Who’s Zooming You?


      Bookmark and Share

The other day I was trying to track down someone I haven't seen in about 20 years. I tried Google and, shockingly, received too many results. The two links on the first page of results that I clicked and skimmed through were old and not terribly useful.

Next, I tried ZoomInfo. I actually find this service to be a bit spooky. It automatically culls bios of people from the web. They have six bios for me. One is fairly accurate, but another implies that I'm the editor of a different magazine, while others suggest I may work for some of the companies in the digital content industry I write for—because I'm quoted in their press releases.

Ultimately, however, the ZoomInfo profile for the woman I was trying to track down was a good deal more useful than the open Google search. Though I knew it was far from gospel, it provided me with a clue that allowed me to track the digital breadcrumbs to the appropriate inbox.

Then my phone rang. (I'm not the only one who is looking to kindle connections.) Terry Frechette, a PR rep from Lois Paul & Partners, which represents a number of companies we cover, wanted to check in on a client I've not heard from in a while. You guessed it: ZoomInfo. Timing is everything. Either Frechette is excellent, or ZoomInfo can tell when I'm searching its site.

In any case, Frechette put me in touch with the company's COO, Bryan Burdick and we chatted about the expansion of the ZoomInfo domain. First, however, I wanted to sort out my multiple personalities. Turns out, I'm what ZoomInfo considers a "celebrity entity." Look ma, my fifteen minutes! Unfortunately, it isn't as glamorous as it sounds. Burdick says, "What we mean by that is you are someone who has a lot of web references."

With its semantic search engine, which crawls the web and extracts information about entities then partners it with other information on that entity, Burdick boasts that "eighty-five percent of the time, we've got the right person at the right company; we have extracted the correct information." I don't know about that, but then again, we celebrity entities simply can't expect the same treatment as the masses.

According to Burdick, ZoomInfo would "rather err on the side of having multiple profiles rather than a combined one that is inaccurate." The company does have a system in place in which users can create or "own" their profiles, so that regardless of automatically generated info, an accurate profile exists. Burdick says that in the year he's been with the company, the number of people claiming or creating profiles has climbed from less than 1,000 to 10,000 people a week.

The company invests most of its energy in improving its algorithms, to increase the accuracy of its results. Of late, however, the company has also done some thinking about what other information its crawlers should be culling. Burdick says they wondered what other content business people seek. The answer: "They are all interested in their careers."

The company didn't build a jobs board; rather, it's working with "a partner who was willing to give us their API," according to Burdick. "We've taken their jobs and matched them with our knowledge of industries, markets, and companies to create an unusual job-list offering," which the company launched in early April.

With the semantic approach, a user can do what Burdick calls "pivot searches" like searching for, say, a small software company in Boston. Users search on keywords, then click on another button—and there are the jobs. Then users can cross reference to ZoomInfo's other information, like who works where, company data, and so on, to help narrow down places they might actually want to work.

ZoomInfo plans to do more than just enhance its content offering, though. In addition to wanting to know more and find better jobs, the company also believes its users want to get more accurate information faster. So, in late April, ZoomInfo also implemented a new database architecture. The key impact will be that the process in which its crawlers cull content, dump it into databases, then churn and combine it before making it available will be reduced from a month to a mere three hours.

The improved speed will allow the company to take steps to significantly impact the accuracy, according to Burdick. He points to another celebrity entity, Bill Gates, as an example. "We'll be able to create algorithms that will rank data in different ways. In the case of Bill Gates, we have data from an EDGAR filing and data that he is a judge for the Webby awards. We can be sure that the new info—from the Webby awards—wouldn't automatically supplant his position as chairman of the board at Microsoft."

Alas, with ZoomInfo zooming in on accuracy, my celebrity status may not even last the full fifteen.