- Google’s Q4 earnings disappointed investors—they had higher ad clicks and lower costs per click (for the first time in a while). It’s hard to understand exactly what happened here, but John Ebbert at Ad Exchanger has an interesting theory: Google may be funneling traffic to PPC advertisers whose cookies can be used by Google’s display busines. Essentially, they’re offering a subsidy to keep users in the display ecosystem. It’s hard to put a net present value on one more cookied click, but the number isn’t high. The strategic value of having the best intent-level information among all display ad markets is extremely high.
- One throwaway line during the earnings call led to much speculation on how many users Google+ has. While it’s not fair to argue that Google+ is a success based on what the numbers could be, it’s worth noting that with Google+ baked into so many other Google products, the numbers will only get more confusing. Facebook’s daily login stat doesn’t include people who use Bing data influenced by Facebook, for example—and if Bing had Google’s search market share, that would distort the issue even more.
- Google has launched an algorithm update focused on page layout, specifically penalizing pages with excessive above-the-fold ads. This was one of the signals they seemed to use in the first Panda update, so some of the victims may fall into the Panda-survivors category. Search Engine Land’s Danny Sullivan has many details.
- A subtle competitive advantage at Google: they optimize their workplace to make it easy to stick around. Some of the highest incremental ROIs in the world consist of tiny perks that convince smart people to stay at work a little bit longer. (This might actually be one underlying cause of widening income inequality: the high bidder for a smart person’s time may be the company that knows they can use more if it; as long as more employers duplicate these strategies, the number of hours worked by top earners should increase. Why blame Bush when you can blame Seamless.)
- Google is running PSAs about data protection and privacy. As we’ve argued before, Google captures a pretty big piece of the upside from campaigns like this (more clicks, more spending), meaning that they have a strong economic incentive to look out for the public interest in these cases.
- Google is selling search engine data to banks to help them make better macroeconomic forecasts. If you’re in the market for similar data—without using private search data—contact us.
- Google is asking searchers to use Google+ for questions that aren’t answered by search queries. This has been a pretty obvious search play for a while—Q&A sites tend to backfill search for more obscure queries. While this might increase Google+ use, the big effect could be on Quora.
- Search Engine Land has a detailed interview with a Search Quality Rater. The main upshot of this interview is that they’re adding subjective judgments and details to what Google already does algorithmically (i.e. measure user reaction to a page or set of search results, and use that to re-rank things). It’s hard to see why this job is still necessary. One possibility: they use search raters when click-data could indicate either a) users disliking search rankings, or b) manipulative clicks.
- Facebook is allegedly negotiating with Vevo to replace Youtube as Vevo’s viewing platform. The only good reason to do this is to game ComScore numbers and become the web’s biggest video destination. Facebook tends to avoid owned-and-operated content; it’s hard to stay lean without doing so, and it makes it far easier for them to eventually eviscerate content owners’ business models.
- Digg is a great example: Facebook auto-sharing traffic is keeping them alive—why would Facebook prefer to own the destination content, rather than being able to replace them with Reddit later on?
- Facebook effectively subsidizes ads that point to other Facebook pages. Entirely sensible: the exit rate on any given Facebook page is low enough that internal clicks likely lead to ‘net present impressions’ that justify the discount, even outside of Facebook’s broader interest in controlling more of the conversion funnel.
- There’s another study of Facebook’s average CPC by market and overall during 2011.
- Facebook is offering ad credits to businesses that get more fans. Perhaps they’ll adopt other parts of the Google playbook—training people to game their algorithm in ways that make the overall FB experience more enjoyable, rather than doing so at the expense of the broader Facebook ecosystem.
- Facebook has dethroned Orkut. Inevitable.
State of Search Marketing 4Q 2011
RKG has released a comprehensive study of the SEM, SEO, and Facebook markets in the last quarter. Many of their predictions were borne out in Google’s latest earnings report (more clicks and lower revenue, for example).
eMarketer has also released a report, projecting ad market growth over the next five years. It’s basically designed around the headline that online ads will surpass print this year, and come within striking distance of TV a few years later.
And Adobe’s data indicate that tablet users are big spenders.
RIM Drops Both CEOs
Research in Motion’s co-CEOs have resigned; their successor has announced that the company does not need to change strategy. He may be right; they are likely beyond the point at which they’re valuable as anything other than a de facto asset play.
Publishers Expect Amazon to Kill Them
Pando Daily has published a somewhat insidery email from a publisher, detailing Amazon’s plans to outbid them for good authors and undercut them on costs, wiping out the industry. This is another situation where unit economics and capital costs have both shifted in a way that hurts middlemen economically. The difference between this and the SOPA debate: publishers don’t have that much political clout, and Amazon doesn’t care to use theirs.
An Ads’-Eye-View of The Economist
DigiDay has a display ad-focused interview with the publisher of The Economist Online. One enlightening bit: “In the breaking-news business you are competing not just against other publishers, but also against almost every other person on the Internet who is able to be a mini-publisher on their own… The business that we’re in, analysis and curation, is different. It has become much more valuable in a setting where there are seemingly infinite content choices.” (And that’s why Digital DD doesn’t break news.)
The day after SOPA protests led to the bill getting pulled, authorities shut down Megaupload, a massive file-sharing site. Megaupload illustrates the need for new regulations: the site had many legitimate users, but derived a decent fraction of its utility and viral coefficient from widely available pirated content. Now legitimate users are stranded, and it’s taken arrests rather than fines to shut down the site.
Group Buying Turmoil Continues
Discount luxury site Lot18 is laying off 15% of its workforce, just a week after Gilt announced similar layoffs. In both cases, the companies are still growing, just reorganizing.
And Daily Deal Media has released a report arguing that a) there’s high turnover in daily deal sites, b) in the US and Asia, the total number of players has dropped in the last six months, and c) users are increasingly likely to be happy with their deals (while merchants are dubious). This fits into our broad thesis on the market: it’s far harder than it looks, and Groupon’s business is commensurately more durable than outside observers might assume.
Impression Asymmetries Make an Impression
ComScore is on a roll: they managed to launch a study on how many ads get paid for but not seen, a product for solving that problem, and a series of interviews with people who deal with the issue, all in the same week. Kudos to their PR team.
The missed-impression problem seems to be self-correcting in a broad sense: it’s not as if ad prices are based on the intrinsic value of an impression; they’re based on the ROI from buying ad views. If we were overestimating the number of people who saw ads, we were commensurately underestimating the percentage of people who bought because they’d seen an ad. The interesting question is which publishers are disproportionately likely to run these phantom ads (and that, of course, is what ComScore is helping to figure out).
What Will SecondMarket Do Without Facebook?
The WSJ wants to know. This is a good question, but it’s one SecondMarket has likely been aware of since they first started facilitating trades in Facebook stock. And it’s obviously a nice problem to have: the popularity of Facebook as an investment has allowed them to build a much broader market, which is surely ahead of where it would be without Facebook.
In fact, SecondMarket has essentially smoothed the continuum between private and public companies—owning that franchise will be enormously valuable over time.
Betabeat has a detailed interview with Jonah Peretti at Buzzfeed. Despite the generally cheesy content, Buzzfeed is worth following; they’re building a new model entirely around a distribution channel. And given how many media sub-industries are literally named after the distribution channel, this could be a big deal: there are “Print” and “Cable” news companies, but you don’t hear about e.g. “Truck” and “Train” clothing companies.
Digital Due Diligence Weekly