9 to 5
Yesterday’s workday.


turkeyandchicken
Turkey inseminator and chicken plucker
are two of the more
unusual job titles analyzed.

66
Projected age of retirement
for current workers.

6 in 10
# of Millennials with jobs.


bone
If metrics are the bones,
data is the marrow.

13
% of workers engaged at work.

8k
# of boomers turning
65 daily in US.

156m
# of people in US labor force.

24
# of newly defined detailed
occupations in 2010 SOC system.


dollarsign
On average, Canadians earned 33.18
per hour in 2013.

1/3
# of buyers indicating no assignment limits.

+9%
In 10 years, projected growth
for on line staffing services.

$70b
2013 global MSP spend.

Staffing Industry Analysts | CWS 3.0: April 6th, 2016

by Jason Ezratty

Staffing firms are drowning in data. Skills, rates, previous jobs, upcoming jobs, client industries and more are the pieces the workforce solutions ecosystem has to deal with. But not all of that data is useful, or even accurate, and that makes putting it to good use challenging.

Talent Data Exchange (TDX), a membership-based human capital data analytics platform and service, aims to solve that problem. Jason Ezratty, Brightfield Strategies’ president and the force behind TDX, explains in Part 1 of this two-part series why data is so important and how TDX can help companies get the best insights from their data.

What is the role of data in our industry, and why has it become increasingly important?

Data has long had a presence in the industry. Vendor Management Systems (VMS) and Applicant Tracking Systems (ATS) provide visibility, the first step in knowing which workers we have, how many there are, where they are, and which suppliers they are with.

Companies using a VMS have always had more data than they know what to do with. The level of importance of the contingent workforce program has elevated.

Slicing and dicing data using traditional business intelligence methods was enough to whet the appetite. Now people are saying, “Great, now let me make conclusive decisions, like I can in finance or marketing, that help me make money or save money.” Those decisions are failing because the quality of the data and the way that data was used was insufficient.

What constitutes a state-of-the-art data science solution?

The ideal solution has three separate pieces, only two of which are broadly available right now: 1) properly curated and conforming data, 2) appropriate statistical methods and calculations, and 3) good visualization and presentation. Currently, most solutions are missing that first piece, so you’re going to run into problems and be unable to come to conclusions.

Prior to developing your workforce analytics platform, how did Brightfield measure performance, given challenges such as lack of data conformity and job taxonomy?

We still had the challenges, but they were less because we were dealing with only one company’s data at a time. The conformity challenge becomes exponentially harder when you’re trying to marry the job taxonomy from one company to another.

We started looking at just one company’s data. TDX was an outgrowth of that experience. The “aha moment” was when we realized that we’re not just talking about job titles — we’re talking about job titles plus a bunch of attributes that we auto-discovered from job descriptions. Our technology looks within the job description and pulls out the variables of what defines this particular type of work.

How would you describe Talent Data Exchange (TDX)?  What was the catalyst for its conception?

TDX is not a single tool, but it enables a suite of different tools — some focused on talent acquisition and some on contingent workforce program optimization — for a total talent view. The heart and soul of it is the way we curate data. Before anyone performs a calculation or visualizes the data on a graph, we have to make sure the data is worthy of those steps. We have developed technology to make sure we know how to connect different pieces of data before we start graphing things.

TDX is a place to make sense of the noisiness of human capital data. How much quality signal do we get vs. noise? That’s a huge problem in human capital analytics. We developed techniques to get higher confidence in our conclusions that we’re recommending to our clients, whether it’s a business case showing how much they could save or a comparison of two locations.

We are membership-based, meaning members are contributing data from their programs. No one knows who is in the pool of data and no one can see any one company’s data.

What are the reasons companies are looking to products like TDX for developing and refining their workforce strategies?

They’re realizing there is value to be had in increased precision and accuracy. Companies can answer questions such as how much does a database administrator with these five skills cost in San Francisco, or it could be someone from finance looking at whether contingents are more expensive than employees. (Yes, but only sometimes.) Another area is looking at the overall sourcing model: are there more cost-effective ways to do it? We’re constantly looking at the spending behaviors and talent acquisition behaviors to find patterns in the data.

What are the most common pains TDX can help companies address?

We can help with the comparison of contingent workers to employees. We can give an employer a comparison of what they’re paying compared to what is being paid by others in the marketplace: Am I above or below? There are dangers to both.

An employer can investigate different scenarios: What if I change my spending behavior? We can model a specific scenario.

Another thing we can answer for people: What are the factors that drive rates? Some people pay more than others for the same thing — the question is why. It could simply be that a recruiter got away with charging more. But there could also be another factor — it might be the industry you’re in, the speed with which you’re expecting the talent to arrive, the size of your supply base relative to your overall spend, or the duration of your contracts.

We can also look at the factors that affect time to fill positions: If we found that one company was much faster at filling positions and they weren’t paying more, we could look at the way they’re running their shop. Do they have fewer suppliers? A better relationship with each of them? More predictable demand so people are better prepared? 

 View on the Staffing Industry Analysts website