Sharing Ad Data

The rapid evolution of Internet advertising is astounding.

A year ago, high click rates were nirvana and killer creative was the road leading there. Today, click-through rates continue to dominate as the metric of online advertising success; the accountability that the Internet offers to advertisers is increasingly leveraged by interactive agencies and online solution providers to maximize a client’s return on advertising dollars spent.

Closing the Loop

Several companies now offer services that report advertising campaign performance based on client defined actions. DoubleClick, MatchLogic, AdKnowledge, Flycast and Avenue A all offer solutions that allow advertisers and/or interactive agencies to create custom metrics for success (sales, leads, page views, and so on) by linking client-defined conversions back to campaign performance.

These products generally work by placing code on an advertiser’s site that feeds data back into the solution provider’s database. The conversion data is then matched back to advertising campaign data (such as click quantities and click-rates) and then evaluated for overall impact using the cost of the ad space to determine a cost per conversion. Sites upon which advertising has been bought can be compared by advertising effectiveness using the conversion metric created for the advertiser.

With names such as Closed Loop Marketing, TruEffect, Advisor, ValueTrack and CustomROI, these advertising campaign evaluation products and services reflect the growing sophistication of Internet advertising. Direct-response marketers are realizing that click-rates alone do not determine the success of a campaign and have recognized the value of including the conversions that define a campaign’s ROI into any evaluation of advertising performance.

Though many interactive agencies (for example: Modem,, K2, Organic and Left Field) use a third-party server to manage ad campaigns for their clients, few have deeply analyzed the collected data (clicks, frequency, unique visitors, etc.), while fewer still have leveraged the resulting conclusions to improve campaign performance and provide the client with strategic market research. The tide is turning as more agencies and online marketing solution providers have begun to recognize the value of data analysis and its impact on discovering ways to increase an advertiser’s return on investment.

As more agencies and their clients embrace conversion-tracking technology, the impact on web publishers is difficult to predict. Initially, some web sites will benefit more than others will. The importance of click-rates as a stand-alone indicator of campaign success will continue to diminish, and sites that perform well in terms of the advertiser’s metric for success will receive more ad dollars. Conversely, web sites that don’t drive conversions will suffer a loss of ad revenue.

Rocky Relationships

In a recent research report, Jupiter predicted that “agencies that do serve ads themselves must ensure ease of use for publishers and share data to maintain good relationships with these publishers.” The report continued to note that “the advent of agency serving heightens the battle over data both agencies and publishers claim rightful ownership to campaign results and user data.” Jupiter duly notes that third-party ad serving is at the core of this growing friction.

Third-party ad technology — the basis for any product that evaluates advertising across multiple, unrelated sites — has had a rocky relationship with web publishers since MatchLogic (then called Media Lab) pioneered the technology when they began serving General Motors ads in late 1996.

Resistance to third-party ad serving began in publisher IT departments. Before business decisions were heavily influenced by ad revenue, webmasters objected to the introduction of technology developed and maintained by a third party. As revenue streams stabilized and actually became predictable for publishers, friction with agencies using third-party servers focused on counting differences which inevitably led to billing disputes.

While counting issues will continue to be a problem, conflict may arise as data that agencies collect on publisher sites is used to add value to agency clients at the expense of future publisher revenue streams. Though publishers can collect similar data from ads running on their own properties, this information alone has limited value.

Publishers can only act upon internally generated data; their analysis will lack the deep insight gained from data collected across many sites. Since advertising decisions are often made based on information inaccessible to individual publishers, an agency may unfairly acquire a reputation for being difficult to please or acting overly capricious.

Armed with in-depth data analysis, an agency or advertiser may choose to reduce the ad dollars placed on a publisher’s site based on poor performance or a diminishing rate of return. The agency may inform the publisher that the cause was poor performance in relation to other sites. Though the reasons supporting these decisions are clear to the agency, the potential frustration on the part of the publisher is easy to understand. A publisher not experienced (with the powerful campaign tracking and data analysis tools available to agencies and marketing firms using third-party ad serving technology) may not understand the underlying rationale.

Additionally, advertisers will exert pressure to move to pricing based on action, while publishers unfamiliar with an agency’s methodology and striving to maintain guaranteed ad revenue streams would continue to push for CPM pricing. Given these seemingly opposite motivations, how can publishers and advertisers using third-party ad serving technology hope to find a middle ground?

Agency-Publisher Collaboration

As the battle over user data intensifies, agencies must find ways to give publishers the opportunity to compete with each other and themselves. One solution may be to supply the publisher with information that will allow a filtered view of an advertiser’s performance metric.

Using some combination of direct response and brand awareness objectives, the agency will have worked with the advertiser to develop an acceptable cost per action. Though the cost is confidential and proprietary in nature, an index reflecting a publisher’s performance relative to the advertiser’s target can be developed and made available to the publisher. This performance index would allow the publisher to use agency-collected data in several ways:

  • More optimization choices. Buys which had formerly been made in predefined locations can now become more flexible. For example, if one slot is performing poorly in relation to the index, a publisher can move the buy to other, possibly better performing inventory.
  • Performance feedback. The index would provide publishers with a real-time indication of campaign performance based on metrics that matter to the advertiser. The publisher, armed with this information, can leverage this learning by producing creative marketing proposals that focus on the client’s defined goals.

This collaborative effort between agencies using third-party servers and publishers offers a way to bridge the gap between the advertiser’s desire to pay for media on a cost-per-action model and the publisher’s preference for cost per impression pricing. An index would allow the publisher to decide the cost effectiveness of meeting the advertiser’s objectives. The publisher can experiment with different slots that perhaps charge different CPM rates. As long as the publisher meets the advertiser’s cost per action goals as reported by the index, the advertiser should no longer be concerned with the CPM rate being charged by the publisher.

Another useful index might be supplied by the agency, which shows publisher performance in relation to the other publishers included in an advertiser’s campaign. Pegged on a baseline average, this index would accomplish two goals:

  • Stir the competitive juices. The competitive index motivates a publisher to find ways to perform better if they are ranked lower than the baseline average.
  • Provide a secondary pricing target. If working only with the cost per action index (and a publisher is performing well ahead of the indexed targets), there will be pressure to raise the CPM rate in order to maximize ad revenue. Though this is reasonable, the competitive index will give the publisher an indication of the impact of a price increase. If the publisher raises the CPM price in order to achieve the cost per action target, the second index may indicate a decrease in the competitive index.

In addition to many other issues that must be addressed in order for this type of agency/publisher collaboration to work, the basic premise requires an agency to employ a third-party ad management system as well as the data analysis tools to produce the necessary indices. A single advertiser could potentially employ such a system. But expertise in data analysis and subsequent campaign strategy implementation is more likely to grow out of an agency with extensive client experience.

The Client Wins

Third-party serving of ads is here to stay. Advertisers like to do an apples-to-apples comparison of publishers using a wide variety of ad management systems and counting methodologies. Additionally, deeper response metrics can be reported and analyzed when all collection of campaign data is consolidated. Many publishers do not have the resources or the business model needed to create in-depth reports for their clients. They are in the business to sell ad space.

However, when the agency/marketing solution provider using a third-party ad server leverages the deep learning taken from the analysis of collected data, publishers don’t always understand why their portion of the media buy suffers (or improves, for that matter).

More importantly, publishers are not given the tools necessary to compete for more ad dollars. The performance indices allow publishers to make informed business decisions based on the quantified business objectives of the advertiser using a third-party serving solution.

Related reading

Present with tag saying 'just for you'