to just one industry, Realtor.com is basically giving away 90%+ of the idea to whoever else is competing with SocialBios.

This kind of startup can be cheap, since the business model isn’t always clear from the start. But that doesn’t mean that buying it is necessarily a good move. A partnership could have conveyed the same economic value to Realtor.com, while leaving SocialBios free to develop a broader business.

AdWords Adds Call Metrics

Google now offers extra analytics for phone calls derived from AdWords ads. This turns more of companies’ marketing budgets into something measurable (i.e. something Google can charge the market price for). For now, they’re charging $1 per call. At six minutes per call on average, that’s too good a deal to last.

TripAdvisor: Payola for Me, None for Thee

TripAdvisor is removing ads that appear to be incented, but also incenting reviewers. This is imminently sensible from their perspective: all else being equal, one more review will make the site incrementally better. But one more review paid for by the reviewee will make the site a little bit worse. That’s bad news for local marketers who rely on incented reviews (often, these are folks in dowdier industries—very few people get online to rave about their plumber, though they might post a complaint).

Zynga + Facebook: When Market-Dominant Players Negotiate

Zynga has just disclosed some fascinating details on their relationship with Facebook. When so much of Zynga’s user growth is due to Facebook, and so much of Facebook’s revenue and activity is due to Zynga, it’s not surprising that they have a close relationship. What is surprising is that Zynga is giving Facebook exclusives on some games, and giving them a heads-up when new games are introduced. In exchange, Facebook is guaranteeing Zynga traffic.

This is the future of tech M&A: no M&A, just very close deals on frictionless platforms.

Amex Gets Deals Right

Amex and Facebook are collaborating on a new deals platform that links Amex cards to Facebook accounts, and then shares deals via Facebook. The Amex cardholder gets credited the value of the deal on their monthly statement. This is a brilliant move, which fixes two problems with typical daily deal sites:

  • Normal people feel a little bit embarrassed to pay with coupons rather than cash, an the swankier the establishment, the more embarrassed they are.
  • Restaurant staff tend to loathe daily deals, because so many patrons tip on the post-discount total.

LinkedIn CEO Embraces “Peak Minutes”

The web is getting more crowded. Jeff Weiner, LinkedIn’s CEO, argues that Google+ won’t succeed because “nobody has any free time”. This is a good argument for niche networks over general-purpose networks, or at least an argument that being a merely good enough general-purpose network is not good enough any more.

This will be a dominant factor in the online media narrative of the future: as content explodes in quantity and time spent consuming content grows slowly, if at all, the tailwinds for new online businesses will vanish. Acquiring users online will eventually be similar to acquiring users in the real world—it’s going to take some kind of expensive “storefront,” and a more traditional media campaign, rather than a low-cost presence that naturally leads to traffic.

GazeHawk on the Human Eye

Designers spend a lot more time thinking about Photoshop than physiology. That’s a problem. YCombinator-funded startup GazeHawk has a great blog post on how the human eye operates, and what it focuses on. Required reading.

Google’s Evolutionary Strategy

Why would Google warn users about malware running on their own machines? It’s not public-spirited; it’s purely Darwinian. All else being equal, if Google users have faster computers and face fewer phishing attempts, they’ll end up doing more searches and being more successful in general. That’s a cheap win for Google.

Bing Testing Inline Ads

Bing is testing ads within search results rather than off to the side. User surveys consistently indicate that some users assume that ads are not, in fact, ads, but are specially-selected results (that’s why they’re highlighted, of course). So this is more taboo than it should be. And this layout should be familiar: it’s how eHow places ads, too.

Social Network Neutrality Considered Harmful

Betaworks entrepreneur in residence Jake Levine calls for social network neutrality. This is weak for the same reason that general net neutrality and search neutrality arguments are weak: there is rarely a business case for exploiting non-neutral social media activities. And when there is, it ends up being net beneficial for users. Google is a better search engine because Google non-neutrally displays weather forecasts rather than adding a click to the process; Facebook is a better social network for putting house ads and new features ahead of client ads when it suits them.

And if social networks are so valuable that they need to be regulated, it seems dangerous to regulate them in the first place: social network neutrality will reduce the flexibility and value of social networks, which is one more way to prevent the next Facebook from getting off the ground.

Associated Press Discovers Linking (Sort Of)

The Associated Press will link to sources when a member newspaper is substantially responsible for the content of one of their articles. However, the AP is using an odd (print-friendly?) implementation: they’re crediting the original source in the text of the article, with a bit.ly link to the source’s article. This may be a clever hack, but it also means that search engines won’t give original sources quite the credit they deserve.

DIY Doesn’t Scale

BetaBeat has a healthily skeptical piece on New York tech darlings Etsy and Kickstarter. They don’t scale well—which is fine, in a way, since they’re startups that own the most scalable part of a largely non-scalable business. Being a real estate broker doesn’t scale. Zillow sure does.

Zillow’s IPO a Success

Residential real estate site Zillow had a great IPO, hitting $60 per share for a $1bn+ valuation, before settling in the mid-30′s—far higher than their most recent $16 to $18 range. Zillow earns a big chunk of their traffic from search, but the ultra long-tail searches in which real estate really shines (addresses and MLS numbers) they don’t do quite as well as their competitors.

Stampt Reinvents Foursquare

Stampt is a new startup trying to recreate card-based loyalty programs. So is Foursquare. Stampt is doing it by faithfully recreating the experience of stamping a card. Foursquare is doing it by creating a massive game whose side effect is a record of which stores users are loyal to, and which stores their friends should be recommending to them. There are some businesses whose model is most visible in comparison to the accidental knockoffs; Foursquare is one of them.

Editorial Note

The Digital Due Diligence team will be experimenting with some new formats for our newsletter. At about 3,000 words, this is probably too much reading for the average Monday morning; we may split the newsletter into more than one part, or shift some of the content to a different venue. Reader comments and suggestions are always welcome (just reply to this email).

Comments are closed.

Share Let others know too.
Apple Is the Macro Situation