Data Analytics in MSP-Based Staffing
According to Webster’s dictionary, data is defined as “factual information (such as measurements or statistics) used as a basis for reasoning, discussion, or calculation”. And there lies the pith of its meaning to the times we live in. In the last few centuries, we figured humans’ five basic senses are all measurable and thus generate data! In last decade or so, new fields of Big Data / Data Science have come into being.
In staffing operations, the business philosophy around a successful recruiting venture is that it’s a combination of Art and Science. ‘Art’ can be said to encapsulate the flairs and strengths of business development, client relations, communication skills, social networking, etc. ‘Science’ helps determine where to apply that ‘Art’ and what to focus on within of that ‘Art’. Art and Science here are synergetic and contention is that quality data and its apt application sharpens decision-making around ‘Science’ part of business.
There’s no end to the variety of data analytics one can go about assessing around Staffing business. For sake of this piece, we’ll keep it around MSP-based staffing. If your operations happen to be engaged with a handful of MSPs where clients have nation-wide footprint / Fortune 1000 listing, you’re mostly dealing with plethora of job orders cutting across many verticals. In order to harmonize business resources with all that incoming business, you need quality data analytics that you can use to take affirmative actions.
Some basic-level data metrics are:
Externally generated:
- Average job orders received from client in defined time-period (day / week / month)
- Average job orders received in a given labor category from client in defined time-period
- Average job orders successfully filled by client in defined time-period
- Average job orders cancelled by client in defined time-period
- Average job orders internally filled by client in defined time-period
Internally generated:
- Average candidate submissions in defined time-period (day / week / month)
- Average interviews per defined time-period
- Average offers per defined time-period
- Average hires per defined time-period
- Average profit / spread per hire
As mentioned earlier, you can extract myriads of data metrics around business. Take any pair of above, and you are looking at a ratio e.g. (2) divided by (1) is your Average Submission-to-Interview ratio.
So, next step is to establish a set of guidelines that can help effectively go about the body of data metrics you want in place. Here’s a few:
- Internal Score carding: Once a business has reached a certain scale, principles of time-management don’t afford you the leeway of hearing out every individual in workforce. A well-defined internal scorecard structured around a set of data metrics helps quantify and thus qualify workforce and their productivity.
- External Scorecarding: MSPs often use scorecards, essentially a tabulation of a bunch of data metrics that client considers to be of significance for success, for rating vendors. While a strong scorecard rating has benefit of bringing laurels to vendor from client, converse side is that a series of mediocre scorecards are used by clients for vendor streamlining i.e. to remove vendors they consider non-performing.
- Continuous Improvement: In order to lend meaning to this concept beyond just being an industry buzzword, you need data that you can act on in order to target / design / effectuate / measure the building blocks of framework of continuous improvement.
- Removal of bias: It’s human to be biased. Data metrics are one of your best tools to mitigate the ill-effects of bias in decision-making. Based on modus operandi of business.
Data metrics are generated from data, which in turn is generated, on a tapering scale based on automation in place, from human input. Quality of data metrics is directly correlated to quality of input. It behooves upon managers to continually educate and remind workforce (base-level creators of data) of, to borrow a term from Computer Science, GIGO (garbage in, garbage out).
Remember, there’s always a risk of getting inundated in the ocean of analytics (read ‘paralysis from analysis’); one could end up tending less to the business itself while spending more time & efforts on its measurements.
Take stock of status quo of business to select / build your data metrics. As status quo changes, so should the kind of data metrics of interest. As example, if Hiring Source costs are on the rise, that’s a strong reason to build data metrics around costs of different Hiring Sources (job boards / social media / networking / referrals, etc.) and their ROI.
Though it has taken to to the end to make mention of this all-important but matter-of-fact truth, data analytics alone wouldn’t ever do the trick! Their main role is to lend you a window of optics into opportunities / challenges surrounding the business.
Karan Bandesha, Director of Operations