Dirty data, A.I & recruitment.

If one thing has become clear to most in 2019 it's that companies are collecting much larger datasets than most could imagine. This data brings both opportunity and responsibility. Opportunity comes for those who are able to put it to good use and secure it correctly. We all know what happens to those who don't.

The challenge.

Gathering data is the easy part especially if this very act is fundamental to your business model. The hard part is firstly how to store and use that data in a way that is both in line with regulations and permissions correctly.

Once this first hurdle is passed the next step is labeling or tagging said data in a way that is both efficient enough so that it isn't hugely labor-intensive but also useful to the business in the future. An example of this would be fast-paces sales environments where hitting targets are the primary goal and data labeling isn't a top priority.

Now this may, of course, vary depending on which company you speak to but 'dirty data' is sighted as the number one challenge faced by data science professionals in 2019.

According to the report from MMC Ventures into the State of AI 49% of teams agree.

Dirty data is the biggest issue because we are coming from an era of data being something we use reactively to make a phone call or send an email (in the case of recruitment) to it being seen as an opportunity to gain valuable insight into the business.

Example, a CV is uploaded to a CRM or ATS without being tagged with the relevant salary indicators. Future queries by data scientists or analysts based on 'job_title', 'location' and 'salary' would exclude this result from the list. Dirty data and skewed insights.

Another key stat from the above and a clear blocker for teams looking to do anything with data is getting the support to do so.

37% of respondents cited the lack of managerial or financial support for their initiatives. This alone raises the issue of the need for education surrounding data science and AI so executives can understand the potential upside of investment into the area.

The opportunity

For incumbents with historical data sets numbering in the millions, retrospectively going back and cleaning data to make it useable is a painful task but one that is becoming increasingly unavoidable. That said few are making the switch decisively.

For new startups, however, putting data integrity as a top priority will provide a disproportionate advantage. Customer relationship management systems and applicant tracking systems are only as good as the input, bad input bad insights. Good input + data science person/team, big win.

Early education of staff on the business value of good data combined with clear processes to follow will lock-in long term upside value.

Clean data combined with an in-house data science team is the ideal situation for any data-centric organisation but failing that engaging an outside entity like Peak.ai is a good place to start.

What about AI?

AI technology is important because it enables human capabilities – understanding, reasoning, planning, communication and perception – to be undertaken by software increasingly effectively, efficiently and at low cost. - David Kelnar.

As things stand a business without any implementation of AI or machine learning is stagnant and relies purely on changes implemented but human input. On the flip side of this, active AI and machine learning systems implemented in a company allow for continuous insights to be identified and acted upon instantaneously.

This could be used to remove manual time-consuming tasks, serve up recommendations or insights based on the query of the user.

In a recruitment setting, this could equate to the system identifying a candidate based off of previous placements and an email automatically served up with the appropriate details attached.

Another application could be populating consultant task lists with automated options attached such as candidate follow-ups, client contact, missed data requests, old CV requests, key stakeholder report generation to name a few.

AI has now crossed the chasm from early adopters into the early majority with over one-third of enterprises having deployed AI in some form.

Artificial intelligence in the unfair advantage of any business that implements it effectively according to its use case and now is the time to start exploring it as a technology if you haven't already.

Because someone somewhere is.

#artificalintelligence #datascience #data #ai #recruitment