The Quiet E-Commerce Renaissance

E-commerce has undergone dramatic change in the last two years.  Zynga has cracked social gaming and virtual goods wide open, producing enormous transaction volume.  Groupon has crushed local service group buying and created a multi-billion dollar cash machine.  Gilt has pioneered invite-only branded high end flash sales.  All three of these companies have pioneered new e-commerce models in a matter of months, have been locked in the tech news spotlight, and have spawned hordes of fast-followers trying to get in on the action.

However outside of the consumer spotlight, there are several companies that have been revolutionizing web based e-commerce tools available for businesses.  Shopify, AvantLink, Shopatron, AdRoll, and FeeFighters are each dedicated to helping their customers sell more and sell better online.  They don’t receive the public fanfare of their consumer facing brethren like Zynga, Groupon and Gilt, but all have offerings that create unique value and some are growing at a similarly furious pace.

Calculating Online Advertising Return on Investment

I’ve been helping a friend write a business school case on his company. I haven’t done too much but it’s been fun to see the process. Recently I helped with an appendix covering the ROI calculation when dealing with online advertising campaigns. I thought I would include it here in case it might be useful for other folks.

Facebook: Reach and Saturation by Country, Part III

I checked in on Facebook’s growth and added to the data set used in Part I and Part II. The data is a bit too big for a spreadsheet so I used iCharts to make a dynamic chart that allows for easy visibility. Use the slider on the side to zoom in on the other 90+ other countries. Notice that their global growth continues to accelerate.

User Value = Demographics + Tolerance

A subject often mulled over by startup founders and investors alike is the value of a user. Web services, just like brick and mortar service providers, try to make money from of those that use a service. Some sites charge users directly, but most rely on selling some form of access to their user base to others. Display advertisements, job postings, and direct messages are the three predominant monetization strategies used today. In the past several months, web services have begun to offer more and more targeted and focused access to their users, a trend that will certainly continue in the social network space in particular as companies begin to focus on monetization.

As this is a popular topic, Arrington recently ran an article entitled “Modeling the Real Market Value of Social Networks“. Ironically, he ignored a large amount of relevant data and instead posited a new way to value social networks…completely disconnected from any and all real market valuations (he missed a few major social networks as well). His end analysis is that demographics are the key factor in determining the value of a user base. I’ve been analyzing valuations, user bases, and user access recently and disagree with this analysis.

In order to examine the true drivers of user value, I aggregated below all of the recent available real market valuations for companies that monetize user profiles. I used a combination of enterprise values, acquisition prices, and pre-money valuations, all of which represent real market valuations. Dividing the number of profiles each company had at valuation by the valuation amount delivers the value the market put on one of that service’s users.

The numbers show that demographics are just part of user profile value. The other driver of user profile value is user tolerance, or what level of access a typical users is willing to endure. Gerson Lehrman (#1) has an extremely small user base in relative terms, but it is a group of professionals (high-end demographics) and, by definition, every user is willing to receive cold calls from Gerson Lehrman clients (perhaps the highest level of access there is). Dice (#3) has only 2M users but all are technology professionals interested in job opportunities (valuable demographics) and all are willing to receive messages or alerts about potential employers (high tolerance to access). In fact, five of the top six on the list allow for direct access to their user bases, while almost all of the bottom half of the group relies entirely on generic, unfocused display advertisements and will likely face much resistance to increased profile access (see: Wikipedia).

Going forward, it’s going to be important for web services to not only enable targeted access, but to also cultivate users that will stick around to receive it.

Company Valuation Derived from… Profiles Value Year GLG $875M Investment 200,000 $4,375 2007 Facebook $15B Investment 50,000,000 $300 2007 Dice $500M Enterprise Value 2,200,000 $227 2008 CareerBuilder $1.55B Investment 24,000,000 $65 2006 Affinity Labs $61M Acquisition 1,000,000 $61 2008 LinkedIn $1B Investment 22,000,000 $45 2008 Geni.com $90M Investment 2,000,000 $45 2007 Twitter $80M Investment 2,100,000 $38 2008 Badoo $300M Investment 10,000,000 $30 2008 Monster $1.88B Enterprise Value 80,000,000 $24 2008 Bebo $850M Acquisition 40,000,000 $21 2008 MySpace $580M Investment 22,000,000 $26 2005 Friends Reunited $208M Acquisition 15,000,000 $14 2005 Fotolog $90M Acquisition 10,000,000 $9 2007 Photobucket $300M Acquisition 40,000,000 $8 2007 Plaxo $150M Acquisition 50,000,000 $3 2008 Tagged $102M Investment 30,000,000 $3 2007 Various $50M Acquisition 18,000,000 $3 2007 BlackPlanet $38M Acquisition 16,500,000 $2 2008

Buzznet $15M Investment 7,000,000 $2 2007

If anyone has hard valuation/user profile data for yelp, myyearbook, orkut, HI5, Ning, classmates.com, friendster, reunion.com, habbo.com, hyves.com, imeem, live-journal, mixi, multiply, netlog, perfspot, skyrock, sonico, stidivz, V Kontakte, wayn, windows live spaces, xanga, xing, or any other companies in this genre please send them on.

Where’s SelfServe by MySpace?

Back in November of ’07, MySpace announced the future launch of “SelfServe by MySpace” which would “allow advertisers to directly purchase, create and analyze the performance of ads throughout the MySpace network.”  It was supposed to be in beta for two months then launched in early ’08.  It didn’t happen and ClickZ reported last month that SelfServe may launch in later summer or early fall.  MySpace is being left in the dust by Facebook Social Ads and LinkedIn DirectAds.  Now Orkut, Hi5, Bebo, Ning and the others need to step up to the targeted advertising plate as well.

LinkedIn quietly launches Research Network and DirectAds…let the monetization begin.

LinkedIn DirectAds
LinkedIn has quietly launched a beta version of a dynamic CPM text advertising platform called LinkedIn DirectAds. No formal announcement of the launch was made on the LinkedIn blog or elsewhere. According to the DIrectAds FAQ, advertisers will be able to dynamically target ads by age, gender, geography, educational institution, industry, and seniority. Minimum order size for an advertisement is $25, with the minimum number of impressions dependant on the targeting audience chosen by the advertiser. The rate that you pay for a CPM (1000 impressions) changes as you add or remove targeting options from your ad. Apparently the product will give click-through rates to advertisers, but billing will be based on CPM. In a unique twist, ads will also include the advertisers name and a link to their LinkedIn profile in hopes of “increasing transparency and visibility into the advertiser.” Much like the Facebook SocialAds platform, advertisers must have a profile on the network to launch an ad, although LinkedIn says they are limiting advertisers by completeness of profile, number of connections, date of profile creation and a number of other factors. I was unable to access the platform through my profile.

The DirectAds platform will bring LinkedIn closer to Facebook’s Social Ads technology, with these two leaving Bebo, MySpace, Plaxo, Friendster and the rest of the social networking world behind for now. I hope to be able to try the LinkedIn platform soon and give a head-to-head comparison. LinkedIn will continue to extract a premium on their advertising, as it seems they will be setting the price per CPM internally. A true market (e.g. Overture/pre-Panama Yahoo) or partial market (e.g. Google quality score) influence on price would likely result in prices lower than they would like, and they are clearly avoiding a CPC model for a reason since they are measuring CTR anyways. I think this slow transfer is very smart on their end especially considering their pre-IPO status, but as an advertiser I wish they would switch to a free market faster. Their ad margins will likely be lower than what they were getting with their rate card (although perhaps not), but the volume of advertisements will definitely spike upwards as you no longer have to go through a traditional advertising salesperson process to launch a targeted ad on their network.

LinkedIn Research Network
Additionally, on Thursday of last week LinkedIn quietly launched the LinkedIn Research Network, a product the company first mentioned back in February. No formal announcement of the actual launch was made on the LinkedIn blog or anywhere else, but the Research Network product page is live and linked to from the Premium Product footer, along with job, corporate, and upgrade links. Also linked is a 2 page product summary PDF. The product page outlines what is essentially a premium version of InMail (pdf). A Research Network subscriber can send send 20 InMails at once, and no monthly or daily limits are mentioned. Previously, LinkedIn BusinessPlus subscribers had the most InMail access and were limited to 10 InMails per month, so this is a dramatic increase in potential InMail volume. In the past advertisers could send targeted InMail blasts through LinkedIn’s advertising platform at $1 – $5 per recipient.

The LinkedIn Research Network is an attempt to move into the expert network industry and will be sold primarily to hedge, private equity and venture funds. According to a recent Integrity Research Associates report, there are roughly 25 expert networks in existence today. Aside from my company KnowledgeBid, every other expert network service operates on a subscription model. LinkedIn is likely gunning for the fat subscription fees that players like the Gerson Lehrman Group are pulling from investors (+$50k for access to one industry vertical of experts for 6 months), but the product they have launched is far more like the resume search/direct email services offered by Monster, HotJobs, CareerBuilder, Dice, etc. than an expert network. Perhaps down the road LinkedIn will try to facilitate the actual expert matching, but this iteration of the product just enables subscribers to send a large volume of cold emails to potential consultants. Additionally, the product page makes no mention of facilitating consultant payment and the only compliance functionality mentioned is a “research history”. Legal compliance is arguably the largest issue faced by expert networks today, and something that expert network users have come to expert from service providers. It’s possible that LinkedIn is intentionally not involving themselves with payment of experts in an attempt to remove themselves from the chain of liability if their service were to be used to facilitate insider trading or the like.

Congrats to LinkedIn on the product launches. I’m glad to see them competing with Facebook on the advertising technology side of things (let’s see an API guys!) and will certainly be keeping tabs on these products as they mature.

Facebook Ads now targeting professional titles, taking LinkedIn head on…

Facebook continues to quickly and quietly improve their advertising platform. In yet another innovation launched without formal announcement, Facebook now allows advertisers to target ads based on professional titles in user profiles. Previously ads could only target by keywords listed in users’ “interests” field. Now Facebook has indexed professional titles and allows for dynamic targeting through the Facebook Ads platform. Perhaps the recent launch of the advertising feedback function was in anticipation of an aggressive move towards monetization via heightened ad targeting? We’re still waiting on the Facebook Ads API but LinkedIn is still using a massively inflated CPM rate card despite their recent $1B valuation and and MySpace still uses Google adwords, putting Facebook miles ahead of the rest of the social networking pack when it comes to advertising technology.

Facebook quietly launches advertising feedback

The Facebook advertising platform continues to advance ahead of the rest of the social network pack. We’re still waiting on the API, but they’ve recently snuck out a feature that allows users to indicate whether they like or dislike an ad served up to them. Where previously there was just a link for “more ads”, there are now StumbleUpon style thumbs. Clicking on one of them pops up a window with feedback options. Screenshots below. The fact that Facebook is implementing these kinds of features before they launch an Ads API shows that they are approaching mass advertising very carefully. They know that they need users to make ads have value, and the better the ads are the more valuable their ad space will be. Also, it’s quite possible that having some interaction with ads beyond just clicking them will incentivize users to click more ads. The Facebook advertising platform continues become more and more interesting.

Facebook: Reach and Saturation by Country, Part II

Updated here.

Roughly six months ago I posted some information that I dug out of Facebook’s then just launched Facebook Flyer Pro advertising platform. I’ve been poking around with Facebook again the last few days and recent talk about Facebook’s valuation inspired me to update my report. Facebook has improved their advertising platform, now just called Facebook Ads, and while ad creation is still manual, they offer CPC payment and social features on top of the incredible targeting they had with Flyer Pro. They still also display the number of users that an ad will target, so messing around with the UI can give you some interesting information about Facebook’s user base. Here I’ve added current data to the spreadsheet I started six months ago. In short, Facebook has added ~28 million users in the last 6 months, growing at an average 140% in the countries where they released data six months ago and today. 70% of their growth was in North America and Europe. I updated population figures so that saturation percentages would be accurate.

Updated saturation leaderboard:
Canada: 28.22%
Norway: 24.04%
UK: 16.88%
Australia: 13.30%
Sweden: 12.20%
Denmark: 10.73%

Trailing 6 month growth leaders:
Turkey: 875%
Israel: 393%
Colombia: 393%
France 294%
Malaysia: 255%
Switzerland: 199%

Imperia: VoIP app backbone

Much of the recent buzz (two more) about VoIP app innovation seems to be brewing around Iperia. Iperia’s website is pretty tough to decipher, but it looks like they provide infrastructure for trackable calling and other VoIP stuff. They gave a demo of a potential use of their product at the VON 2007 fall conference which was for realtors to track calls to the homes they had on the market. Much of the criticism about the Iperia is focusing on the app for realtors, not the actual products Iperia provides. I’ve written about the benefits of trackable calls for local advertising. It seems Iperia is positioning themselves to be a service provider that can step in and help application builders integrate VoIP features. I think that VoIP features will continue to be grow, but Iperia will be fighting against the tide as the technology behind them becoming easier and easier to implement.

Facebook: Reach and Saturation by Country

Updated here.

Facebook’s recently launched Flyer Pro advertising platform offers a phenomenal, albeit manual, mechanism for targeting advertisements on Facebook.  It also shows some interesting statistics about Facebook’s reach. I ran some quick numbers to see which countries were most saturated with Facebook’s +48M users. The winners:

Canada: 22.12%
Norway: 18.68%
UK: 10.58%
Sweden: 9.09%
Australia: 6.64%
United States: 6.57%

Blended iPhone

This is going to be all over the place today and I saw it on TechCrunch, but it’s too funny to not post it.

100 billion cookies

Seth Goldstein posted, in his usual bizarre fashion, about the billions of cookies that lie behind the billions of dollars which make up the recent ad network acquisitions. Seth’s venture AttentionSoft is trying to empower users to take control of their “attention” (embodied in cookies) because, as these acquisitions show, their attention is very valuable.

While I admire Seth’s unique take on the market for attention, I think the more interesting issue in play here is the fact that these networks thrive on the fragile world of cookies. Banner advertisements generally have miserably low CTRs, although the networks that serve banners argue that they can increase clicks by “targeting” ads served to users. This targeting is done through cookies. If a user navigates to a site that contains a banner ad from one of these networks, the ad “sees” the cookies in the users browser, and if it recognizes any of them it can use that data to serve up a relevant ad. For example, if a user is shopping for a car, he may visit several car manufacturer sites. If Ford is using an ad network, they may try to plant a cookie in the users browser so that when the user navigates to a site that serves ads from the network they run ads in (NYTimes.com, whatever), the site will see that the user has been to the Ford site recently and will serve up an ad for Ford. Additionally, some ads themselves can plant cookies. These ads analyze the content of the page a user has navigated to and plant a cookie that contains information about the ad and the context the user saw it in. You likely have hundreds of cookies in your browser right now – to see them go to Preferences and then look under Privacy or something similar. Some of these cookies are useful and enable autofill or autologin features you likely agreed to, but the majority are likely not.

This system hinges on the cookies making it into the users browser in the first place. Different browsers have different default options for cookie acceptance. Firefox currently only has one option: “Accept cookies from sites.” A user with this option selected will see targeted advertising as they navigate around the web while someone who doesn’t select it will see no targeted advertising. Safari’s options are: “Always,” “Never,” and “Only from sites you navigate to (for example, not from advertisers on those sites) .” Internet Explorer has a scope of options from “Low” to “High” to “Block All Cookies.”

The way in which these options are framed has a large impact on how users choose between them (i.e.: the fact that the “Block All Cookies” is past the “High” security setting may make a user feel like an extremist for selecting it). Similarly, the default security setting on a browser like Internet Explorer, which tends to be used by less savy web surfers, will have a huge impact on how cookies are logged. Now that Microsoft has acquired aQuantive, which relies heavily on cookies to target ads, you can bet that the default IE option will be to allow cookies.

Comscore has done lots of interesting research on user cookie deletion from the perspective of traffic measurement systems. Their study states that 3 in 10 internet users delete their cookies every month. This seems quite high to me. At the end of the day, for the vast majority of internet users the default configuration of a browser has the most impact on cookie acceptance and retention, and with browser makers moving into the advertising world, you can bet that default browser settings are going to become more and more cookie friendly.