Investing social network Roboinvest now allows users to copy one another’s trades with a single click. They’ve prudently limited this to small dollar values, but it’s still a dangerous plan: market swings already tend to happen when investors overestimate the information content of others’ trades; letting them do this by default seems even more dangerous.

Reed Hastings: Comcast Violates Net Neutrality

Netflix’s CEO argues that Comcast is violating net neutrality by not applying bandwidth from their Xfininty app to their bandwidth cap. This is hard to buy as a serious complaint. Netflix, for example, doesn’t let other streaming services piggyback on its own recommendation engine—it’s a valuable asset as long as it isn’t shared.

Meanwhile, Netflix announced that it didn’t use its million-dollar ranking algorithm improvement, though it did use some of the runners-up.

Could Ad Targeting Hurt the “Conference” Business Model?

Many online trade magazines (you might also call them “tech blogs”) make a significant fraction of their revenue from hosting conferences. The basic business model of a conference is to take the set of people who all care about the site’s topic and are willing to spend a couple thousand dollars to interact with one another—and then charge them a couple thousand dollars to interact with one another. Add some extra fees for vendors, while you’re at it, and the conference business starts to make sense.

But it only makes sense if it’s hard to target these audiences in other ways. As Scout Analytics argues, you can reverse-engineer the entire conversion funnel for conference attendance. As they don’t proceed to argue, you could then target prospective conference attendees at an order-of-magnitude lower CPM. While this doesn’t destroy the economic rationale for conferences, it does erode their pricing power. In the case of TechCrunch, they were bought based on their strategic value to AOL as a media property. But the “cover bidder” was that they could stay private and reap dividends from their conference. So this could negatively impact valuations for business-focused blog networks.

LinkedIn Launches Follower Targeting

LinkedIn’s ad product has added a few new options, including targeting people who follow a particular company. VMeanwhile, the number of people who use this tool is staggering: IBM, for example, has 6X as many LinkedIn followers as Facebook followers.

Google Updates

• Do searchers care about Google+ integration? Many angry nerds think so, but most searchers claim that they haven’t noticed, and those who have noticed are evenly divided, pro and con.
• N=1, but this Google+ user is amazed at how bad Google+ is at serving its intended purpose. For consumer tech, stories about how a user set up the product for his mom should be the kiss of death.
• Danny Sullivan demands real user data on Google+, arguing that Google’s statistics are largely meaningless. Which is true. Their explanation of their usage numbers is incredibly evasive.
• RKG has some always-interesting stats on click distribution and cost per click across different Google products. One major highlight is Google’s dominance in mobile clicks (and their consistent growth).
• Google has run the numbers on customer service for AdSense, and found that it doesn’t pay.

Facebook Updates

• Facebook ads have higher click-throughs on weekends. Facebook users are more boring on weekends (or rather, they’re doing the stuff they’ll upload pictures of on Monday), and some advertisers aren’t clever enough to cut bids when there’s low traffic. This anomaly should fade.
• Facebook is re-launching offers, which now look far more Facebooky than their previous iteration. No effect on revenue just yet.

Smartphone User Adoption

Asymco argues that smartphones have had faster adoption than any other consumer technology in history. This is, of course, fairly dependent on what you call a smartphone, and when you think they started becoming available. But it is striking how quickly smartphones became ubiquitous. In part, because most of what they did was a seamless replication of what other devices did—they handled calls and texts, just like a phone; they did websites and games, just like a desktop. So smartphones haven’t had to overcome anyone else’s network effects, the way e.g. radio and VCR (other rapidly adopted technologies) did.

Meanwhile, cell phone sales are not holding up well.

What is Amazon’s Actual eBook Strategy?

Charles Stross, a prescient science fiction writer who happens to be on the front lines of this sort of thing, argues that Amazon wants to create a monopoly with respect to selling ebooks to readers, and a monospony with respect to buying them from writers. Where his critique falls flat:

Monopolies suck for their customers because they don’t have to give a shit about product quality or price: they have you, the customer, over a barrel with nowhere else to go.

Monopolies exist on a continuum, since “products” are not discrete categories—if Amazon had complete control over the written word, they’d still have to compete with television and conversation. And Amazon is clearly willing to ratchet up quality as long as it increases purchases—their ideal customer is someone who gets in the habit of clicking the “buy” button, not someone who does so reluctantly after scouring the web for alternative options and finding that everyone but Amazon comes up short.

Amazon is in a position to create this monopoly because of their earlier monopoly on print books—not a monopoly in the sense of 100% market share, but a monopoly in the sense that the most useful strategic planning that anyone else could do in the industry was to figure out what Jeff Bezos was going to do to them, and then determine how to minimize the damage. It’s unclear that this monopoly hurt readers—of whom there are more than ever.

In other Amazon news, they’re now offering internal search. That’s a significant move: Google uses site-search as a data source; if Amazon can accumulate search data (from people who are already happy to pay them for cloud computing), that will make their internal search more effective, and mean that they can consider launching another consumer-facing search engine (like their ill-fated A9 project).

And More on Monopolies…

Peter Thiel is on a mission to rehabilitate monopolies. He’s currently teaching a Stanford class on startups (while, apparently, running a US VC fund, a New Zealand VC fund, and running a hedge fund that now makes VC investments).

One student in this class is taking what are apparently verbatim notes: here’s the most recent lecture. One key point: monopolies are the only way that startups become successful, and they’re more socially useful than people give them credit for. This echoes some startup advice that Paul Graham gives in one essay: that startups should address hard problems, not because they’re hard but because everyone else will give up in advance.

In legal terms, it might eventually make sense to view startups as a clever way to game anti-trust law: as long as your business is so new and confusing that nobody at the DoJ understands what you do, they can’t understand that you have a monopoly on it.

Fred Wilson: Leveraged Recaps as an Exit Strategy

Fred Wilson has a simply stunning post.

In a way, it’s inevitable: if companies a) achieve success faster than they used to, and b) don’t need to IPO so quickly, it follows that there will be plenty of startups that evolve into mature, predictable companies without ever needing to list their shares. Usually, VCs put pressure on them for an exit, whether that’s through an IPO or an acquisition.

Wilson proposes a third path: leveraged recaps. These are highly contingent on being able to effectively model future returns from starutp-y companies, but that’s an increasingly solvable problem, especially in verticals like digital media and lead generation. But it will substantially change the narrative. Firms that can stay private indefinitely can build a very different culture from firms that need to focus on an exit.

Strangely enough, one great example is HotOrNot, whose managers chose a compensation structure focused on huge bonuses rather than equity.

No Buy Ratings for Yelp

Yelp now has analyst coverage: four different analysts have three different terms for “Hold.”

Yelp, meanwhile, is launching more refined rankings for idiosyncratic searchers, like people of specific ages or dietary tendencies. This is more of a defensive move: none of these demographics are huge, and Yelp only has this kind of data because such people use the site already. But it does help them stay ahead of Google, which will eventually use search customization to deliver the same kinds of content (e.g. when vegetarians search for “pizza near Union Square,” Google will incorporate the vegetarian-friendly recommendations of their disproportionately vegetarian friends).

Patch’s Editor in Chief Quits

Embattled Patch has lost their editor in chief. His exit email is a long justification for a) how wrong the media are about how badly Patch is doing, b) how it’s now doing much better than it used to be, and c) why, despite that, he’s moving on.

TripAdvisor Highlights Friends-of-Friends

TripAdvisor’s new reviews use friends of friends to help users find trusted opinions. Since reviews from friends are the most trusted advertising messages, this could pay off. The immediate effect should show up in their conversion rate from visitor to hotel booker, rather than in a higher visitor count—so the success of this feature will depend on whether or not that is the bottleneck.

Microsoft to Charge for Bing API

Microsoft will now start charging for use of Bing’s search API. DuckDuckGo was, as of fairly recently, a major user. This could slow down their growth.

Groupon Apologetics

A former Groupon employee has a lengthy essay on Techcrunch detailing how unfair he feels their media coverage has been, and how impressive their employees are. One point he makes—which is worth emphasizing—is that the pure “offers” business is just a loss leader for Groupon; if they can become the default direct response marketing and yield-management solution for small businesses, they will earn a massive amount of money. The Yellow Pages still generate leverable cash flow even though they’re technologically backwards; combining that market position with modern technology could be incredibly lucrative.

9GAG Sets a Precedent

All Things D has a brief profile of humor site 9GAG, which makes one important point: “9GAG (the name sounds like “making fun” in Cantonese)…”—9GAG may be the first major Internet brand in the US to originate in China. But it clearly won’t be the last.

One symptom of founders getting stronger relative to VCs is that VCs have been forced to expand their repertoire. YCombinator has an in-house designer, Sequoia has an in-house PR person, and of course many early-stage VCs are informal recruiters (in the sense that they find good employees for their hot startups, and find cushy exits for the early employees of their less hot startups—it’s a brutal business).

Now, Ron Conway is getting attention for his lavish campaign donations to San Francisco politicos. This is an odd side effect of the geographic concentration of many VCs’ portfolios: with enough investees in the Bay area, it is indeed in Conway’s interest to lobby local politicians. But it isn’t working too well. In this case, AirBNB still needs to pay a hotel occupancy tax despite Conway’s efforts.

Google Updates

• Larry page has posted an annual letter on Google’s progress this year. It’s a fairly banal piece, but it’s notable that Google+ still gets top billing. For investors who expect Google to pull back from plus and aim for more immediate monetization, that’s a warning.
• Android is on 50% of US smartphones, according to ComScore. But it monetizes worse for Google than the iOS does, according to some well-crunched numbers from Asymco.
• If Facebook is pursuing search—which they are, but not urgently thanks to Bing—it could be good news for Google since it would theoretically show that other big companies are competing in search. On the other hand, they already have Bing.
• Google has summed up 50 changes to their algorithm in March. Many of them deal with synonyms, specifically toning down Google’s too-smart-for-its-own-good synonym matching. This won’t have a measurable effect, but could pull early adopters away from alternative search engines, which will pay dividends in the future.

Facebook Updates

• At last, ad benchmarks: Social Fresh has released a survey pegging Facebook’s average CPC at $.80, and average CPM at .041%.
• After the new fan page layout, Facebook engagement for brands has universally dropped. This is expected, and good news for Facebook; they’ve successfully found a way to charge for what they used to give away.
• Business Insider Intelligence surveys Facebook advertisers to figure out what they hate about the site. Their big issues: poor APIs, uncertain data ownership, and slow analytics.

Is Advertising too Fragmented?

Terry Kawaja argues that there are too many adtech companies each stage of the ad sales process. (And he certainly has good financial reasons to think so: he gets paid basically every time anyone decides to reduce inefficiency in the ad market by buying somebody out.)

It’s hard to make the case that there are too many companies in this space, though; startup costs are still fairly low, and rapidly-declining transaction costs mean that it’s more affordable to add layers to each transaction, as long as each layer is adding at least some sliver of value. Given how automated the business has gotten, we could easily have a situation where half a dozen large display ad companies are replaced with dozens or hundreds of smaller players which have parcelled out tiny pieces of the full chain of transactions. That outcome makes sense given that costs are getting lower and coordination is getting easier.

And it’s not awful news for investment bankers in the space. There’s still plenty of room for scale acquisitions (#1 player buying the #2 player in an otherwise fragmented business) or talent acquisitions (treating the company as an extended audition for the next gig).

How Do Groupon Deals Affect Yelp Ratings?

Finally, we have data! Groupon deals tend to hurt Yelp ratings, especially immediately after a deal is launched, and close to the deadline for using the deal. The study has more details on exactly why this happens.

HomeAway’s Travel Roll-Up Continues

HomeAway has purchased a European vacation home rental site for roughly $19mm. As Paul Graham pointed out about AirBNB in 2009, Europe can be a great market for vacation home rental: “Just wait till all the 10-room pensiones in Rome discover this site.” And, of course, there are natural economies of scale for a site like this: it shouldn’t be too hard to guess which HomeAway users would consider a European vacation. Add in airline deals (also easier at HomeAway scale than at this site’s scale) and it’s even easier to tweak subscribers into making one more purchase.

Why Doesn’t the Philippines Sleep?

Based on ODesk data, the Philippines has a comparatively even sleep/wake cycle compared to other countries. It’s a place with a large English-speaking population and a fairly weak local market, so this makes some sense. More than a few SEO-driven companies have either a) Philippines offices, or b) an unusual amount of traffic to their sites originating in the Philippines.

In three years, a larger population of companies ends up with double the valuation. It’s hard to explain this through market sentiment alone: even if investors are twice as optimistic about the future, or about technology in particular, that kind of optimism should be reflected in much higher levels for other market indices. It’s startups in particular that have benefited, not technology stocks in general.

Why are startup valuations, in particular, so much higher? Because capital costs are steadily approaching zero, and marginal costs are easier than ever to predict.

YCombinator’s original thesis focused on the first point. Paul Graham notoriously pointed out that, for many consumer web startups, starting a company had about as much overhead as being unemployed. In the early days of YCombinator, starting was cheap and scaling was comparatively hard. But now it’s possible to arbitrarily scale up infrastructure through Amazon’s cloud computing offerings. And it’s easy to arbitrarily scale up marketing through AdWords. So once a startup has gotten to the point that “customer lifetime value” and “customer acquisition cost” are meaningful concepts, it’s entirely plausible to scale up by an order of magnitude or more.

And that means there’s more embedded optionality in the first round of funding. Early stage investors aren’t buying x% of a real business: they’re buying x% of a y% chance at a really amazing business. All else being equal, a higher y% and a higher value for “really amazing” will mean that x can drop without blowing up the deal.

Investors tend not to couch their decisions in terms of explicit optionality, but the standard practice among lots of VCs is to “date” aggressively in the early stages, and then marry cautiously later on. Increasing the information content of dates encourages more dating, and higher demand for dates raises the cost of admission. Even if they’re not explicitly running these numbers, it’s hard for trends to work out any differently: investors often chase “the next [startup of the moment].” If AirBNB had started in 2004, “the next AirBNB” would have been orders of magnitude less exciting than AirBNB is today. And AirBNB is exciting not just because it’s an obviously good product, but because it’s easier to model than previous iterations on that business would have been.

In a sense, this is a strong argument against a bubble. If there’s one positive thing you can say about today’s crop of startups, it’s that investors have better information—in terms of aggregate data and in terms of fast access to company-specific numbers—than they did in previous years. If better information leads to higher prices, that’s a sign that startups were excessively undervalued years ago.

In a site like, Pinterest, anything that gets enough fake votes to matter will be visible to real voters. In sites that dynamically reorder stories based on votes, ostensibly spammy content will actually get the vast majority of its votes from legitimate users. It’s hard to create a pattern of spammy behavior that makes something look like the most popular posts that day, but comparatively easy to make it look like one of the most popular posts of the last few minutes.

So Pinterest manipulators don’t push actual spam—you won’t see mortgage refinancing lead-gen rank well, for example. Instead, they’re pushing pseudo-spam: something that could have gotten popular naturally, but is much more likely to thanks to the spammy behavior. Take a post that has a 95% chance of being ignored, and spam it until it’s in the top 5% of that distribution: that can sneak below the intuitive filters people use to fight spam (i.e. “Is this getting way too much attention given its objective quality?”), and actually subverts them through signalling: if someone sees a mediocre-looking post that’s getting popular, they may suspect that they’re missing something, and give it a second positive look.

While this whole process makes sites like Pinterest resistant to the worst kinds of spam, it also makes them vulnerable to manipulation and mediocrity. There isn’t a reliable way to distinguish between mediocre content that got spammed up and mediocre content that lucked out, which means that sites will either irritate their power users through false-positive spam flagging, or end up relentlessly mediocre. That’s not the worst fate—Reddit is doing well—but it’s interesting to consider whether that’s a necessary part of social network ranking algorithms, or something a smart startup can solve.

• Facebook now offers ads optimized for “Objectives” such as likes, installs, or ad clicks. That should bump up the low bids, raising the effective CPMs for their pages. It’s unclear how popular any of these options are, though.
• Zynga’s OMGPop purchase has raised their estimated Daily Average Users by about 25%, according to AppData, which does not appear to adjust for overlap among gaming audiences. That’s something Zynga probably did adjust for when they bought the studio, of course.
• Facebook users are vastly more likely to use privacy settings to hide their friends lists. Ultimately, acceptable levels of privacy are a user interface problem, not a regulatory one.

ScreenLeap: A Threat to Citrix?

YCombinator-funded ScreenLeap offers free screen-sharing across multiple devices. Citrix’s GoToMeeting offers expensive screen sharing across a more limited set of devices. A few other companies offer similar software—Skype, Cisco’s Webex, Adobe Connect—but a company that flattens pricing in the category could be pretty damaging to them.

Most companies in this field have great pricing power because one common use case for screen sharing is enterprise software sales. But a free product gets rid of lots of the frictional costs of switching to something new. For now, ScreenLeap is very much below the radar, but it’s worth tracking monthly searches for it (compared to GoToMeeting, Webex, and Adobe Connect) just in case.

Congress Cites Low Demand for Privacy Laws

Adweek reports. Money quote: “Where is the public outcry for legislation? I haven’t gotten a single letter from anyone back home urging me to pass a privacy bill.” This might not be enough to sway things: the people who do care, care pretty loudly. But it’s fair to say that no one’s election chances will be much improved by taking a strong stand on the issue.

BankRate buys InsuranceAgents.com

BankRate has bought InsuranceAgents, a successful lead-generation site. They actually have an impressive PR record, with relevant mentions from Reuters, Inc., and the 37signals blog. Beyond that, their success has been fairly by-the-book, with lots of middling-quality links from press releases and directories. It’s not an amazing acquisition, but BankRate will do well if they can buy up small companies like this, without the risk of a blowup.

Groupon Revises Earnings

Groupon has made a comparatively minor adjustment to their earnings—they’re raising reserve requirements for returns. It doesn’t affect their operating cash flow, so it’s a weak argument for the theory that Groupon is bound to collapse. It’s a much stronger argument for the theory that Groupon needs to get its act together operationally. Their stock price responded to this news by collapsing to levels not seen since two weeks ago.

SecondMarket Scales Back

SecondMarket has laid off about 10% of their staff now that Facebook has finally halted trading ahead of their IPO. This will leave them with about 135 people, or several times as many as they had the first time commentators got worried by their dependence on Facebook. As before, the lesson appears to be: SecondMarket’s business wouldn’t be where it is today without Facebook, but “where it is today” hardly relies on Facebook. (It is somewhat surprising that trading in Facebook alone could support fifteen full-time people, so this might be a broader fat-trimming exercise.)

In another last-gasp of private Facebook trading, the most recent Sharespost auction ended with Facebook sporting a $103 billion valuation. This is not strictly indicative of the IPO price, though: the question is whether bragging rights influence a trade of $6mm worth of shares more than a rabid retail investor base will influence prices when liquidity is orders of magnitude higher.

Apple’s Bot Crackdown hits iOS Developers

Apple’s crackdown on automatic download bots (designed to manipulate app store rankings) has affected developers: the cost of acquiring users is up month-over-month, but the volume of users acquired has dropped by 6.5% from last month.

TripAdvisor Notes: Vacation Home Rentals Growing

A recent TripAdvisor survey shows that vacation homes are growing in popularity, with 40% of the travelers surveyed saying they stayed in one in the last year. (Last year, exactly 40% planned on doing so, making the survey eerily accurate.)

Demand Media Launches Tech Vertical

Demand Media has launched a new technology vertical, targeting the usual long-tail, semi-monetizable stories they’re known for. (Their top story today: “Best Smartphone Navigation Tools”) Demand does not appear to be using their traditional contributor model. They’ve hired a few high-end contributors, and recycled some content: an article on “How to Choose a Digital Camera,” also one of the ten most popular on the site, was written in 2004. Helpful tips include “If you will only output pictures to a computer monitor (for viewing, Web page use or e-mail), an inexpensive digital camera with a 640-by-480 pixel resolution will provide very satisfactory results.”

Efficient Frontier: CPCs are Flat

Efficient Frontier sees generally flat CPCs for Q1 2012 over Q1 2011. This trend has been going on for a while, though: Google more than makes up for low CPCs through higher search volume.

Yes, Networks on Networks Work

Fred Wilson asks if it’s really possible to build a network entirely within another network. At some point, this turns into a hair-splitting problem. But Facebook is a network built on another network (the Internet), which itself is actually several interoperating layers of networks. So it’s entirely possible. It’s just imprudent, unless the network you’re building on top of is either religious about not competing with you directly, or sclerotic enough that it can’t win if it tries.

Click-Throughs and User Intelligence are Inversely Correlated

SearchQuant compares state-by-state data on intelligence and Facebook ad click-throughs to demonstrate that stupid people click on more ads. This is pretty well-known, but it’s unclear whether this is because of higher accidental clicks, more interest in advertised products, or some other factor. Unfortunately, the plot compared smartness-ranking to CTR-ranking. It would be interesting to compare this to other ad targeting that acts as a proxy for smartness.

Netflix Buys DVD.com

Netflix has bought the DVD.com domain name, in what is hopefully an effort to more seamlessly split their DVD and streaming businesses. This is worth keeping an eye on.

Who at T Rowe Price is Doing These Late-Stage Deals?

DST. Andreessen-Horowitz. And T. Rowe Price? The mutual fund company has always seemed like an odd participant in huge late-stage rounds. PEHub has a quick profile of Henry Ellenbogen, the fund manager who runs these deals.

Mobile Search Ad Click-Throughs: 4.1%

Marin Software has released a clickthrough study on search ads, showing that mobile searches get a 4.1% clickthrough rate, compared to 2.4% for desktop and 3.1% for tablets. But that translates into a far lower conversion rate: 5.2% for desktops, 4.9% for tablets, and 2% for mobile. These numbers also imply that tablet searches are the most valuable on the web (they get ad-driven 1.5 conversions per thousand searches, compared to 12.5 for desktop and .82 for mobile).

The Atlantic channels Paul Graham to claim that this would be generally bad for the web: even though user-level data supports the idea of customizing searches based on the searcher’s behavior, that ends up being a negative externality since we don’t have a common set of data to refer to.

But that begs an important question: should Google be the standard of truth? That sounds like a noble mission, but it also sounds like a special case of what Google is: a way to get answers that you’re looking for. To get answers that you’re looking for, you need two ingredients: an idea of what answers are, and an idea of who you are. Right now, Google knows what the general answers are for any given category of queries; they need to know who users are to put together the rest of the puzzle. That works at a large scale, but even at a smaller scale it’s powerful: Google Docs now uses the web for spell checking.

For anyone who’s used Google long enough, it’s unfortunate that Google is no longer a universal arbiter of truth. But that’s best understood as their wedge product—it’s annoying the way Facebook got annoying to Harvard students once it started letting Yalies in. As it turns out, universally agreed-upon truths are less popular and less commercially viable than individualized answers. (And of course, Google offers un-customized search results for anyone who’s really intested: just log out and add “&pws=0″ to the end of the URL for your search, and you’ll get something close to universal results.)

If Facebook can link up their data across properties—Google+ data for social connections, Youtube and Google Play data for leisure, Doubelclick and AdWords for commerce—they’ll have an unparalleled ability to show people the ad messages they’ll respond to, in a medium they like. As always, the question is whether Google can figure out social faster than Facebook can figure out search. Google’s big bet was that if they figured out the web, they’d have a superior product for users; Facebook has made them worry that they should have figured out users, instead.

Ultimately, a Google that dominated social would be about as useful as a Facbeook that co-opted search.

SpyFu has some amazing stats on which sites most frequently show up in top Google rankings. The usual suspects are well-represented: Amazon owns product searches, Wikipedia owns proper nouns, and a smattering of other obvious sites own various common query categories. Interestingly, Amazon is trending upwards: whether this is due to better SEO or a broader range of products is open to debate.

Meanwhile, Amazon has acquired Kiva Systems, a warehouse logistics company. Amazon has also pushed their local deal service with a $10 for $5 coupon deal. $5 leads for a local ad service is actually quite cheap, especially if the “$5″ consists of money they’ll have to spend at the same parent company, regardless. Amazon remains a few chess-moves behind local deal competitor Groupon, which recently acquired a small business credit card processing company. If Groupon is in the coupon business, that’s a meaningless gesture; if the coupon business was a way to get into the business of using software to fix problems that vex local businesses, it fits right in.

eBay Divests Rent.com

eBay has sold Rent.com to Primedia. While Rent.com is an interesting property, it’s gradually losing ground in search results. Since eBay doesn’t have a strategic focus on this kind of commerce lately, the deal makes sense.

Pew: Social Media is Not For News

Part of the problem with the news business is their obsession with building a better distribution model for news. Facebook and Twitter have built a fantastic system for sharing news, but a new Pew study reveals that most people get news from non-social sites. One easy explanation: this great medium reveals that most people aren’t interested in traditional news—the industry thrived because distribution was broken, not because it worked.

In semi-related news, AOL is launching a new iPad magazine: “Huffington.”

A Data-Driven Approach to Local Rankings

Andrew Shotland has a nice summary of statistically valid Google Places ranking strategies. Some of the heuristics are pretty simplistic—five reviews matter, four don’t—but others are clever. For example, Google Local doesn’t care much about the actual reviews, unless they’re extremely negative. That’s a good way to avoid gaming.

Free 4G from NetZero?

NetZero, the unnaturally hardy budget ISP, has launched some free 4G plans, with rapidly escalating prices for higher bandwidth. NetZero’s parent company, United Online, has a surprising amount of cash on hand, and solid earnings; they’re gradually shrinking, but it’s interesting that they’ve chosen to experiment with this. Their target market isn’t obvious, though; smartphones tend to get sold with data access, and tablets tend to get sold either with data access from the tablet owner’s smartphone account or with Wifi-only data. NetZero’s place in this system isn’t obvious.

Pew: Texting Rises Among Teens

The Pew Research Center reports that texting is up among American teens. That might not be surprising, but it’s down year-over-year in Finland and other markets, which could be a negative harbinger for American markets.

Google Continues Blog Network Crackdown

Google is still shutting down private blog networks (i.e. collections of poorly-written or obviously-spun articles). One blogger offers a list of some of the recently-hit sites, and another tries to explain why Google cracked down. And Warrior Forum has a few posts from panicked link-buyers wondering why their sites no longer rank.

Interestingly enough, blogger AffHelper claims that his site was demoted because a competitor bought links pointing to it. While this is a common hypothetical claim—and typically the main reason some people argue against “outing” sites that use black-hat tactics—this may be the first case in which someone plausibly claims that it actually happened to a specific site. Even so, the economics don’t favor outing sites: no one competitor will exact a big portion of the benefit from this.

Validated Display Ads: Micro-Significant, not Macro-Significant

Forbes is now among the sites using ComScore’s validated Campaign Essentials. ComScore put out a major PR push a few weeks ago, highlighting the fact that many banner ads aren’t loaded, and many of the ads that are loaded aren’t seen. Which is true, but not that interesting since advertisers are ultimately paying for user actions, not ad views: overall, it will mean a lower number of impressions, a commensurately higher CPM, and roughly the same level of spend and ROI. It will mean a slight shift in where ads are deployed, but even that’s easy to overstate: most of the obvious candidates for lower budgets—user-generated sites, sites with pirated content, and adult sites—already attract low CPMs. They certainly won’t “improve ROI for advertisers by approximately 68%,” though.

Google Updates

• Firefox may encrypt all Google searches which would be painful for a large number of sites. This would likely reduce the number of search referral keywords available in any given day by another 25% (accounting for Firefox users’ search proclivities and likelihood of being logged into Google already). While that’s not a complete disaster, it will slow down sites’ ability to adapt to changing search behavior.
• Google’s location-based customization is getting a little too trigger-happy by showing sub-pages for branded queries when the sub-pages have links that mention specific places.
• Google is rolling out yet another Panda update, though there’s no word on whether or not this is connected to the blog network crackdown.
• Google now integrates display ad buying with AdWords, which could encourage more synergistic ad campaigns: using display to influence searches, then capturing the effects of those searches.

Facebook Updates:

• Facebook is promising to fight back against employers demanding Facebook passwords before employment. (In a somewhat sleazy gesture, ZDNet willfully misinterpreted this as a promise that Facebook would sue people who asked for passwords, and ran that as their headline—then got another inflammatory headline out of announcing that this was not, in fact, what Facebook intended to do.)
• Facebook is pricing homepage and logout ads at $710K per day: $550K for the homepage and $160K for the logout screen.
• Facebook has lost its top position on job review site Glassdoor. The new winner? Google.
• Facebook is shrinking ad copy and image sizes on its standard ads, perhaps in an effort to push marketers towards new formats.

How Netflix Uses Its Own Servers And Amazon

ZDNet interviews Netflix’s “Cloud Architect” on how he uses Netflix’s in-house assets in combination with Amazon’s cloud services. It’s largely what one would expect, but underscores a broader point: now that cloud computing has a spot price, it’s possible to distribute calculations over time to in a far more market-aware manner.

Is Apple Really Enterprise?

One outsourced IT service provider says no. Sample error may be an issue, though.

A Skimlinks API

Skimlinks, whose business model we find confusing, has released some APIs. Given what they’re currently doing, this makes sense: as long as they can take a piece of existing transactions, they should insert themselves into those transactions as cheaply as possible.

Zynga Pays $210mm for #1 Game Maker Status

Zynga purchased OMGPOP for about $210 million, based on OMGPOP’s enormously popular “Draw Something” game. “Draw Something” is a huge success—as app store watchers relentlessly note, it’s the #1 free app, the #1 paid app, and the #1 top-grossing app. And it hit that status far faster than other products. $210mm is expensive for a single hit game, but it’s a cheap way to remain the top hitmaker.

DST Raising $1bn

DST, infamous for their very late-stage investments in Facebook, Groupon, Zynga, and Groupon, is raising $1bn. One interesting tidbit: ” DST first invested in the social-networking site in May 2009 and purchased “substantially” every Facebook common share that was traded from September 2009 to February 2010, according to the investor presentation.” This timing implies that DST might have been almost singlehandedly responsible for the growth of SecondMarket’s private company stock market, though DST might not have executed transactions through them.

Monster Tries to Sell Out

Monster’s CEO is interested in selling the company. In what is surely unrelated news, they are trotting out some of their software products to impress unspecified parties. Monster is an interesting asset, but it will be hard to sell: the most optimistic story about it is that its decline will be slower than some investors might anticipate.

What AOL Should do with AIM

AIM has just laid off a substantial number of staffers, and is now running into login issues. It’s obvious that AOL can’t do anything interesting with AIM other than monetize the pageviews. And yet, AIM has a large legacy userbase. The smartest solution: sell it to Bloomberg. AIM and Bloomberg are probably two of the top four communications tools for investment professionals (the other two being phone and email). And while AIM is expensive as a standalone product, it’s cheap as a lead generator. Bloomberg could trivially identify AIM accounts belonging to prospective Bloomberg users, at which point the only question is whether house ads are a better deal than targeted ads from third parties.

Digiday notes that car comparison sites are price setters, unlike nearly everyone else in the online ad world. This seems counterintuitive, especially since it doesn’t apply to sites one level downstream (e.g. eBay Motors) or one level upstream (like auto blogs).

Turntable.fm Signs Record Deals

Turntable.fm has finished one leg of their licensing deal arbitrage by signing deals with major labels. Now that their costs are predictable, all they need is a revenue model.

Can Bing Compete with Google?

Wired has a PR puff piece from Bing arguing that, at a technical level, it can. And this blow-by-blow analysis of Bing and Google’s performance during breaking news stories also indicates that Bing is more willing to push news-driven results into high-volume queries.

It may just be their PR, but Bing appears to be on the right side of an innovator’s dilemma: they have a worse reputation than Google, so they have less to lose if they promote new pages in search results long enough to measure user click behavior. And that means their search results will be more adaptable during breaking stories—which is an indirect subsidy for the technical improvements that Bing cites in the Wired story.

Meanwhile, a smart webmaster who was hit by Google’s various Panda updates argues that they’re more based on user clickthroughs than any other metric.

How Not to Invest in Facebook

Andrew Chen walks through his thought process when he decided Facebook didn’t have great potential as a business based on their metrics when they had about a dozen employees. Every part of his analysis makes sense—it’s actually much more reasonable than what Facebook’s early VCs seemed to think, which was that a) there was a new business model in there somewhere, and b) that Facebook’s growth could continue indefinitely. As it turns out, those fairly irresponsible assumptions were exactly right.

Sharespost, EB, and Felix Investment Charged by SEC; SecondMarket Unscathed

The SEC has looked into the private stock market, and spotted some dubious actions: Sharespost was dinged for failure to register as a broker-dealer (they’ve since fixed that). EB Financial made some more serious mistakes, including buying Facebook stock from an entity controlled by the wife of the manager. And Felix Investments (as one might have predicted) got accused of attempting to sell stock it hadn’t actually acquired from investors—apparently shorting non-public equities is possible, but probably criminal. Felix intends to fight the charges.

SecondMarket’s CEO has a detailed explanation for why they weren’t charged. Their model is much closer to the way a typical exchange would work; the companies that were charged seem to operate more as principals than as agents.

(The SecondMarket answer is an interesting case study for another reason: SecondMarket employees used Quora’s “Credits” system to promote the answer, which is apparently how some media outlets found it. Given that Quora could theoretically charge cash for credits, a business model naturally suggests itself.)

Netflix has Recovered from “Quickster”

PaidContent cites some Citigroup research indicating that Netflix’s usage and brand perception have both largely recovered from their attempts to split the service into separate streaming and DVD companies last year. Consumers have short memories for PR debacles, and long memories for products they actually like.

Android Tablets Growing Faster than iPads

IDC argues that Android will be more popular than the iPad by 2016. But as long as Apple can avoid Android’s device and OS fragmentation, it will remain the default choice for developers for longer than that.

Meanwhile, iPad users have a high lifetime value, at least on design-centric site Fab.com

“How Garbage Ranks: a Case Study”

This is worth reading: it’s a look into how a site ranked near #1 for terms like “Car Insurance” and “Auto Insurance” despite having weak content and few to no brand signals. (The author didn’t publish the piece until Google had demoted the offending site.)

The Most Trusted Tweets

Apparently social connections weigh more heavily than expertise when people evaluate opinions in social media. That’s empirically true, in the sense that a huge fraction of social media activity involves following a known stream of users, instead of browsing for expert opinions. But in some ways, that’s part of the medium: recommendations from social connections are trusted in social media because, when users want expert opinions, they use search instead.

Twitter Buys Posterous, Aimed for Tumblr?

Twitter has acquired Posterous, and will likely shut it down and have its staff focus on Twitter. Ryan Tate claims that Twitter was really hoping to buy Tumblr—which hardly makes sense, given that Posterous is a talent acquisition and Tumblr would be a product buy.

MTV Targets Artist-Name Searches

MTV is using their new .MTV top-level domain for a product targeting artists. This will be an interesting case study in how Google treats branded top-level domains: optimizing thousands of new dot-com sites would be tough, but the .mtv domain is a strong signal that these sites fit under MTV’s umbrella.

FAA May Review Tablet Policy

The FAA may reconsider its rules against using tablets during the beginning and end of most flights. Internet ads are worth roughly $.70 per hour, takeoff and landing take about an hour per flight, and there are about 637 million enplaned passengers per year in the US. So the potential marginal revenue for the US Internet advertising industry is measurable, though not significant—marginal revenue for companies like Boingo Wireless could be a bit higher. And that ignores the fact that more passengers choosing to turn on their tablets in flight means more reasons to target that specific demographic—a hotel offering spa treatment might get far more takers in hour six of a six-hour flight than they would offering the same deal a week earlier.

Is AOL Shooting the Messenger?

They are, at least, laying off many of messenger’s product staff. Since messenger is integrated with Google Chat (and plenty of other chat clients), this shouldn’t have a huge effect on users—but it’s surprising to see AOL give up on this kind of asset. Perhaps DST wasn’t interested in buying another legacy chat network from AOL.

The Internet Bubble: A Rational Tournament?

Albert Wegner makes the case for the bubble: since successes can be so huge, so quickly, it makes sense to value more companies based on the small chance that they’ll suddenly turn into growing businesses with defensible margins. That’s a good first-order explanation, but what happens when investors raise money knowing that this is how VCs think?