Who Else Wants Big Data for Recruiting from WANTED?

I first fell in love with WANTED Technologies in 2007 after a demo in a dinghy shared office in New York City. For 10 years, long before the buzzword existed, the Canadian company (now public) has provided Big Data on job demand and more recently on candidate supply. About 10 percent of the Fortune 500 and nearly every big Silicon Valley brand already use WANTED Analytics. It’s the clearest example of what Big Data can really mean for HR.

Eight years ago, I was already arrogant enough to expect all but the largest vendors to come to my home office in Westport, Conn., to brief me. But a WebEx demo that seemed magical (presented by what I thought was a start-up) put me on the commuter train to New York City, which truth-be-told, is only an hour away, plus a five-minute subway ride to the then-dinghy shared offices in Silicon Alley of WANTED Technologies.

What I saw, and didn’t have a name for at the time, was the first use of Big Data for HR. Naturally, it was for recruiting (then specifically for governments), measuring with startling precision the demand for very specific positions and skill sets everywhere in the country.

Later, I invited Wanted (as with COMPAQ decades ago, I won’t continue writing their name in jarring all caps) to be part of the second “Cool New Technologies for HR” breakout at the 2009 HR Technology® Conference, a session that also showcased Rypple for the first time.

Later, I renamed the session “Awesome” in a pathetic attempt to sound younger, turned it into a general session and now Steve Boese has wisely split it into two general sessions: one for disruptive start-ups (Zenefits was a prescient pick last year, considering the current lawsuit from ADP!) and a second session for new innovations from established companies.

Wanted’s clients now include every big brand name in Silicon Valley too selfish (or competitive) to let a smaller vendor profit from touting its business relationship, plus some equally good companies (10 percent of the Fortune 500) that are more generous, including Thomson Reuters, Intuit, Lockheed Martin, Unisys and Kelly Services. It’s a list consistent with its history and various functionality across different industry verticals.

What do customers get?

As recruiting expert and blogger Tim Sackett wrote: “… a complete picture of your sourcing environment, which allows you to build and plan the best strategy to attract the talent you need. What it delivers, in terms of data, charts and information, is completely insane!”

Chief Product Officer Jean-Pierre Rabbath, based in Quebec City (also once Taleo’s global headquarters and still the source of shamelessly high IT salary subsidies), built the current Wanted Analytics product and offers more specific details.

Wanted Analytics constantly spiders 15,000 corporate career sites and 10,000 job boards, aggregating about 15 million new and still-open unique job ads in the United States, all placed in its own very specific categories. In short, this reveals the demand in your area for the same jobs you (or your recruiters) might be trying to fill.

The company first started collecting job-advertising data in 1999, surviving its dot.bomb-era birth by providing sales leads to newspapers—which, 16 years ago, still ran substantial numbers of “Help Wanted” (hence the name?) classified ads. Wanted told them which companies in their region were hiring a lot, so they could sell to them. Still collecting that data, some of Wanted customers, like some staffing agencies, still use them for sales leads.

By 2005, it had created a vertical application for government agencies, and then, in 2011, the corporate recruiting app. Reflecting that history, about 4,000 individuals at 320 organizations use Wanted today: about 110 corporate recruiting departments but mostly other entities such as staffing agencies (like Kelly), media companies (like Thompson Reuters), 22 state and Canadian provincial governments and financial-service firms wanting to know how much hiring companies are doing as yet another data point for their analyses.

The Conference Board uses its data for The Conference Board Help Wanted OnLine Data Series, which is not quite the big report that keeps CB in the national spotlight every month, but an important service.

Rabbath says enough ads feature salaries (about 15 percent to 20 percent) to allow Wanted Analytics, using a rolling three years of data (representing 125 million posts) to display accurate line charts of salary ranges being offered for a position locally or nationally. Heat maps will point recruiters to other areas of the country where demand (or salaries) are much lower—if the company is willing to relocate new hires or let them work remotely.

What’s more, the recruiter can see which companies are advertising to fill the same positions, and how many have been open for how long, and can even read the actual competitive job descriptions! Naturally, Wanted Analytics first parses clients’ own job descriptions to make sure the comparisons are valid.

So tell me what else a recruiter needs to know to turn back to the hiring manager and say, “Fuhgedabboutit! We’re never gonna fill your open position at the salary you’re offering when we’re competing with Goldman Sachs and Google. Plus everyone else is taking 44 days to fill it, and you want it STAT?!? C’mon.”

Figuring out demand is a relatively straightforward Big Data process of gathering, normalizing, de-duping, categorizing and presenting the data—except for detecting and removing the evergreen postings, such as for customer-service reps at constantly hiring call centers.

“The demand side is based on facts and postings,” Rabbath says. Certainly not easy at such a large scale, but also not so tricky. Wanted has demand data available now for 22 countries.

Determining the other side—the available supply of employees (working or not)—is a lot more difficult. There, Wanted has had to make some assumptions and deploy the second part of Big Data: algorithms to determine some probabilities, which may or may not be right.

The first assumption is that when a job posting is removed from a career site or a job board, it means the position has been filled and someone has actually been hired, not that the company gave up! Pretty reasonable assumption, allowing the application to present a list of companies with the largest number of employees doing the job you want filled—and even include a four-year assumed “hiring” history for that position.

Combine it with a cleverly designed LinkedIn search by company name, and a recruiter already has “A Field of Dreams for Poaching!”

The Bureau of Labor Statistics provides worker numbers for 1,100 Standard Occupation Categories (SOC). Wanted has made those more granular and detailed by creating its own 11,500 SOC’s, using algorithms to add skills, and obviously not blowing out each one 10 times. “Software developer” may have 30 new SOCs. “Childcare worker” maybe only one or two more. Users don’t see the details of those—part of the secret sauce.

Wanted presents labor-supply reports on six increasing narrow levels with decreasing numbers. The first four levels are straight from the BLS: country, more specific location, function and occupation. Some or all of the next two—seniority level and keyword/skill—are derived from its algorithms.

Put supply and demand together, and Wanted displays its Local Hiring Scale, factoring in the number of potential candidates in your area for each job, the number of companies competing for them and the number of open jobs posted. And out comes two very stark numbers, from 1 to 100, indicating the difficulty of filling it locally and nationally.

Finding the actual humans represented by those candidate numbers is done by partners, currently Talent Bin (owned by Monster), Swoop, ZoomInfo and data.com (part of salesforce.com). LinkedIn used to be included until it told all its partners to get lost earlier in June and withdrew use of its APIs.

All of the others use various methods, including “social exhaust,” to create their own worker profiles, often reflecting the fact that filling IT jobs everywhere (not just in Silicon Valley) is the bloodiest sport in recruiting right now.

Wanted customers need separate subscriptions to each service to view their individual profile cards displayed within the Analytics application.

Wanted sells up to two seats on Analytics for $18,000 a year, exactly (by chance, Rabbath says) the per-seat price LinkedIn charges for Recruiter: $9,000. Wanted offers deep discounts above that number ($36,000 for up to 10 seats), while, to my knowledge, LinkedIn offers none. They are clearly in different areas of the talent acquisition business.

So wouldn’t you think the ATS vendors would be falling all over themselves to partner and integrate with Wanted? Well, you and I would be wrong.

Rabbath, who also heads business development, has not vigorously pursued partnerships. Wanted has done them, by customer request, using the public APIs to integrate with Taleo (just four or five times), IBM/Kenexa/BrassRing (twice) and even once with Avature whose ATS is relatively new and lightweight.

But with a new CEO hired six months ago and various reorganizations, Wanted will likely be wanted by a lot more of these companies very soon.

HR Technology Columnist Bill Kutik is co-chair emeritus of the 18th Annual HR Technology® Conference & Exposition, with this year’s agenda available online. Listen to The Bill Kutik Radio Show® for his provocative interviews with HR thought-leaders, now featuring Gartner’s Ron Hanscome and his best advice for choosing among SAP, Oracle, Workday HCM. Watch the fifth episode of his new video series, Firing Line with Bill Kutik® featuring Josh Bersin. He can be reached at [email protected].

Copyright 2015© LRP Publications