Collective Intelligence and Analytics
In today's collective economy where we see diverse applications and heavy information consumption, it is imperative that we seek patterns and linkages in data.
In the last few years, our ecosystem has experienced massive changes owing to the collective market forces getting stronger. Malcolm Gladwell established the flat world; Zuckerberg established a highly networked one! And I couldn't agree more on how networked, collective and multi-dimensional our lives are today! Look around you within your respective zone of operation to see if you mirror these seemingly non influential impact waves such as bulk download of mobile applications as guided by friends, "liked" more than ten posts a day on social sites, have been bombarded with emails and newsletters, have reviewed more product portfolios, you are "the collective atom" (as I call it) of the social era.
Who is a "Collective Atom"?
We all derive intelligence in multitude and magnitude by viewing content on Facebook, newsletters, updates from LinkedIn, advertisements on apps and the web, HD television, newspapers etc. We do not necessarily seek this information yet receive it from various sources in various formats. A collective atom is a system that is a receptor, filter and influencer of information. Collective atom is free to respond positively or negatively to every binary that is consumed. Everyone's responding, observing and influencing- I like / I dislike/ I post / I share. In other words, we all are collective atoms. We are constantly connected on LinkedIn or Facebook. "Someone posted a picture of a dress - I like it too, someone shared a song but was not my type, a new game has 200 good reviews and so I need to download it too, lets attempt this online shopping, the movie got boring towards the end, Gangnam video is fantastic" and so on. Information is flowing everywhere and one seems to be influencing the other! But not all of it seems applicable and relevant. There is loads of data out there that need not make sense to one, but is extremely critical for the other, thus making data subjectively intelligent.
A friend of yours "likes" an advertisement of paintball on Facebook. What does that tell you? Does he like "the paintball as a form of entertainment" or does he like the "advertisement"? Do you want to gift him a coupon for a paintball event on his birthday or do you want to share similar whacky advertisements with him to please him? However, if I choose to go through his earlier posts, I realize that he has already shared further similar posts about the game. Now this gives me the confidence to decipher that here is a fellow with similar liking for paintball as mine. I "derived" this intelligence and used it to my benefit when the two of us had a gala time at the sport later.
While the above is a very tactical example; the same is applicable in our areas of work as well.
Recruitment is one space in an organization where we see loads of information seeping in. The top of the funnel is always heavy. Recruiters typically download at least around 2000 candidates and add them to database every other day in large organizations across the globe. An Applicant Tracking System (ATS) is then used to filter duplicate entries from this bulk download and remove them. This clearly is very basic information. However, analytics enabled ATS should have derived intelligence that tells us for example that a candidate is a duplicate entry (basic input), rejected earlier (basic input), rejected due to reason (basic input), yet, is a good candidate due to reason and can be considered for another role (detailed input), and as an organization, we already have x% candidates who are duplicate entries, yet, highly probable candidates for future requirements. Having this input, recruiters need not download and reprocess resumes on daily basis. Advanced analytics and derivations from recruitment systems through Big Data should enable organizations to:
- Reduce the spend on time and effort to source from job boards for each new position by highlighting candidates in existing system
- Predict joining probability of candidates there by reducing cost of hire
- Establish indicators on market demand of the organization by tracking number and quality of applicants; and many other indicators linking data points to establish structure and impact organization effectiveness.
Another example in effectively driving employee value proposition through analytics and derived intelligence is that of a talent management helpdesk. We have all seen dashboards that give operational data points like: # of employee queries received v/s answered; First time resolved v/s transferred; Net Promoter scores etc. If the same system is equipped to classify and link data points like: Employee named Nick raised a repeat query on career progression and role change in 5 different organization forums, across the last 4 months, followed by requests around exit policy, then as Nick's manager, I can fairly establish a deduction that he is keen on career progression or external movement. Such a data point can help managers plan and manage career progressions better.
However, today this information remains scattered across employee helpdesk, HCM, Talent management software. There are several such data linkages that can be established within systems to convert data into intelligent systems and influence organization effectiveness.
In today's collective economy where we see diverse applications and heavy information consumption, it is imperative that we seek patterns and linkages in data. Impactful analytics has huge potential to drive derived intelligence. Human resources function needs to evolve to establish these data patterns and create newer, finer systems.