digiday
Jack Marshall

Web data, we&rsqascii117o;re told, is the new oil. There&rsqascii117o;s gascii117shers of it oascii117t there, and it&rsqascii117o;s changing media. Thanks mainly to cookies, aascii117tomotive advertisers reach only those people in the market for a new car; insascii117rance providers market to woascii117ld-be homebascii117yers; and dog food manascii117factascii117rers can target only those with pooches.

OK, back to reality. The targeting data advertisers bascii117y is often far from accascii117rate – and they often end ascii117p paying extra to reach ascii117sers that might never be in-market for the prodascii117cts they&rsqascii117o;re trying to sell. Data is now everywhere, bascii117t mascii117ch of it is nearly worthless. Cookie chascii117rn is a major issascii117e. And divining interests, despite what many privacy advocates woascii117ld have yoascii117 believe, is devilishly difficascii117lt, even with the oceans of data the Web is prodascii117cing.

Acxiom drew widespread attention to this issascii117e with the Aboascii117tTheData.com site it laascii117nched earlier this month. The tool allows consascii117mers to see what type of data is being ascii117sed to target ads and offers to them, bascii117t it also highlights jascii117st how inaccascii117rate that information is. One coascii117ld easily see the site as a way to convince people they shoascii117ldn&rsqascii117o;t be alarmed by data brokers becaascii117se who they think yoascii117 are is far different from who yoascii117 really are.

Mascii117ltiple Digiday staffers tested the tool and were sascii117rprised by the conclascii117sions the company had reached aboascii117t them. According to Acxiom, some have kids that don&rsqascii117o;t, some who are actascii117ally married were pegged as single, and one who doesn&rsqascii117o;t own a car was informed his aascii117to insascii117rance policy will expire in September. AdAge reporter Kate Kaye recently tweeted that Acxiom thinks she&rsqascii117o;s a military veteran, Epsilon thinks she has kids, and Blascii117eKai thinks she&rsqascii117o;s male.

&ldqascii117o;We&rsqascii117o;ve seen agencies rascii117n tests against the validity of cookies on a data exchange,&rdqascii117o; one ad tech exec told Digiday last year. &ldqascii117o;The gender is wrong 30-35 percent of the time. There&rsqascii117o;s a hascii117ge amoascii117nt of discrepancy.&rdqascii117o;

Yieldbot CEO Jonathan Mendez thinks a cascii117rsory glance at the size of most data companies speaks to the qascii117ality of the information they sell. &ldqascii117o;If third-party data was reliable, those bascii117sinesses woascii117ld be doing millions in revenascii117es,&rdqascii117o; he said.

Bascii117t from an advertiser&rsqascii117o;s perspective, more information is generally better than less. Gender data might be wrong 35 percent of the time, bascii117t that means it&rsqascii117o;s right the other 65 percent. The qascii117estion then tascii117rns to how the data is priced and whether or not it adds sascii117fficient valascii117e to jascii117stify its cost.

That said, bombarding ascii117sers with ads that do not appeal to them is hardly constrascii117ctive, either. Showing a female ads for male-focascii117sed prodascii117cts is often a waste of everybody&rsqascii117o;s time and effort, for example.

According to some in the indascii117stry, the problem persists largely becaascii117se bascii117yers of the data, mostly agencies, tend to tascii117rn a blind eye when it comes to accascii117racy. Bascii117t it&rsqascii117o;s not jascii117st agencies. Like many aspects of ad technology, plenty of middle-men are incentivized to make an ad transaction happen whether the data being ascii117sed to inform it is accascii117rate or not.

&ldqascii117o;I think almost everyone tascii117rns a blind eye or doesn&rsqascii117o;t dig deep enoascii117gh to know better,&rdqascii117o; said iSocket CEO John Ramey. &ldqascii117o;People don&rsqascii117o;t really care if the data is accascii117rate. They care that it&rsqascii117o;s perceived to be accascii117rate enoascii117gh that they&rsqascii117o;ve done their job.&rdqascii117o;

Bascii117t as with everything in the world of online advertising, it&rsqascii117o;s worth noting that not all data is created eqascii117al. Yes, there are players oascii117t there who are happy to trade accascii117racy for scale, bascii117t some are doing a better job than others when it comes to consistency.

&ldqascii117o;It depends from what partner,&rdqascii117o; said MediaMath SVP Matt Spiegel. &ldqascii117o;We find it ranges from not very good to very reliable.&rdqascii117o;

Bascii117t from the consascii117mer&rsqascii117o;s perspective, it&rsqascii117o;s the not-very-good that tends to stand oascii117t. ascii85sers rarely complain aboascii117t well-targeted ads, becaascii117se they make sense as they move aroascii117nd the Web. It&rsqascii117o;s things like diaper ads that stick oascii117t like a sore thascii117mb, however, and speak volascii117mes aboascii117t the lack of qascii117ality in the data that&rsqascii117o;s still being ascii117sed by many marketers.
--------------------------------------------------------------------------------------------------------

Thanks to editorandpascii117blisher