Five Details You Don’t Know About Data—But Should
Marketers need to err on the side of caution when buying and using data. It is important to know the red flags to look out for so you can identify them when they inevitably appear.
As data and targeting have continued to steamroll their way to the top of every marketer’s priority list over the past year, there is more scrutiny than ever around the information being bought, sold, and targeted against.
This is a good sign for improving data quality in the long term, but marketers still need to err on the side of caution when buying and using data. It is important to know the red flags to look out for so you can identify them when they inevitably appear.
With that in mind, here are five things you probably don’t know about data—but you really should:
The Value Of Third-Party Data Is Out Of Whack
I remember testing a sample of gender data from a leading DMP to find out that it was only 33% accurate when targeting either male or female. With stories like this becoming the rule rather than the exception, marketers are getting increasingly disenfranchised by the third-party data marketplaces that have reigned supreme for so long.
Questionable collection and ownership, inaccurate information, and lack of performance once a data set is targeted have spurred the movement to walled garden environments like Facebook and Google, but something else is happening too: Publishers are offering up their own valuable CRM data as an alternative to the subpar segments available in market—not to mention the additional revenue it puts in their pockets.
Expect this trend to continue in 2016 as more pubs take control and more marketers look elsewhere for data sets.
Cross-Device Data Is Directionally Correct
The proliferation of smartphones and cross-device interest has finally reached a tipping point. Probabilistic matching across devices is in every ad technology toolbox and, in some cases, can provide great value.
At the same time, advertisers are entrusting millions of dollars to companies that claim to know users based on information that is directionally correct. Companies that are transparent will ultimately prevail— but only if CMOs stay informed about how data sets are collected and modeled in order to make big assumptions in the most accurate manner.
Email Is Valuable
The right email address can transform the accuracy and impact of targeting in immensely valuable ways. Environments like Google, Yahoo, Facebook, and LiveIntent all target based on email and have garnered significant praise from advertisers due to the quality of their performance.
Also consider that as the marketplace shifts more toward mobile, the consistency of email provides a baseline for marketers to build upon. For instance, in the past nine years, I’ve had at least five new devices, each with a new device ID, but I’ve carried one, maybe two, email addresses during that time. Device turnover, especially in the U.S., is not slowing.
Expect the increased use of email-based advertiser CRM data to separate high-value market opportunities dramatically in the next 12 months.
More Targeting Is Not Always Better Targeting
I recently had a client inquire about targeting against females who were ages 35 to 54, had made a purchase on an iOS device in the past 30 days, identified as mothers, and lived in three separate zip codes.
Remember, just because one-to-one marketing is starting to be possible, it does not mean that is where we should always hang our hat. Look-a-like models, cohort testing, and channel experience can provide marketers with more reach and is an alternative means to finding new converters rather than going really granular on a perceived behavior or demographic set. The most important thing a marketer can do is test, test, and test some more.
First-Party Data Is King, But You Can’t Win With That Alone
In a 2015 study conducted by Winterberry Group in conjunction with the IAB, 41.7% of marketers cited that more first-party data would significantly advance efforts to achieve competitive advantage through the use of data to support media. While that seems obvious, it also clearly denotes that marketers don’t feel they have enough first-party data, even if it is showing success in their current endeavors. Select third-party data providers and recent first-party data co-ops have a real place in this market to ensure the success of advertising performance.
There will never be “enough” data, so advertisers must look to model via external partners or internal investments in talent and tools to increase the scale of their valuable information living in-house.
From AOL’s investment in Taboola and Time’s acquisition of Viant to Media Math’s spinout of Helix and Kochava’s Collective—expect more advertiser and publisher first-party data initiatives, investments, and talent acquisitions to be a key theme as everyone looks to navigate a rapidly and continuously changing digital advertising ecosystem.
See what the Twitterverse is saying about big data: