MetaMuse

Current State of Web Privacy, Data Collection, and Information Sharing

Very interesting research report at KnowPrivacy.org on the current state of web privacy, data collection, and information sharing. The project was to compare users' expectations of privacy online and the data collection practices of web sites, identify specific practices that may be harmful or deceptive and attract the attention of government regulators, and to produce recommendations for policymakers.

The key takeaways for me were:

• Users are concerned about data collection online (duh!), want greater control over their personal information, yet lack the awareness or initiative to do anything about it. (I found it interesting that the report seemed to focus on personally-identifiable information (PII) and not distinguish that from non-PII.)
• Web bugs/beacons are ubiquitous.  All of the top 50 websites contained at least one web bug at some point in a one month time period. Some had as many as 100.
• Google is the dominant player in the tracking market; it operates the top three trackers and four of the top 10. Among the top 100 websites this project focused on, Google Analytics appeared on 81 of them. When combined with the other trackers it operates (AdSense, DoubleClick, FriendConnect, etc.), Google was on 92 of the top 100 websites and 348,059 of 393,829 distinct domains reviewed -- that's 88.4% reach across the Web!!
• Most of the top 50 websites collect information about users and use it for customized advertising.

Other various data points and comments I noted:

• The number of user complaints made to the various organizations is extremely low relative to the number of Internet users. The FTC had only 6,713 for five years (in the General Privacy category), the PRC had 2,202 for the same period and the COPP had 1,152. TRUSTe had 7,041 that it categorized as privacy related. The largest numbers of complaints at all four of the institutions we received data from were concerned with public displays of personally-identifiable information.
• Only 23 of the top 50 affirmatively stated that users could have access to some portion of the information the website had collected about them. The remaining 27 policies lacked mention of access or their statements about access were unclear. None of them explicitly offered users the ability to view or delete click stream data.
• The Network Advertising Initiative (NAI) currently has an opt-out mechanism that requires users to download a cookie, which will let direct advertisers know not to install any third-party tracking cookies on the user‘s computer. This method of opt-out is unacceptable. First, it only governs members of the NAI; tracking companies that are not members will still be able to use cookies and web bugs to collect data about users. Second, users that delete cookies on their machine may delete the NAI cookie inadvertently and open up their machine to third-party tracking again.
• Only 27 of the top 100 Web sites provided a P3P policy, and only a subset of those were valid according to the P3P standard.

The final recommendations as a result of the research?

• Regulation by which both websites and third-party trackers must allow users to see all the data that has been collected about them, not just user-provided information. Additionally, users should also be allowed to see with whom their data has been shared.
• That companies request permission from users before sharing data about them with any outside party, regardless of affiliation.
• Privacy policies should be readable for average users.
• Users be given clear and proper notice as to whom the data will be passed, regardless of affiliation or method of sharing.
• That the practice of third-party tracking be made more transparent.
• That the FTC create an opt-in standard for enhancement -- the practice of buying information about users from outside sources.
• That all browser developers provide a Ghostery-like function in their browsers that alerts users to the presence of third-party trackers.

Increasing Social Network Ad Revenues

Ever since I heard about the MySpace HyperTargeting program last December, I've been looking for information about how it's been performing.

The theory, which we're proving to Others Online partners and which HyperTargeting is testing, is that ads in social network situations are better targeted not by the context of the page you're on, but on the content you write within your own profile/pages.

Looks like they're reporting 300% increase in clicks in initial tests.

UPDATE: Found a little more information on how MySpace HyperTargeting works and how they're charging. Looks like they are segmenting users into proprietary categories based on the information in their profiles, then selling those categories on a CPM basis at a 50% premium.

Affinity vs. Engagement in Advertising

Others Online has been working on some interesting reports for our partners to summarize in aggregate form what their users care about. As a matter of labeling, we've discussed the term "affinity" vs "engagement". The latter has been thrown around the online advertising space quite a bit recently, so I did some poking around the Web to get a better feel for how it's being used. Usually, going with some convention in labeling in more advantageous, rather than creating your own definitions (which markets get annoyed with IMO).

Stumbled upon an interesting and intelligent blog post on How the New Engagement Metrics Can Impact Advertising Decisions. In it, "engagement" was defined as the nature of visitors’ relationship with a site and how that is expressed in the full range of user interaction, involvement and connection. Kevin then went on to describe categories of engagement that effectively describe the kinds of engagement illuminating and differentiating Web sites for advertisers: loyalty, recency, click depth, interactivity, duration, and subscription. I should also point out that Kevin's work leveraged the work of Eric Peterson. Kevin then wrote:

How  can publishers collaborate with advertisers to yield actionable intelligence?

“We’ve developed Web technology to the point where we have an astounding wealth of data about audiences. Publishers can tell us what content audiences are consuming and the share of content downloads among competing advertisers. All this has been great. But what does it all mean? How can we turn that information into something we can act on?”

Brandon Starkoff, Vice President/Global Director at Starcom Worldwide

Starkoff’s point is critical to the whole point of seeking to establish a definition and a set of metrics for engagement. What does it matter if, as far as media companies are concerned, it doesn’t produce better insights into what will make advertisers successful with even the most “engaged” audiences? The kind of audience knowledge Starkoff says he is seeking is “predictive intelligence-advice on what kinds of advertising will work with a particular audience or audience segment.”

Advertisers right now think about engagement as a way to distinguish sites from each other.

It was in the last sentence that I realized what I felt was wrong with their line of thinking. It's not that advertisers need to distinguish sites from each other, but rather people from each other. Engagement is a term being applied as an attribute of a site, not a person. That's just wrong IMO. That's online advertising 1.0, not 2.0. As long as the industry continues to think in terms of targeting sites and not people, there will only be small incremental improvements to online advertising performance.

Affinity is a term which we feel best describes attributes of people, and what we actually care about. Since Others Online is all about targeting people and not sites/pages (based on our understanding of what they care about), that's clearly the label we need to use.

Attention Data and User Value

I believe most people on the Web are lazy, and would prefer services to read their minds instead of requiring them to expend more effort. I'm talking about most uses of the Web. This is why social bookmarking and tagging in general isn't mainstream -- at some point, the explicit effort gets to be too much work. That's why I believe so strongly in the power of the Implicit Web, and using Attention Data as the backbone of Web services (in addition to explicit user preferences/indications).

I just read a post from Marshall at R/WW on different things you can do to "fall in love with tagging again" -- in general providing good ideas for how to tag content for improved productivity/efficiency/communication. He starts off saying that he stopped using social bookmarking services last year because of the cost/benefit (cost being the effort required). Ok Marshall, I'm with you so far. His 5 useful ways to use tagging weren't very interesting to me, though I can see how it would be for some content junkies.

His "reason 5 1/2" I thought was weak though. In it, Marshall discusses the future where we'll be able to influence our attention profiles through tagging. I can't disagree with that, but I thought his example was thoroughly uninspiring. Marshall is very smart and tuned in, so I suspect he just lobbed the example out there instead of applying much thought to it. So I thought I would provide some examples ...

Ok first of all, tagging is inherently an explicit action -- an extra step -- and therefore (in my mind) just one small way to influence your attention profile. I believe your existing actions, such as reading content, searching the Web, publishing, shopping, etc. represent a much more powerful way, since it's more thorough and doesn't require any incremental user effort.

So let's assume we can leverage explicit and implicit user behavior within an attention profile. And let's also assume we have many many attention profiles and an infrastructure/ecosystem in place. Here are a few more examples of user benefits, which I think are more compelling that the "African photoblogging" example Marshall gave:

• Connections to people -- what if I was able to connect with people who were most relevant to the Google search I just did, or the camera I'm looking at? From within an "attention warehouse", one could even introduce you to someone who owns that camera, ALSO uses it primarily for outdoor adventure photography, and whose blog displays many pictures.
• Connections to content -- what if you could connect with the content being viewed and/or published by groups of people who indicate the same level of attention in certain areas? You could view the Web through a filtered lens of those with an attention affinity to "Barack Obama, Seattle, readwriteweb.com", and even do keyword searches of content viewed/published within that congregation. There are many jewels in the long tail, often hidden due to lack of popularity -- viewing the Web through the eyes of the many instead the eyes of the few improves discovery.
• Attention metrics -- do you ever wonder what your readers are paying most attention to these days? Do they care about Dick Clark? What % of them show an affinity to UK Premiership Football? Understanding what people care about gives you a much different view of the world, and helps break up all the echo chambers.

I believe 2008 will be the year of Attention Data, and user benefits therefrom. Just as the content Web is comprised of HTML links, I think we'll finally see the emergence of the "people Web", comprised of similarities and matches between attention profiles.

Online and Social Network Ad Spend

eMarketer predicts online ad spending to double in the next 4 years, from $21.4B in 2007 to $42B in 2011. Forecasted breakdown below. They obviously think "rich media and video" to be a large market opportunity. I think for both this category and "display ads", the key will be increased targeting. All marketers are interested in more and better targeting.

In a separate report, eMarketer forecasts phenomenal growth for social networking ad spend. Their research shows that 37% of the US adult population and 70% of teens use social networks. They predict almost 4x growth in the next 4 years, from $1.2B in 2007 to $4.1B in 2011.

Funny thing about social network advertising -- users' social network profiles contain so much explicit interest information yet the eCPM for most social networks is $.20 - $.40.  Does this represent inefficiencies in the online ad market, or is this simply because (as Alexander surmises) the harder advertisers try, the more they are ignored -- it's just the nature of social interaction.

I think it's a little of both, and outside of my general belief that explicit interest data is much less valuable to advertisers than implicit interest data (need to blog about this someday) I agree with what Scott says:

Social profiles have economic value, but as remote targeting data, not as local targeting data. The information they contain is useful for ad targeting but on pages/sites other than the social profiles/networks themselves.

What If PageRank Didn't Exist

What if PageRank didn't exist? Wouldn't that solve a lot of problems if there was no longer a "score" for people to manipulate, and bitch/whine about?

This morning I woke up to find that everyone seems to have a PageRank of zero. It was after reading  Tony's post on what PageRank says about Google. I used 3 different tools, such as Check Page Rank, What's my Page Rank, and Smart Page Rank. I checked 5-10 different sites, including my own and popular sites such as WSJ and Techcrunch. All zero.

I suppose this is only an anomaly -- perhaps the PR algorithm is having a bad day.

Person-to-person Targeted Advertising

Ok, so I've got this company that will hopefully disrupt the world of advertising a little -- creating the concept of person-to-person targeted advertising. There are more people creating content on the Web today than businesses, and many more individuals looking for Web site traffic than Google advertisers.

Anyway, we're just getting started and haven't made any real noise, but my man Mike recently informed me that we're at almost 500K targeted "introductions" per week now. Our engagement rate on our widget is about 4%, and 40% of people who view a user profile go to that user's blog.

Cool, but not nearly good enough. We're working on it ...

Search Engine Frustration and Fatigue

Interesting survey on the current state of search, from a user experience standpoint -- apparently from a poll of 1001 US adults conducted by Kelton Research. Key takeaways:

• 65.4% of Americans say they've spent two or more hours in a single sitting searching for specific information on search engines.
• 72.3% experience "search engine fatique" when researching a topic on the Internet.
• 75.1% of those who experience search engine fatique report getting up and physically leaving their computer without the information they were seeking.
• 78% "wished" that search engines could (in effect) read their minds to deliver the results they were looking for.

Also some info on why these people were so frustrated -- the issues look to be somewhat evenly distributed, instead of just one problem:

When asked to name their #1 complaint about the process, 25 percent cited a deluge of results, 24 percent cited a predominance of commercial (paid) listings, 18.8 percent blamed the search engine’s inability to understand their keywords (forcing them to try again), and 18.6 percent were most frustrated by disorganized/random results.

Using Attention Data for User Value

Jeremy Liew writes a post demonstrating the wise use of attention data, and adding value to users. I've been critical of many viewpoints on the "attention economy" (see my post regarding virtuous profiling and the implicit web), but Jeremy echoes the sentiments I wrote for how implicit profiling ought to be done for user value:

This is a good example of attention data in action. Some people call it the implicit web. There are three reasons that this creates a good user experience:

1. It requires no effort from the user (because it uses implicit metadata that is gleaned from watching normal activity)
2. It is presented contextually (so you’re not overwhelmed with data when you don’t need it)
3. It informs decision/actions that the user might take.

The only thing he misses is the importance of privacy controls, but that was due to the context of his post. In a company like Others Online, clearly he'd see the importance of user privacy controls.