Obviously b2b print publishing has been hit just as hard as the consumer side in recent years. Business information ad pages are down about 30% across the board in the first half of 2009. And online, many of these niche publishers are just trying to get up to speed after years of digital obliviousness. To be sure, there are exceptions. The multi-title tech publisher IDG, for instance, was online hard and early, and now has many a highly evolved network of sites and databases. But across the many nano-niches b2b represents, there is a massive amount of valuable user data waiting to be aggregated and leveraged at scale. There have been many attempts to do just that over the years, including Web-based B2B portals like VerticalNet, which tried to serve many trades at one site and then network advertising across the niches.
The latest attempt to aggregate data and serve targeted advertising b2b publishers is Bizo. The company claims to partner with 300 publishers like McGraw-Hill to gather highly granular data about business users, from the company they work for to their job type. Bizo says it is targeting not just trades but people in what it calls "Bizographic" profiles, a kind of massive database of anonymized business cards. "It's a combination of things," says CEO Russell Glass. "It is firm data -- how large is it and where is it located and what [fields] is it competing in. And is it psychographics on the individual -- how senior , what job function and educational level."
Glass says the system has about 45 million targetable uniques and is about to serve over 1 billion ad impressions a month with clients like UPS, J.P. Morgan and American Express. At Bizo.com, the company maintains an ongoing view into the data sets and how it breaks down. As of today, for instance, about 5 million in the pool are from the Media and Internet segment, while 3.3 million are from Education. 21 million are in "Micro-sized" businesses and 3.8 million in "Extra Large" firms. About 24% are C-Suite executives and .4% are from legal departments. About 5.5% hold doctorates, 11.76% have MBAs and .52% are MDs. Bizo categorizes over 110 industries for targeting and 50 job functions. "We can target cardiologists or pediatricians or a company size from Fortune 500 to micro," Glass says.
Glass says that Bizo amasses such granular detail about audience by aggregating multiple data types. General site data can render IP addresses that often can be sourced to specific companies. Many of the publisher partners are also collecting registration data from their qualified audience. "We take the non-PII [Perosnally Identifiable Information] parts of the profile, like company names, domain, titles, and we anonymize it to further classify it," he says. "Microsoft doesn't look like Microsoft but like a Fortune 100 software company, and that is what we store and allow to target."
And so the idea is to bring all of these niches together and leverage the data in multiple ways. Small-business publishers can monetize their own qualified audience by contributing data to the system and get paid when the network targets them elsewhere. B2B publishing partners can also use the network as an inventory extension to sell their own clients into larger audiences. The data can also be used with existing data providers and networks. Bizo partners with BlueKai and AudienceScience, for instance. Glass admits that the classic limitations of a B2B network remain. With nano-targeting comes the challenge of scale. "If a marketer wants to put DMA targeting on top of Bizographics, you can get down to some small numbers. DMA targeting is too narrow."
Glass claims that so far his targeting is proving more effective for advertisers like UPS, J.P. Morgan and Monster than many of their other ad networks. When messaging is specifically tied to data points like job types the results can be especially strong. "UPS can run campaigns where the messaging is specific to time-critical and sensitive shipments and target that to people and professions who have time-sensitive shipments," he says. "The reason this works is because the data is not guess work," he says. That is, you aren't inferring an interest because of a past online behavior, but working from a granular profile that targets users' real business concerns and job responsibilities.
But that targeting doesn't mean the specific audience will respond in the usual and usually measured ways. Among the most interesting learnings from running a B2B network with such a granular understanding of audience is just how irrelevant standard ad metrics can be. A highly targeted viewer is likely to recognize and record the ad impression, but may also be more likely to take an important action on that ad later.
"We have learned, for instance, that CFOs don't click," says Glass. As you come to know your audience more intimately, you can see different behaviors and responses to ad campaigns. That deeper level of understanding has to be built into the metrics. "When we run campaigns that target CFOs, they are not clickers. It doesn't mean they aren't getting to that advertised Web site and learning more about that brand. But you need to measure in totally different ways. You need view throughs to see that someone has seen an ad and converted later and that you can attribute it to that campaign," he says. "Everyone has realized that clicks are largely fraudulent and a lot of activity is happening after seeing an ad. Any time there is a brand name in a search box, some other advertising channel deserves the credit for that."