What to look for when picking your ad network

I hope every person involved with the ad network selection process reads Tom Hespos's article, titled "Why ad network reach reports are stupid." It is refreshing to see someone challenge the validity of ad network evaluation based on reach and unique user counts.

To start, reach does matter. Scale is one of the benefits of working with an ad network, and it should always be considered in the buying process. However, given the volume of inventory that is available today and the lack of marketplace barriers to entry, scale has become a commodity between the networks and should not be a primary consideration factor when differentiating among or selecting ad networks.

There have been many instances where a network launches and, soon after, it reaches 100 million unique users, as ranked by an audience measurement service such as comScore. Yet the network is a 10-person company with little revenue, infrastructure, or history. Selecting that network because of its reach statistics would be a mistake. 

Today, planning is often done against the top ad network based on reach. Networks know this, and many of them simply manipulate the system to enhance their comScore statistics. This is done by dropping pixels without serving ads, buying deeply discounted media just to serve a counting tag while an ad is rarely seen, and/or participating in daisy chaining, which is the practice of network A working with network B, who works with network C and so on, with each network getting credit for the unique user. While comScore has begun to address the reporting methodology, it is in a difficult position because many networks exploit the system but also happen to be comScore customers.

On the other hand, advertising agencies like to use scale to justify vendor selection on a plan.  Too often, an agency's media plan will consist of the top networks ranked by comScore, which is the equivalent of only eating at McDonald's, Burger King, or Subway because they have the most restaurants. This is done because the buying community is expected to know how to differentiate networks and what the buy rationale should be, so they rely on something that appears to be measurable and valid.

You can blame the buyers for relying on these statistics, but I hold the networks accountable. Many do a poor job of explaining their differences and appeal to clients as the solution to everything and anything, instead of honestly representing when and how to use their network. 

ComScore has proposed a new methodology that measures actual versus potential reach, but unfortunately, it won't go far enough in turning a manipulated report into the valuable one we would all like it to be. So the question becomes, how should an advertiser or agency select a network if reach isn't the appropriate criterion? 

  • Inventory acquisition model
    How does a network acquire inventory? Is it through an exchange, direct relationships with publishers, daisy chaining, or a combination of these approaches? Ask for a clear answer to this question and have it put it in writing, as a forced ranking or allocation percentage, so you understand where your ads will be running and can make sure that is appropriate for your brand and objectives.

  • Capabilities, tools and compatibility
    It is important to also consider targeting capabilities, ad type and placement, pricing structure, and ease of use. These attributes are often what initially attracts marketers to ad networks. From ensuring that you reach your target audience, to accurately measuring and optimizing campaign performance, the right combination will affect campaign success. But be careful, because nearly every network offers some form of behavioral targeting, so the secret sauce is becoming less and less secret.

  • Customer feedback
    There is no standard way to track renewal rates in the industry, so that is not an option. Relying on case studies is helpful and important, but if you have doubts or want further assurance, I recommend taking it a step further. Ask for client references that you can contact. Talk to agency peers in other offices or ask the network for names of clients who can share their experiences. Verify what the performance and service was like and then decide if that network will meet your needs.

  • The account executive and support team
    When it's all said and done, the people you are working with will have a significant impact on your campaign, how it is approached and managed, and its success. This factor, while often overlooked, directly impacts your experience and should be a main point of consideration. Try to get a good sense of this team's knowledge, work ethic, and accessibility, along with its passion and interest in your business. When everything else is equal or close to it, making a decision based on people is a reasonable rationale, and one an agency should be able to confidently present to a client.

Ultimately, selecting a network comes down to what is most important to you and your client. Once that's determined, it becomes easier to evaluate a network by identifying how it matches up to your priorities and how it will help you accomplish your goals. As this is accomplished more effectively, marketers will realize that reach via the comScore ad network report has become a flawed evaluation metric when it comes to network selection, and one that will not accurately predict or definitively lead to a successful campaign.

Michael Cassidy is CEO of Undertone Networks.

 

Comments

Ekaterina Tsvetkova
Ekaterina Tsvetkova April 21, 2009 at 3:31 PM

I'd like to add one point for consideration: look for opportunities to work with sites that have undergone a third-party audit of web traffic data. Visit www.BuySafeMedia.com if you want to know why.

Eileen Lichtenfeld
Eileen Lichtenfeld April 13, 2009 at 3:56 PM

I'm not clear why how they acquire their inventory is so important as I always ask for a list of the sites we'll be placed on - and then if possible where on the site. This review has been extremely valuable in assuring we're going where we want to go and hitting the target we're paying for.

As for the other comment about why not just video placements - who has the time to watch a video if you don't even know if you have a need/interest in the product??