Data: The Big, the Bad, the Beautiful

Modern Hire

Big data is more desired than ever, its promise of access to the insights that lead to business advantage irresistible. Recruiters are increasingly expected to leverage big data in their talent acquisition efforts. But the question of where to begin is enough to discourage even the most forward-thinking hiring experts. After all, by definition, big data = lots and lots (and lots) of data.

Amid the daunting claims and clamor, know this: bigger data doesn’t always mean better data.

One of the central challenges in any big data initiative is navigating the enormous landscape of bad data. Any attempt to gather candidate intelligence as an HR practitioner will inevitably result in collecting information that isn’t relevant to what you’re trying to accomplish the data is random and inconsistent and lacks conformity across all candidates. The data is often collected passively from the candidate or harvested from social media or LinkedIn profiles, resulting in a mixture of good and bad data. Your mission is to discover how to uncover the relevant and meaningful information – the beautiful data.

The successful journey to beautiful data begins with experiment design. You first must create a data collection and scrubbing methodology to differentiate the meaningful from the meaningless, the useful from the useless. Recruiters too often rely on parsing technology to look for job titles, experiences, and other presumably relevant keywords in a candidate’s resume. Distorting such data further is that candidates are often aware of how to optimize their resumes and social media profiles for such analysis. Clustering and organizing candidates by specific terms they may include in their resumes is a primitive evaluative method that will provide little, if any, evidence of what a candidate can bring to a role – meaning, you’re collecting bad data.

How do you get to the beautiful data? Implement a process of data capture that gathers the least amount of required information in a format that facilitates ease of analysis. Analyzing resumes or LinkedIn profiles for arbitrary terms like customer focus or collaboration won’t tell you anything about the candidate’s expertise in customer service or working well with others. Instead, adopt an assessment methodology designed to capture specific information about a candidate’s work history and previous experiences.

What will give you more insight: Scanning a resume for the keyword coaching? Or asking a candidate what percentage of time they spent on coaching in their last position? What do you think is the most effective method of gathering evidence of a candidate’s supervisory experience? Searching for key terms in a list of skills, or directly soliciting information about how many people the candidate has supervised, how many they have hired, how many they have separated?

Consider similarly useful ways of evaluating a candidate’s approach to handling on-the-job challenges. Again, rather than relying on keywords to highlight experience, consider how presenting candidates with simulated role-specific scenarios actively engage them in making choices about how they would handle various situations, enabling you to collect relevant data about how they approach customer and coworker interactions. And beautiful data becomes even more beautiful when it is aggregated compiling and comparing candidate responses to create ranking is an easy way to see the meaningful differences among candidates, helping you determine who is likely to be the best option for what you seek.

Beautiful data is also consistent over time and across candidates. Give two recruiters stacks of the same resumes and instruct them to divide them into yes, no, and maybe piles how likely are they to end up with the equal piles? Using relevant, deliberately and methodically collected data ensures candidates are placed into the same piles regardless of who is doing the sorting or when they do it. Such methods also are more objective and fair, treating all candidates the same regardless of class, age, gender, or ethnicity.

As big data continues to get bigger, the risk of bad data grows just as high. Getting to the beautiful data is more urgent and crucial than ever. To obtain the relevant information you need to enhance your recruiting process, you must develop a purposeful data collection strategy, one that is relevant to the job you want to fill and gives all candidates a fair opportunity to demonstrate their expertise. Intentional data collection not only improves the candidate experience, but it also delivers the beautiful data you want at the heart of your recruiting routine to make it a process that is better for candidates, better for recruiters, better for companies.