Internal Big Data decision making
As we continue to upload, type, share and work on digital spaces, a digital trail is laid of our movements. We have looked at how big data analytics are important tools for consultants in order to predict customer behaviour and improve supply chain management. But can big data also be used within companies themselves? For consultants making recommendations to business organisations about how to make them more effective, change in the digital sphere does not necessarily have to be focused on the external factors.
In business decision making about internal matters, big data could play a key role. Using the vast amounts of big data that is stored within a company about it's internal workings, we are able to peer into the 'business beehive' and start to consider, and try to understand, the ways in which the worker 'bees' are performing and impacting on the success of the business itself.
Businesses are starting to recognise the power that big data can have at optimising businesses from within - KPMG has appointed a department in 'decision science'. Bill Nowacki working for the firm in London works to try and deduce how the employees in a business work from following the data trails they leave behind. These 'artefacts' as they are known can give great clues to the overall 'performance puzzle' of how employees' behaviour patterns make them more or less successful at their tasks and results. Big data algorithms created by people such as Bill Nowacki in the decision science department at KPMG can start to figure out who works best. Then, by knowing the intricate data details of who works best, a profile of the optimum employee can be created. This is useful in several ways.
Firstly, by knowing the key ways in which the best employees complete their tasks, this knowledge can be used to help review and continue to train others within the business to work to the same performance targets. This kind of knowledge improves the organisation as a whole internally. As the saying goes, 'you can't manage what you can't measure' and this phrase means so much more that we are now able to measure everything that a person does in the digital sphere – the sphere where more and more business is being conducted. With the big data that can be collected, managers are now able to have a clearer insight into what all their employees are doing at all times – what time they log into their computers, where they go within the building with their keycard, what websites they use, what media channels they effectively deploy and so on. Big data capability means that daily, hourly, even minute by minute digital usage can be data mined and analysed. It may seem like trivial aspects of working behaviour at the time but it is measurable. Managers can transform the information that they can gather from big data in order to improve decision making and therefore performance.
Next, using the algorithms giving information about the key attributes of the best workers, recruitment can be made far more effective. Would managers even need to be involved? Could an algorithm complete the hiring and firing process all by itself? By using the number-crunching power of an algorithm to assess the people who are excelling within the company already, new recruits can be compared with the already established outstanding performers. Some are big advocates of the algorithm recruitment option such as Sabre that has been trying to create specific questions and aptitude tests using the data of current high performers that allows the filtering out of applicants before their information has even been cursorily looked out by a human in HR. With the current job market being so globalised and competitive, thousands of people can apply for a job posted on the internet, particularly on a site such as LinkedIn where your CV is already visible. Big data specialised algorithms are like the 'online dating' process for businesses – managers can input the information from their internal analytics about their prime workers and the algorithm will find a good 'match' out of all the applicants.
But is it right that you should have to impress a computer rather than a human being? Although algorithms can 'learn' as they continue to mine more and more data to increase the accuracy of their predictions, they cannot interpret creativity or originality very well at all, for it does not fit the mathematical equations they work to. Some say that the algorithm process is unfair to those who are new to the workforce and may not be able to write applications that fulfil the keyword criteria or have practice at the aptitude tests, algorithms finding proofs of already held aptitudes rather than potential. Others rally against the depersonalisation of the recruitment process, not allowing for the individual consideration and intuition that can be gained from reading CVs and conducting interviews.
Using algorithms though does not mean that it has to be at the expense of any human interaction. Interviews can still be conducted after the filtering process has been completed. Realistically there are filters already in place by HR departments and the preferred answers to questions, the algorithms simply streamline the process, making it easier and more accurate based on real mathematical data. Robotised recruitment also has a great advantage of eliminating forms of bias and discrimination that human influence can have. It is a well studied fact that people hire what they know - white middle aged men will be more likely to hire other white middle aged men. A huge variety of variables could come into play in an interview to sway opinions such as similar mannerisms, shared hobbies or noticing the way that an applicant speaks. Within 30 seconds of walking into a room many interviewers already have a degree of intuition about the person that will develop as the interview progresses. Being totally blind to this type of human influence, computerised processes could be seen in contrast rather like a recruitment version of the talent show 'The Voice' where the panel of judges have their backs to the singers, relying purely on their talents rather than first impressions. Algorithms will pick out those candidates that fit with the profiles of the most high performing current employees, whatever their other attributes are.
Finally, if the algorithms can be used for new employees in the recruitment process then it makes sense that they should also be used for the corporate hierarchy in determining the next leaders. The performance indicators of their work or the attributes they should need in order to continue to improve performance in the business will be measurable in the internal big data network of the company and should help to get the right people on their onwards and upwards careers.
With many decisions in businesses still relying purely on 'HiPPO' (the highest paid person's opinion), it seems unwise to not take into account the vast amount of precisely accurate information that can be gathered from big data internally. There is much emphasis on the information big data can gather about the external factors of a business such as customers, but the working habits of the employees within a business are equally important at understanding the performance of a company. Looking at an organisation through all angles of the big data prism is sure to lead to some enlightening results.