Ever since retargeting stormed the U.S. online marketing space, using just one retargeting provider was standard practice. But why is this? To be honest, it doesn’t really make sense. In Europe, for instance, running with multiple retargeters is standard. Thousands of companies do it, including the biggest and brightest. How can we explain the discrepancy in the U.S.? Let's explore.
Healthy competition: My main argument involves the concept of competition. Healthy competition is a great catalyst of innovation and always benefits the customer. It's basic economics.
If you are running with only one provider that’s getting more and more comfortable in its position over time, do you sincerely believe this company would run the extra mile for you? Competition, on the other hand, will force rivals to be at the top of their game.
When it comes to our own industry with a still prevalent last-click wins approach, the concept of competition between providers has its own benefits (especially for the advertiser). Let the boys play -- and the winner takes it all.
CPA: Encouraging competition between performance-driven CPA providers has worked well in affiliate marketing, so why not in display? Whether you prefer a standard CPA or true CPA model in which you pay only for post-click conversions, there's no risk in it for you. So why not test your providers head-to-head and "may the best man win"? It will force them to draw their top guns and resources to generate the best optimization, creative, account services and project management.
To avoid paying twice for the same conversion, make sure you are running with providers who can and are willing to de-dupe for no overlap.
CPA & CPC: Running with different pricing models can also work well. Running with several retargeters (especially CPA and CPC) will always deliver greater reach than with a single provider. This is obviously true when referring to non-automated media buying, but it is also true when buying media on automated exchanges.
In most cases, CPC and CPA companies will not target the same users, since each provider's optimization and segmentation algorithms are different because there outputs are different.
For example, a CPC company will target those likely to click, and therefore will use a "spray and pray" approach buying a huge amount of cheap impressions. A CPA company, on the other hand, will target those most likely to convert and can therefore manage its bids based on a user's actual economic value. A conversion-centered CPA company serves banners with a surgeon's scalpel, not a sledgehammer.
Raising bidding prices: It has been argued that if you run with multiple providers, each provider will be bidding for the same spots on the same websites, driving up media costs. First, as mentioned above, CPC and CPA providers do not usually target the same users.
If several retargeters do end up bidding for the same user on the automated exchanges, the price will increase but only temporarily. Every advertiser knows what is the maximal eCPC or eCPA it is willing to pay and this is usually determined pre-launch. When over-bidding occurs, providers will quickly reduce their bids to meet their client's as well as their own goals.
Overhead: True, working with a single provider is less of a hassle than working with several. That's why I believe the ideal number should be two -- or maximum three. This will enable healthy competition, while keeping overhead to a workable minimum.
How do you prevent consumers from getting bombarded with retargeting ads if run with multiple partners? Setting up frequency caps is helpful, but if you have multiple partners then consumers are still seeing an overload of retargeting ads. Also, with the amount of data that proves consumers are no longer clicking on ads but getting to sites through some sort of view-thru metric, why would you suggests running a CPC campaign? It seems that a large portion of those consumers that click will not be in your target audience and therefore less likely to convert causing conversion rates to drop.
This is absurd. The definition of retargeting says you are going to hit the same people if you use multiple providers. It NEVER works to the marketers advantage, and certainly doesnt benefit the consumer.
In my experience, the CPA is done on a view-based attribution model while CPC is obviously clicks. CPA deals tend to spray and pray MORE than CPC as they need to drop cookies wherever they can to try claiming credit for orders that would have happened anyway. We're a mass advertiser though, so YMMV.
LOL. He's right!
Thanks all for your comments. Meliah, indeed banner bombardment should be avoided as it generates poor user experience and above all it is counterproductive to achieving performance. As stated, due to the different goals and algorithm outputs between different models, the overlap in reach is rather minimal. If some users do end up being targeted by several providers, which in any case should ideally be two (three tops), marketers can always control the frequency by enforce a cap for each provider. I am not suggesting running CPC campaigns but rather only pointed out the type of scenarios marketers face. You’re right that most clickers do not end up converting. myThings, on the other hand, specializes in conversion-driven CPA models.
John, we have empirical data across multiple markets including the US that running with two retargeters can be done with very little overlap. Different algorithms, different goals and different outputs generate different users. One of our customers, among the biggest retailers in the market, told us the reach overlap with another leading CPC provider ran as low as 4%. In fact, scale will only grow when using two retargeters. I’m not saying it’s a perfect solution but its competitive advantages certainly outweigh the disadvantages.
Alex, I disagree with your claim. CPA done right does NOT spray and pray. On the contrary. Performance display is all about generating sales/conversions, not only clicks. That’s a much harder goal to achieve and requires the most sophisticated algorithms, which are wired to target users most likely to convert. As such, we bid higher than CPC. Even if you disregard the fact that it is a poor tactic, CPA cannot sustain the media costs of a spray and pray approach. A conversion-centered CPA company serves banners with a surgeon's scalpel, not a sledgehammer. It’s all about quality rather than quantity.
It's not a claim, it's what I've seen in my own data. Our view-based CPA attribution retargeting campaign's effective CPM is consistently a fraction of that of the CPC's with *billions* more impressions. The network is giving us terribly inefficient impressions, which I've lovingly nicknamed "Dumb Targeting" and welcomed as branding. Good article, I'm adding another CPC retargeter next Q to compete with current state because of it.