Google Shifts Data Focus From Retargeting To Remarketing, Targets 3rd-Party Data

The advertising industry rallying around improving the quality of data to target ads continues to make improvements to processes and techniques. Google began rolling out a remarketing tool in Google Analytics on Friday to help marketers gain insight into targeting ads. But if the brand's data mixed with inadequate information from a third-party data seller,  a company's retargeting strategy will deliver less than stellar results.

At Guthy-Renker, Colette Dill-Lerner, vice president of Internet marketing, said she has looked at data files "where nearly 50% of the gender is wrong." Some marketers think this characteristic remains one of the most difficult to identify.

Aside from gender, other discrepancies exist in third-party data that limit ad efficiencies.

It led Guthy-Renker to create a scorecard with help from company partners like demand side platforms and data management platforms. The major gap between the person's actual characteristics and how they are portrayed online requires cross-checks. Dill-Lerner said data may give company marketers insight into the 35-year-old mother of three, but it's not clear how that translates digitally.

Marketers need to identify accurate data providers. Leon Zemel, chief analytics officer at [x+1], also points to gender as a difficult attribute, suggesting that brands should focus on household rather than personal data. "User-level data is less accurate than household level data," he said. 

In general, data providers rely on cookies to retrieve data, but gender is a tough one because with multiple people in the household, the data could be accurate at the time retrieved, but then someone else uses the computer. "We don't have the total answer yet, but we're looking at it," Zemel said.

On Friday, [x+1], which supports a data management platform, expanded capabilities of its Origin Enterprise Data Management Platform, adding Creative Decisioning across its audience and media types, including direct purchase media and real-time targeting for ad impressions. It uses first-party match keys, which means it's not dependent on third-party cookie data.

Similar to data quality, cookie identification needs to improve, said David Norris, CEO at BlueCava, which supports device identification for mobile and desktop devices. He said the average U.S. site can only accurately identify between 10% and 20% of visitors because consumers delete cookies from browsers or Internet protocol addresses. Better identification translates into higher CPMs and better targeted ads.

Quality inventory and data doesn't support real-time ad targeting, either.

The need to protect brands and celebrities running campaigns prohibits Guthy-Renker from buying as much remnant inventory as Dill-Lerner would like. "We have a project to profile our customers and conduct attribution modeling against the profile before rolling it into a dynamic creative solution," Dill-Lerner said. "We're nowhere near ready for market."

Data quality supporting ad targeting also challenges the industry's focus on "scarcity," the notion of reducing the number of ads served and available units on publisher sites to improve conversations, conversions and prices. The ad industry served more than 400 billion ads monthly on major U.S. publisher networks between June 2011 and June 2012, estimates Mike Rich, vice president of media at comScore.

Since too much inventory drives down prices, Rich said online advertising can't survive unless CPMs rise.

2 comments about "Google Shifts Data Focus From Retargeting To Remarketing, Targets 3rd-Party Data".
Check to receive email when comments are posted.
  1. Ray Kingman from Semcasting, July 30, 2012 at 8:56 a.m.

    Good insight into the issues with data accuracy. Especially relevant is the comment regarding HH level data versus individual. Ultimately we want to provide insight on the drivers of consumer transactions – bemoaning the accuracy of gender on who is using a device is secondary to can they afford it and are they in-the-market. At a HH level this can be done accurately online using other methods than trying to imprint cookies with a precise intelligence that will only have limited reach anyways. There is an alternative that doesn’t rattle the FTC = http://semcasting.com/solutions/consumer-ip-zones.html

  2. John Grono from GAP Research, July 30, 2012 at 9:29 a.m.

    Ironically, as consumers adopt more 'personal' devices (i.e. everyone in the HH has their own smartphone and tablet device) then the misattribution declines - leaving just the over-estimation due to multiple device usage by the same person and due to cookie deletion. Empirical data collected from identified participants needs to be used to quantify these parameters to adjust the traffic-based cookie data.

Next story loading loading..