How Online Advertising Ecosystems Evolve and the Death of the Ad Salesman


Last week Clearspring Technologies [disclosure: I used to work there] publicly announced its new direction as an audience buying platform, leveraging the widespread distribution of its AddThis social sharing tool (which I include at the end of each of my blog posts) to aggregate intent-oriented data from keyword searches performed by web users (AddThis accomplishes this by capturing search query information contained in the referring URL string when a visitor lands on a web page where the tool is embedded). Search re-targeting has become a big driver in the growth of data-augmented display ad campaigns this year as advertisers look to find consumers that exhibit particular characteristics across the web versus targeting visitors to a particular website based on traditional geographic and demographic parameters. The incorporation of data into online advertising has been greatly aided by the creation of self-service platforms that allow advertisers and their agencies to define their audiences and buy access to these users, as well as the accompanying ad inventory, in the process.

These platforms are able to bring efficiencies to the demand side of the of the equation by automating components of the online display advertising ecosystem, something that wouldn’t have been possible without the standardization of display advertising units (which the IAB has achieved by defining such things as ad unit pixel dimensions, file weight and animation length). Without this type of inventory standardization neither the evolution of ad networks, ad exchanges and now demand side platforms (DSPs) would have never occurred, nor the ability to leverage data sources like Clearspring when buying online ad impressions.

This type of evolution hasn’t only been limited to the online display market though.  The IAB began the process of standardizing video advertising in online video players two years ago, creating guidelines that enabled online video ad networks to emerge. These specifications have matured enough to enable the likes of and BrightRoll, who coincidentally enough is a major video ad network itself, to launch video ad exchanges in an effort to bring efficiencies to the scale already available through these video ad networks. On the heels of launching BrightRoll Exchange (BRX) last week, BrightRoll announced this week that it will be leveraging search data from Magnetic to allow video advertisers to re-target audiences across the BrightRoll Video Network as well as BRX in the same manner they do with display advertising today.

As you can see the online video advertising market is following a very similar path as online display advertising has in its maturation- leveraging ad unit standardization to bring scale to the industry, which in turn has led to platforms being launched in order to bring efficiencies into the marketplace and incorporate data to enable audience targeting at scale- albeit in a much more condensed time line. Where online video advertising trails display advertising in delivery effectiveness is in the nascent state of its exchange marketplaces and integrations with DSPs and data sources, which should both evolve rapidly over the course of the next 12 months.

Based on this pattern, mobile applications should be the next advertising segment to follow this evolution as mobile ad networks focusing on the Android and iPhone platforms have proliferated. The biggest thing holding back the mobile application advertising industry from further efficiencies is ad unit standardization as the IAB is not yet willing to go down that path, only providing best practices as of yet.

This automation and scale being brought into buying online advertising inventory has started coming at the expense of ad sales people. Case in point, in June the Fox Audience Network disclosed that it would be laying off 5% of its staff, all from direct online ad sales, as a result of the success the company was seeing from the self-service display advertising part of its ad network business. So is there a future for ad sales people in online advertising or will they be a casualty of efficiency like blue-collar line workers of the industrial age?

To survive and thrive, ad sales people need to re-orient their thinking from selling impressions to creating experiences for brands and advertisers that focus on two core concepts: integration and social. Integrated campaigns enable advertisers to achieve higher engagement and mind share than through individual ad placements. In traditional display advertising this can be accomplished by implementing branded skins into websites or sponsoring sections of content. In video this might mean product placement in episodes or storylines and for mobile it might involve sponsoring the give-away of previously paid apps or premium features. The key is subtly associating the advertiser with the site and content so as to create a positive connection versus an annoying one elicited by standard display and pre-roll video ads.

In terms of social, I’ve previously written about the lack of innovation in online advertising since its advent 15 years ago and that the focus should be on creating socially-oriented ads (since social networking is what most web users are spending their time doing online these days). Developing ways for users to interact with and provide feedback on ads in real-time as well as leveraging a web property’s user base to collaborate in the creation of campaigns, which I’ve also written about, creates engagement because the users who participate have a vested interest in the outcome- just ask the Old Spice Guy.

Regardless of finding the right experiences to drive success for advertisers, ad sales people need to evolve ahead of these online advertising ecosystems or they will end up like Willy Loman.

Liked this post? Share it with your colleagues and friends below.

[tweetmeme style=”compact” only_single=”false” source=”alexcalic”]

Bookmark and Share

Online Display Advertising’s Data Game- Who Will Be Left Out?

2010 is likely to go down as the Year of the Audience in online display advertising as marketers, looking for better returns from their investments in a challenging economy, are turning to search advertising’s strengths to help reach specific individuals across the internet. Search is the largest, and continues to be the fastest growing, segment of the $20 billion online advertising market in the U.S. because it brings advertisers results. The move toward delivering ad impressions on a unique visitor basis across disparate websites is an effort to improve the results of display advertising campaigns by leveraging what makes search advertising so effective- matching web users’ displayed interests and intentions with an advertiser’s defined audience.

This trend in data-driven ad targeting, which is expected to be the primary growth driver for display advertising going forward, has given birth to a number of new intermediaries in the online display advertising ecosystem:
As one can see, depending on an advertiser’s needs and requirements, there are many components to delivering ad impressions to targeted audiences. The new companies that have launched to fulfill the roles of these new intermediaries, especially around data, are capturing most of the additional value being created in the ecosystem. With so many new entrants competing for a piece of the display advertising pie who are the winners going to be?

Before answering this question, it’s worth defining what data is actually being captured and how it’s being leveraged by intermediaries on behalf of advertisers. When someone performs a search query through a search box on a browser or through a website the URL associated with the results page will contain the keywords used in the search request.

When the searcher clicks on any link on the results page, this URL string is passed to the destination website along with the user. While the keyword data is being captured by the web publisher, social tool services, such as commenting and sharing services, can also gain access to this data if their service requires JavaScript implemented on the web page. Marketers, through various demand side intermediaries can reach this searcher by having the intermediary place a cookie on that individual’s computer once they land on the publisher’s site to identify that person when they visit a different website where the advertiser has the ability to serve a banner ad based on the interest the user has shown through their search activity.

Here’s an example of how it would work. Johnny searches for “cell phone” on and clicks on a link on the result page that sends him to, where a cookie is placed on Johnny’s computer by Invite Media on behalf of AT&T’s agency. When Johnny visits, a website through which the agency has the ability to buy inventory via Invite Media, he sees an ad from AT&T for a cell phone.

Because audience targeting revolves around intent-oriented data, the intermediaries that have arisen within the ecosystem to fill various data needs are going to experience the greatest consolidation as some of the services being provided morph into one another or become more standardized across other demand side intermediaries. Anticipating and addressing the needs of this evolving marketplace will be the only way for companies to survive and prosper.

Stand-alone data selling is not a viable business. While selling intent-oriented data to online display advertising intermediaries can be a low-effort revenue stream, it’s an ancillary business even for the largest data providers. As more web publishers and social tool providers begin to offer advertisers access to user search data, that data starts to become commoditized as advertisers and their intermediaries have more partners to choose from to create their audiences. Automation around identifying data from appropriate partners and delivering audiences for campaigns will only hasten the commoditization of keyword data. Google on the other hand, by keeping its search-related data proprietary rather than selling it to third-parties, has been able to determine the value of its data through the development of AdWords. Companies that sell their data to third-parties are allowing these parties to determine the value of the data to their own detriment.

Data exchanges must evolve or die. Being a broker between data suppliers and intermediaries is a short-term business model. Because the whole notion of using intent data to target users is in its infancy, data exchanges have become an easy starting point for advertisers to find data to test display campaigns against. The problem is that as other intermediaries within the ecosystem get more experience and smarter about audience targeting, they will seek out direct relationships with the largest and most effective data providers, thus bypassing data exchanges all together. For data exchanges to survive they need to evolve to provide value-added services to their clients such as those being offered by Demand Side Platforms (DSPs) and Social Data Targeting companies.

Adding social data points will prove to be valuable. While keyword data has the potential to become commoditized as previously explained, data culled from users commenting on articles and sharing links into Facebook and Twitter provides unique additional value to display advertisers. Continuing with the “cell phone” search example, if Johnny ends up on a web page after searching for “cell phone” and then shares a link  to a positive article about the iPhone, the additional information associated with the link being shared (iPhone versus just cell phone) helps better define Johnny’s interest and provides a stronger signal of his purchase intent. Even though social data can provide a higher degree of confidence related to search intent, the data itself is not as structured as search data. As a result, being able to package the information effectively and make it actionable will be the key to success.

But can social data targeting companies find the holy grail? A number of companies are exploring how to leverage social data, in combination with search data in many cases, to provide better conversion and larger audiences for targeted campaigns. While the approaches to finding the best algorithm might differ (Media6 Degrees looks for network neighbors while 33Across creates influencer social graphs and Lotame categorizes user activity on social networks), any sign of superior, and repeatable, results will quickly drive acquisitions of these companies by one of the GYMs (Google, Yahoo or Microsoft) to be leveraged internally or by their respective ad exchanges. DSPs looking to expand and enhance their platform offering could also be an acquirer, but would need to do so before valuations get to high.

Demand-side platforms’ dilemma. The proliferation of DSPs is not without warrant as they hold the promise of tying together disparate ad exchanges and ad networks, as well as data providers, into a single interface to enable real-time bidding of online display inventory for targeting purposes by media buyers. The key to how this market evolves will depend on which companies will be the first to be acquired and which ones decide to make a go of it alone. The two most natural types of acquirers, GYMs and agency holding companies, each face their own potential challenges in purchasing one of the players in this space.

Agencies would benefit the most from owning one of these platforms, but are unlikely to pay the full or potential value that the venture-backers of these companies are looking for because any revenue being generated from competing agencies on these platforms would disappear upon acquisition by another agency. Even though Google is a likely acquirer at some point this year, they, along with Microsoft and Yahoo, risk alienating clients and partners of potential acquisition targets by bringing the neutrality provided by the DSP platform into question. Preferential treatment of intermediary services from the GYMs, such as ad exchanges that are integrated into the DSP, would destroy the platform’s business and partnerships. Companies such as AppNexus, Invite Media and MediaMath have the early client adoption and capitalization (i.e. they haven’t raised too much money) necessary to be likely acquisition candidates.

Because DSPs are already enabling data to be combined with inventory acquisition on their platforms, one or two of these companies have the potential to incorporate the capabilities of Data Exchanges as well as Social Data Targeting companies (the latter through acquisition) to create a more robust offering that simplifies the demand side of the display advertising equation. With its early success and strong management team, AppNexus could be the one to create a viable, stand-alone entity.

The supply side will strike back. The early days of data-driven targeting has enabled advertisers to find audiences across premium websites that charge higher CPMs for ad impressions and target those same individuals across ad networks and websites with cheaper advertising inventory. This has created an opportunity for inventory yield optimization companies to help publishers retain some of the revenue opportunity and CPM value being lost to the demand side platforms. This is likely to include enabling audiences to be created and targeted across disparate websites of premium publishers as well as the development of supply side exchanges, as suggested by Will price, CEO of Widgetbox. Companies such as AdMeld, Rubicon Project and Yieldex will be the biggest benefactors of this in addition to their publisher clients at the expense of ad networks and ad exchanges.

As competing offerings begin to look and sound the same within and across intermediaries, analytics and transparency will be the keys to building a successful business. Analytics will not only need to serve as a dashboard for campaign results, but also provide insight into which aspects did or did not perform well and potential reasons as to why. With so many parties involved in every transaction, transparency will grow in importance from a trust and verifiability perspective as well as enable insights to be drawn from each aspect of the value chain.

Beyond this, determining the winners in the online display advertising ecosystem will be somewhat dependant on Google’s actions as they have made it apparent that display advertising is the company’s next growth opportunity. Google has not been afraid to pay top-dollar to acquire the pieces they need or build these services themselves, thus driving potential acquisition targets into the arms of Microsoft or Yahoo and leaving the rest of the competition out of the game.

[tweetmeme style=”compact” only_single=”false” source=”alexcalic”]

Bookmark and Share

Now That Banner Ads Have Turned 15, It’s Time for Them to Get Social

Head in the SandA couple weeks ago the advertising industry celebrated 15 years since the first display banner ad was presented online. In the years since then as the ads themselves have become more creative and dynamic through the use of Flash and JavaScript technologies, and the units through which these experiences are being delivered has been standardized across the web, how consumers engage with these ads hasn’t actually changed.

For the most part agencies and their clients have treated advertising on the internet much the same way they have older content mediums like print, radio and television: as a one-way channel to broadcast a marketing message to consumers. Since the internet has been a read-only environment for most users over much of its existence, it’s easy to see why advertising online evolved in the same manner as these other content channels. With the rise of blogs and social networks though, web users now have both read and write capabilities that allows anyone with an internet connection and keyboard to give their two cents online. Advertisers have been slow to acknowledge the two-way relationship that now exists on the web with consumers, whether they want to take part in the conversation or not.

Some social media-focused companies have taken it upon themselves to develop more engaging ad experiences on behalf of advertisers, such as enabling video ads to be shared across social websites [disclosure: my company Clearspring powers this feature for VideoEgg]. While this does create value for advertisers through individual endorsement, since the ad is being perpetuated by a person versus an ad server, the messaging doesn’t provide for any feedback. The same could be said for ads which aggregate Twitter commentary or Dugg articles around a particular brand, event or topic. Even though these ads dynamically insert content from specific sources into traditional banner ad units, the information is  moderated before being broadcast and isn’t necessarily oriented to the actual campaign.

Getting agencies and advertisers to embrace the idea that making their ads social will actually benefit their business requires participation from the largest social media sites with the necessary social capital (i.e. a big or growing coolness factor) to experiment with non-standardized advertising. Facebook and Twitter are obvious candidates to lead this effort not only because of their large audiences but because they incorporate the most prevalent user experiences on the social web: community-oriented, information streams of shared content.

Facebook has already put a lot of effort into creating new display ad units and ways for advertisers to engage with their audience, allowing Facebook users to not only interact with ads (by watching videos, RSVP-ing to events, voting in a polls, becoming fans of companies, etc.) but also provide feedback on uninteresting ads.

Facebook Ads

Since Facebook has created a self-service platform to manage the entire advertising process, ads can automatically be delivered at scale across  the entire site. And with Facebook focusing on providing the social identity layer to the web via Facebook Connect it’s easy to see how they could standardize and distribute their own ad units and engagement across participating Connect sites- much like Google has done with search and AdSense.

While Twitter has thus far avoided placing ads DiggAdson its platform, many Twitter apps are primarily monetizing their service through traditional display advertising units. To create a unique and more valuable advertising experience though, ads should be integrated into the actual functionality of these apps. Since tweets consist of text and links, the most logical type of ad unit would mimic sponsored search ads. Digg, whose community is similar to Twitter’s in that they share the most popular content on the web, offers the best example of what socially oriented, stream-based ads might look like.  As with Facebook ads, Digg allows its users to provide feedback on the sponsored articles on the site in real-time.

Whether it’s Digg, Facebook, Twitter or someone else, whoever can define the new display and in-stream social ad standard has a tremendous financial opportunity as Digg understands in contemplating syndicating their ad format to third-party websites via its own ad network. Developing ad standards are important for agencies as it allows them to execute campaigns on behalf of their clients at scale, with minimum creative friction, across a wide variety of websites.  For  most social media web properties that can’t command their own ad standards this gives them a framework for incorporating more relevant monetization experiences into their sites and services. Let’s not forget that Facebook leveraged standard display ads as a way to generate revenues when the site first launched.

These examples are just a starting point for social ads and will evolve over time. The key is that experimentation is occurring now with willing advertisers (whether they are participating because they truly care about the feedback or just want access to consumers on these sites is another story).  While some advertisers will be brought kicking and screaming into the socialization of advertising, early adopters will yield the greatest benefit from capturing the data and engagement directly from their audience versus pretending the conversation doesn’t exist.

[tweetmeme style=”compact” only_single=”false” source=”alexcalic”]

Bookmark and Share