Projected age of retirement
for current workers.

Data that is false or fabricated.

The best data scientists turn
distilled information into pure gold.

Too much churn and
companies lose the cream.

Guatemala has the largest CW
compared to population in Americas.

1 in 3
# of working Americans in
the contingent workforce.

As some jobs become out of date,
others emerge.

In a conformity string, we call attributes
that impact cost and availability of
qualified job candidates "pieces of work".

Projected growth office/clerical
staffing 2013.

Companies implementing proper
measures during offboarding.

Singapore was world's top CW
productivity market 2014.

Data Scientist: the most wanted
job by employers on LinkedIn
in 2014.

Belgium has the highest tax burden in EU.

Ratio of robots to employees in Korea,
highest level in the world.

Employers who find paying
freelancers cumbersome.

The big star in our universe is Data Centauri.

% of American workforce projected
to be freelance by 2020.

Predictive analysis is only as
insightful as the analysts.

Data should never be sugar coded.

A good strategy stretches without
changing its basic shape.

Average length of unemployment
of managerial candidates.

# of workers with tenuous
ties to employers.

% of senior HR officers identifying
talent management as top HR issue.


To find answers, we formulate questions.
Then question the questions.

< 20
% of private sector workers receiving
employer sponsored health insurance
by 2025.

CW population at average
large company.

France has the highest
tax burden in EMEA.

% of Fortune 100 who’ve
implemented a VMS.

Shortage of US managers able to
analyze big data and make decisions
based on findings.

Amount NHS spends on
temp staffing.

Independent contractors can
be reclassified by Irish courts.

Talent Tech Labs: October, 2015

by Jason Ezratty

More Than a Trend — Big Data is Poised to Make a Bigger Impact

Big data has achieved celebrity status of late as a harbinger of a more empirically-optimized future. The applications within talent acquisition are profound, most specifically focusing on better and faster ways of aligning job opportunities and job seekers.

The previous wave in talent acquisition analytics has been about broadening the data set for visualization. Looking forward, the next phase will be focused on gaining depth of understanding. Expect the next generation of data software to enable businesspeople to focus on the business meaning resulting from data and statistics rather than requiring them to be statisticians themselves.

These four trends should be on your radar as we await the next wave of data analytics software and solutions for talent acquisition:

1. Taxonomic conformity.

In human capital data there is no universal standard that all organizations adhere to for master data standards (the labels we place on our data). While data elements like location have a universal standard of nomenclature, elements like job titles, industry and source type are often a mixed bag within a single organization, let alone across multiple businesses.

Emerging human capital data platforms are addressing the issues of conformity head on, by looking within requisition data elements, including job descriptions and resumes, and using natural language processing to more granularly and precisely define a taxonomy as seen in the actual data, with even greater detail.

2. Highly filterable aggregated market analytics.

A core but often misapplied diagnostic method is to benchmark one’s own performance against that of other organizations. Determining which organizations and records to include in your data set for a given analysis is actually a more complicated issue than performing the analysis itself. Should you only include companies of the same size? Okay, but size measured how—revenue, employees, total workforce?

The tools of conformity described above offer tremendous opportunity here. Now, we can rethink the idea of trying to find data points that feel like “needles in a haystack” and instead focus on a highly fluid set of data that can be as broad or specific as you like, and only containing needles. This also includes the ability to use data across multiple organizations—so you can find all the relevant data to your organization, depending on which parts of the market are most relevant to you under which circumstances.

3. Total workforce management.

Given the fluidity with which managers turn to employee and non-employee worker categories these days, it would be incorrect to exclude either from the other’s analysis in some way. If your employee headcount is remaining steady due to a hiring freeze but your contingent population is rising, is it accurate to say the hiring freeze is working? Total Talent Management (TTM) is an emerging perspective that looks at the collective components of a workforce through a single lens. Accordingly, data science tools are critical to comparing the “when” and “how” of workforce mix optimization, and can offer guidance to end users directly within the instance of their raising a requisition.

While these techniques can be applied to any aspect of human capital analytics, in talent acquisition, predictive analytics has predominantly centered on matching algorithms, matching candidates to requisitions (for optimal placement), requisitions to requisitions (for conformity of grouping as a “market” inference) and job descriptions to job titles (for automated classification). There are many techniques in predictive analytics and none are easy or universally work best. Use with caution until the methods and interpretations are more mature than the hype.

4. Tying in performance.

Performance and/or quality of hire data (of candidates, workers, or suppliers) is still where things are most immature. There are companies with methods of looking at volume of work in certain fields (like programming) but it is an imperfect and indirect measure of performance or quality of worker. Without this, while we can look at a how long or difficult it is to bring in talent we will never know which were worth the additional time, effort and expense. It’s a big gap and one that will likely smack against privacy sentiments.

Seems hard, why bother?

Data science applications are in their infancy but are making great strides. As with computer technology before it, the next wave of analytical innovations are going to start removing the technical burden on businesspeople to more directly provide the answers they have been seeking