MetaMuse

New Challenges for Web Publishers, Reminiscent of Adware/Spyware Market 5 Years Ago

The market for 3rd party audience data continues to grow, but I'm seeing evidence of illicit activities that are somewhat reminiscent of the adware/spyware activities of 2003-2005. It's disconcerting, because the 3rd party data market has developed the right way so far -- for ALL parties involved, and especially for publishers. But we are now at a point where a few bad apples could spoil the whole bunch, and are causing new challenges for premium publishers.

In the last two weeks, I've been in San Francisco and New York, taking part in 3 different online advertising conferences, and meeting with top 3rd party data companies (such as eXelate, BlueKai, AlmondNet, TargusInfo, Rapleaf, etc.), ad networks, and publishers. A few observations ...

First, I'm starting to see more evidence of illicit data collection from publishers. Just as we saw advertising-based applications illicitly installed on consumer's computers 5 years ago, collecting consumer browsing behavior (without knowledge of the publisher or consumer), valuable audience data is being taken from premium publishers and used by 3rd parties without the knowledge nor the remuneration of the publisher. It's being done via ad tags, which if not properly reviewed/screened, may contain Web beacons that pass audience data to ad networks, advertisers and 3rd party data partners for re-targeting, re-use, etc.

As a result, premium publishers need to be on the lookout for unscrupulous ad networks and advertisers that bring short-term revenue lift but then leave with valuable audience data. And as the market for data expands and $ start to really flow, 3rd party data partners will need to be extra careful to control their data sources/channels. When publishers see these 3rd party Web beacons showing up on their pages via ad tags, it casts a negative light on otherwise reputable data companies -- not the advertiser or ad network who passed the data to the data company in exchange for payment.

Secondly, I noticed some 3rd party data providers are offering an API into their data store. Basically, it's a small snippet of JavaScript that, when placed in the ad tag, pings the data service for any user-specific information which is then appended to the ad server request. This is a fine model and service for publishers when they are using it on their pages, but publishers need to look out for 3rd party ad tags that contain these API calls -- every request out to a 3rd party from their site represents potential "audience data leakage".

Lastly, I was discouraged to hear that some unscrupulous advertisers are buying cheap run-of-network inventory via ad exchanges and running the NAI "opt-out" scripts within ad tags. Just as 5 years ago when spyware companies were detecting/uninstalling competing applications as an offensive maneuver, seems now we're seeing companies attempting to reduce the targetable population for their competition, increase their data acquisition costs, etc.

In the online advertising market where "audience" has emerged as king, safeguarding audience data is quickly becoming a core challenge/risk for premium publishers. This makes it doubly important for publishers to control their sales channels, work with trusted partners only, and leverage technology approaches to ad quality -- all hard things to do for publishers already burdened with increasing ad sales and operations challenges. Yet another reason why premium publishers will increasingly turn to sell-side technology platforms like REVV for Publishers. And for those publishers that don't, they not only risk their own brand and core data assets, they also impede the proper development of the market as a whole.

What is the definition of "Audience"?

So much talk this year in the online advertising market about "audience" -- audience insight, audience-centric planning and buying, audience targeting, audience optimization, etc.

But what is "audience"?

Of course there are multiple definitions, but in the context of online advertising I define it as a "self-selected group of people sharing similar attributes". It's not content. It's not context. It's about people, their self-identified attributes, and their attention. Can we agree?

Ads More Effective on Social Networks than Portals

According to this article and eBay Advertising's 2nd survey of European shoppers, 60% of consumers are most receptive to targeted display advertising when they are on an e-commerce site, compared to 5% on portals and 7% on social networks.

I take this to mean that e-commerce sites are 12x more effective for targeted display ads than portals (and therefore the eCPM cost of portal advertising has to be 1/12th in order for the ROI to be the same), whereas they're 8.5x more effective than social nets.

More shocking is their conclusion that targeted display ads on social networks are 40% more effective than portals!

How Data is Revolutionizing Advertising

I'm serving on an interesting panel October 29th in NY to discuss how targeting data is revolutionizing the concept of “audience buying” and fueling better performance for advertisers. Really good set of panelists with representatives from ad networks, exchanges, advertisers and optimizers! See below (click for larger version).

Sponsored by eXelate and moderated by Forrester Research.

Others Online Acquired by the Rubicon Project

On September 15, 2009, the Rubicon Project announced the acquisition of Others Online, a company I founded in 2006, the 3rd company I've founded, and the 4th startup I've built up.

Others Online was founded on the vision of advertisers being able to reach people, not pages. When I had looked at the revenue growth of Internet advertising revenues from 1997-2007, I noticed two clear growth trends. The first growth trend, from 1997 through 2000 (v1.0), was primarily fueled by content targeting -- Web sites and content were the proxy by which advertisers reached people.

The second growth trend, starting in roughly 2002 (v2.0) was primarily fueled by keyword targeting -- search and context keywords were the proxy by which advertisers reached people.

My belief, and the bet we made with Others Online, was that the evolution of the Web would result in a third growth trend (v3.0) -- the ability for advertisers to reach specific audiences directly, thereby supplanting the use of proxies. Indeed, the industry is seeing tremendous growth in the area of audience targeting. However, the ability to correctly identify specific audiences is dependent on two things:

1. Lots of audience attention data, representing what we're devoting our precious attention span to online.
2. Technology to aggregate, process, score and summarize the audience attention data.

Audience attention data is everywhere. Every online media company has a fragment of attention data, and we've seen a number of companies form in the data space, each offering valuable data for sale to advertisers. Not just "behavioral" data either -- contextual, demographic, purchase intent, location, search keywords, etc. But that's the problem; no one has a complete picture, merely fragments. The true market growth opportunity can't be realized unless/until there's an aggregation point for audience data, coupled with media inventory, and the mechanism by which advertisements can be targeted based on the multiple facets that define us as people, at scale.

Others Online built a powerful technology platform to aggregate, process, score and summarize audience attention data, with services that helped publishers better understand, segment and target audiences. In our discussions with the Rubicon Project, we quickly realized the power of combining their global reach and scale (45B monthly impressions to 500M users, across 30K Web sites) with our technology to not only provide incremental value to everyone connected to the Rubicon Project platform, but also to fuel the next growth trend in audience-targeted online advertising.

As with all announcements though, there are bound to be misconceptions. Though future press releases will certainly add clarification, I thought I'd at least try to dispel a few potential misconceptions upfront!

• This acquisition does NOT make the Rubicon Project a "data exchange". We do not resell publisher or 3rd party data, nor will we. We provide a platform through which 3rd party data providers are developing an incremental distribution and sales channel.
• The Rubicon Project (still) does NOT sell media directly to advertisers. We remain 100% focused on providing value to premium Web publishers, helping them make more money, protect their brand and save time.

The Others Online team couldn't be more excited to be a part of the Rubicon Project. But mostly, we look forward to adding value to their publisher partners and in doing so, hopefully realize the market opportunity at hand.

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.

Latest from the World of Online Advertising

Catching up on some blog posts and thought I'd note a few things that caught my eye:

• Roughly 8% of the total ad dollars in the US is spent online, while consumers spend about 30% of their time surfing the Web, according to MediaPost.
• About 85% of total ad dollars spent online in the UK is performance-driven, compared to 50% in the US.
• ValueClick launched ActiveAds, which allows marketers to display customized messages in ad creative w/out the need to develop and manage multiple creative units. Some launch though ... there is nothing on their Web site about this!
• eMarketer forecasts global ad spending on social networks to increase 17% in 2009 over 2008, revised down from their previous estimate of 32%. Increase in US only social network advertising is expected to be 10.2%, reaching $1.3B. MySpace and Facebook account for two-thirds of this in the US.
• According to Nielson Online, time spent on social networks and blogs grew 3x faster than the overall Internet rate, and are now the 4th most popular online category -- AHEAD of personal email (and behind search, portals and PC software).
• Read in eMarketer that 41% of US Internet users surveyed said they paid more attention to advertising that was personalized. And nearly the same proportion of respondents (39%) said they were more willing to click on such personalized ads.

Why Ad Networks Need to Co-op Data

With the understanding that online advertisers are shifting towards buying specific audience behavior instead of impressions by the tonnage, it must certainly follow that access to user behavior is critical to ad networks' ability to segment and sell on that basis. That access generally comes in the form of raw page view data gathered across their network.

I took a look at the top 40 ad networks in the US in December 2008, according to Comscore data. While Comscore measures uniques, % reach and page views per month, they don't measure "page view reach" or the number of "touch points" per user (which is my term for any individual behavioral data point per user). The more data points they have for each user, and the more reach across the Web, the stronger their position insofar as behavioral profiling and targeting. So I calculated that myself in the table below.

Here are a few things that stood out for me:

• Yahoo has access to the most user behavior, reaching 51% of all pages viewed, and with almost 1,500 touchpoints per user per month
• Platform-A and Google are not far behind, each with more than 10x the behavioral data that Specific Media has.
• Most of the ad networks listed don't have even 5% of the access to page views that Yahoo and Google do.
• Only if you combine all the page view data from Specific Media and every one of the networks in positions #6 through #40, do you approach the reach of either Yahoo or Google.

Insofar as behavioral targeting, the "rest" of these networks aren't competing with each other as much as they're competing with the "big 3". In order to compete most effectively with the "big 3", ad networks should co-op their raw behavioral data and thereby compete (better) as they always have -- on the basis of their sales process, ad operations, and publisher network.

Note: seems that Comscore doesn't include ad networks who haven't subscribed in some way to their service.

Can Display Ads Outperform Search?

Fred's views of "predictions" match mine, as does his prediction. In this post (where the comment stream is even better than the post), Fred predicts:

display advertising will get so cheap and the tools to target it will get so good that it will be shown that it can outperform search

I wrote about this at the beginning of 2008 in my post CPM Rates Will Rise, They Have To (which means either I was exceptionally wrong or exceptionally astute), where I posited that high CPC rates (in SEM) and better targeting/technology will be the major forces behind a rebirth in display advertising. In any event, the comment thread in Fred's post was rich in opinion and dialogue, which led me to comment: 

CPM rates are already extremely low. I don't think remnant rates are going to fall much more. And an average CTR of .1% means that banner blindness can't get much worse either. In other words, I think we're at the bottom for eCPMs and CTR (for the majority of inventory out there), and as Fred points out, the tools will pave the way for a rebirth in display.

As I wrote last year, high CPC rates (SEM) and better targeting/technology will be the major forces behind a rebirth in display advertising. If you're a marketer seeing your bid rate for keywords exceeding $2 per click, the effective CPM rate you're paying is high. At .25% CTR, you're paying $5 CPM. At 1% CTR, it's $20 CPM. If you're paying $2 CPC rates , all you need is a measely .05% CTR to put $1 CPM inventory on par (which is what 83% of the inventory is selling for, and also 5-10x the price of most social media inventory these days)!

Display offers >25x more volume than search, and new targeting techniques for CPM ads will shift dollars back towards the display market -- assuming ad networks rise to meet the market opportunity. I think they will, slowly but surely. They have to, in order to grow their business.

I was very pleased to see his reference to the estimation that Google has over 1.3M advertisers -- I've been looking for that data. We've also seen at Others Online that display ads are the only advertising format to monetize social networks -- text ads don't work at all. So in order for display to make a big comeback, it has to cater to this long tail of advertisers. Which means to me there are three ingredients necessary to offering advertisers the performance of search:

1. Targeting - Advertisers must have access to deep and more standardized targeting. Proprietary categories/channels/audiences are arbitrary and non-standardized. Keywords are the lingua franca of SEM and SEO, and everyone knows what they mean. In the absense of a search query, the targeting can only be based on context (targeting content) or affinity (targeting the person).
2. Self-service buying - Long-tail advertisers want to dabble first with low $ amounts, and the friction needs to be taken out of the process -- they need to be able to place a buy in minutes, as they now do with AdWords.
3. Ad creation tools - Indeed, part of the friction is the design/dev of the ad creative. Tools must make this as easy as creating/optimizing text ads.

Recent Tidbits: Online Advertising Blogs

Just catching up on a slew of posts from blogs that cover the online ad space. Here are some tidbits that caught my eye.

Interesting comment from a recent Adotas blog post regarding targeting people instead of pages:

In a recent eMarketer study, 90% of advertisers said they plan to use ad networks in the coming year and the top two factors that they will use to differentiate networks were the quality of inventory and the audience targeting. Networks that have quality inventory and rich audience targeting capabilities should be some of the first places marketers and agencies look in 2009 and beyond to keep revenues flowing at lower costs and greater efficiency.

From Fred Wilson summarizing the comScore whitepaper on the unseen benefits of display advertising, compared to search:

It’s clear that display advertising, despite a lack of clicks, can have a significant positive impact on:
- Visitation to the advertiser’s Web site (lift of at least 46% over a four week period)
- The likelihood of consumers conducting a search query using the advertiser’s branded terms  (a lift of at least 38% over a four week period)
- Consumers’ likelihood of buying the advertised brand online (an average 27% lift in online sales)
- Consumers’ likelihood of buying at the advertiser’s retail store (an average lift of 17%) 

The whitepaper also reminded readers that while search advertising was far more effective than display, display ads have 10x the reach that search ads do. I've pointed out before the difference in volume  and eCPM rates of search vs display advertising and comScore points out that marketers ought to utilize both for maximum effectiveness.

I also noted in the comments two specific quotes from Gian Fulgoni, comScore chairman:

• Ffour independent studies (one of them conducted by comScore) have now shown that at least 30% of Internet users delete their cookies in a month.
• Average click through rates on display ads have dropped to under 0.1%