Assessing the effectiveness of different targeting strategies

ComScore and ValueClick have released a pretty extensive study on the performance and effectiveness of some of the most common ad targeting strategies currently employed in the digital realm:

  • Audience Targeting – Ads served based on online behavioral data, such as a past interest or interaction with a related product
  • Contextual Targeting – Ads served to related page-level content
  • Efficiency Pricing – Cost-per-click engagement with creative
  • Premium Pricing – High visibility placements on premium publishers
  • Retargeting – Served to users that have previously visited an advertiser’s site
  • Run-of-Network (RON) – Ads which appear anywhere in the network, often optimized by conversion

The study compared these strategies based on 4 metrics: ‘Cost’, ‘Reach’, ‘Lift in Site Visitation’ and ‘Lift in Trademark Search’.

Among the six different placement strategies, Retargeting generated the highest lift in trademark search behavior at 1,046 percent. Despite this considerable lift, the impact of retargeting is limited by its low relative reach, although its cost is moderate in relation to the other strategies.

While Premium and Contextual placements are the most expensive strategies, the advantage of Premium is the higher average lift (300 percent), while Contextual has the ability to reach substantially larger audiences (Reach Index of 73, second only to RON). RON and Efficiency are both high-reach, low-cost strategies, but their lift performance lags that of the other strategies.

The results aren’t much of a revelation. They confirm what we already know about the relative benefits of high reach/low cost strategy versus a targeted/high cost strategy.

What this study does provide though (for once), is some real numbers behind each of the strategies that allows us to further rationalise our approaches.

What’s even more important though, is that the study considers the reasons for each of these strategies being deployed. For instance, if an advertiser has an offer based, direct response campaign objective, then they’re likely to be satisfied with a RON or efficiency-pricing strategy… even if these strategies provide only a short-term impact to visitation and brand lifts compared with the other strategies.

Furthermore, when you evaluate each strategy in conjunction with the purchase methods (CPC, CPA or CPM) most commonly used, it paints a good picture on how each strategy is being used to achieve different marketing objectives.

As evidenced in the above chart, each placement strategy is purchased differently, and often a
combination of strategies is deployed to help a client achieve a specific objective. Marketers typically use
CPM pricing when they seek to create a branding impact, while CPC and CPA are used when the intent
is to elicit a direct-response action, such as visiting a site immediately or purchasing a product online.
Interestingly, ‘retargeting,’ ‘RON’ and ‘efficiency’ are primarily sold using direct-response pricing models,
indicating a desire for a specific action. Using these direct-response pricing models helps to guarantee marketers are only paying for the results they are seeking in terms of immediate response.

I strongly suggest reading the study if you’ve got the time. There’s a lot more detail if you’re interested in seeing the breakdown of each individual strategy compared to their respective cost, reach and brand lift metrics.

It also notes at the end that of the 103 campaigns measured in this study, 90% utilised a combination of 1 or 2 targeting strategies, while about 10% of campaigns employed a mixed approach of 3 or more strategies. In that 10% bucket, there tended to be “one metric that disproportionately beat the norm for lifts in site visitation and search” compared to campaigns that only utilised a single strategy… therefore opening the door to the possibility that a mixed approach could yield greater benefits, but more studies would need to be done in that area to identify the common causes for those abnormal lifts.

Full study available here.


Lift in Trademark Search Within 4 Weeks of Exposure

Total U.S. – Home/Work/University Locations

Source: comScore AdEffx™ .

Lift      Reach Index*    Cost Index**

Retargeting    1,046%        30              373

Audience         514%        30              329

Premium          300%        21             1471

Contextual       130%        73             1473

RON              126%       100              100

Efficiency       100%        57              140


*Reach Index = Avg. Reach of Strategy/Avg. Reach of RON x 100

**Cost Index = Avg. Cost of Strategy/Avg. Cost of RON x 100


2 Responses to Assessing the effectiveness of different targeting strategies

  1. teknophilia says:

    Very interesting, do you know about any opt-out studies showing how opting out of advertising trackers affects their accuracy?

  2. -fb- says:

    None that I can think of or have the links to. But generally speaking, if you’re getting a high proportion of “opt outs” relative to the sample, then you can just keep increasing your sample size to account for it and/or introduce a weighting for adjustment (assuming you’re talking about how opt-outs affect the confidence level of the results).

    If you’re talking about the proportion of people that will “opt-out” in any given sample, then that would vary depending on the segment you’re sampling… There are probably some studies out there which have canvassed certain demographic groups and evaluated their tendency to participate in/opt out of a tracker or survey…

    Generally speaking though, there tends to be higher tendencies to “not opt in” versus tendencies to “opt-out”… Ie. the ‘status quo bias’, a decision-making behavioural process that’s been hotly debated in the organ donation arena (see here)

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: