05
These days, most HR teams are already data rich, but that is not the same as being insight rich. To be insight rich, you need to be able to turn the data you collect into valuable insights that answer your strategic questions and help you to deliver your strategic goals. In this chapter I look at some of the more advanced data analytics methods available today, and explore some of the most valuable and useful people-related analytics. I also discuss why it is important to combine analytics to get a complete picture. Finally, I look at how to turn data and analytics into insights, and communicate those insights to the people that need them. As you read this chapter, keep in mind that it is very easy to get caught up in the exciting opportunities that analytics bring. It is therefore important to keep the focus on insights to avoid getting side-tracked by new and exciting analytics approaches that do not actually have an impact on performance. Organizations are doing very cool things with analytics, but what works for one business may not work for yours. Your challenge is to identify the best, most accessible, most achievable analytics approaches for you. In recent years, many companies have sprung up offering all sorts of analytics services, tools and platforms. So, for most of the options outlined in this chapter, there are numerous commercially available tools on the market that simplify these analytics processes.
Looking at the latest analytics techniques
As illustrated in Figure 5.1, some of the key techniques for analysing data include text analytics, sentiment analysis, image analytics, video analytics, voice analytics and predictive analytics. Let us look at each area in turn.
FIGURE 5.1 Analytics techniques
Text analytics
Text analytics is the process of extracting value from large quantities of unstructured text data. Most HR teams have either ownership of or potential access to large volumes of text-based data, including e-mails, survey responses, job applications, performance review files, social-media posts etc. Until recently, however, these data were not always that useful from an analytics perspective. Access to huge text data sets and improved technical capability means text now can be analysed to extract high-quality insights over and above what the text actually says. In this way, text analytics helps us to get more out of text, so that we can understand more than just the words on the page or screen and identify larger patterns. This makes text analytics especially helpful for understanding more about your employees. For example, text can be mined for patterns such as an increase or decrease in positive feedback from customers, and this may help to identify customer-service representatives who are performing well and those who may need extra support to improve in their role. I know of one organization that uses text analytics tools to scan and analyse the content of e-mails sent by its staff as well as their social-media posts. This practice allows that organization to accurately understand the levels of staff engagement, meaning it no longer needs to carry out traditional (and expensive) staff surveys and it no longer has to wait to assess staff engagement on an annual basis.
Sentiment analysis
Sentiment analysis is closely related to text analytics, since it helps to extract subjective opinion or sentiment from text (it can also be used on video or audio data to assess the sentiment behind spoken words). The basic aim is to understand the attitude of an individual or group regarding a particular topic (such as proposed changes to the company’s incentive scheme) or overall context (eg wider company culture) and whether that attitude is positive, negative or neutral. In this way, sentiment analysis helps us to get at the real truth behind communication. And, in this age of constant digital connectivity and our increasing desire to share our thoughts and feelings about all sorts of things – including companies – on social media, sentiment analysis has become mainstream.
You would use sentiment analysis when you want to understand stakeholder opinion, with stakeholders primarily being your employees, but also your leadership team and the company’s customers. Advanced sentiment analysis also can go further by making a classification as to the emotional state involved. For example, in Chapter 4 we looked at the Avatar system used in US immigration. Not only does the system analyse what a passenger is saying, it can also analyse the tone of their voice, facial expressions, body language etc to determine whether the person is calm or under stress (indicating that they might not be telling the truth).1 The vast majority of our communication is picked up non-verbally through body language and tonality, and this now can be analysed on a large scale.
Image analytics
Image analytics – the process of extracting information, meaning and insights from images such as photographs, medical images or graphics – relies heavily on pattern recognition. In the past, the only analysis that was possible on images was via the human eye (such as a doctor looking at a patient’s scan). If computers were used, they could only categorize images using descriptor keywords (tags) that were manually added by a human being to each image. Advances in image analytics mean computers now can understand and recognize the content of an image (eg an individual’s face), as well as analysing the digital information associated with the image (date taken, location etc). Image analytics can be used in a number of ways, such as facial recognition for security purposes or recognizing your brand or product in photographs shared by your employees on social-media platforms.
Video analytics
Video analytics is the process of extracting information, meaning and insights from video footage. It includes everything that image analytics can do, but additionally it can measure and track behaviour. A good example of this is the use of CCTV cameras that can detect when an employee is not wearing the appropriate safety gear, such as a hardhat.2 You could use video analytics if you wanted to increase security or understand more about how your employees behave when they are on site. You could also use video analytics to reduce costs and risk and assist in decision making. For example, a number of providers now offer software that allows you to automatically monitor a location 24/7. That video footage is then analysed using a video and behavioural analytics solution which alerts you in real time to any abnormal or suspicious activity. Once installed and provided with the initial video feed, the software observes its environment and learns to distinguish normal behaviour from abnormal behaviour. The system is also self-correcting, which means that it continuously refines its own assumptions about behaviour and no human effort is required to define its parameters.
Voice or speech analytics
This is the process of extracting information from audio recordings of conversations. Voice analytics can be used to analyse the topics or actual words and phrases being used (content), as well as the emotions behind that conversation (sentiment). You could, for example, use voice analytics to help to identify recurring themes regarding employee satisfaction, or, in the case of a customer call centre, employee performance. Voice analytics also can help you to identify when your employees are getting frustrated or angry. By analysing the pitch and intonations of conversations taking place in your call centre, you can gauge the emotional state and performance of customer-service representatives and identify those who are top performers, as well as those who may need additional training or coaching.
Predictive analytics
Predictive analytics uses data, statistical modelling and machine learning (see Chapter 2) to predict the likelihood of future outcomes based on historical data. By understanding as much as we possibly can about what has happened in the past, it is possible to identify patterns and build models for working out what will happen in the future. This is an incredibly useful tool for working out how likely something is to happen and the level of risk, particularly when it comes to identifying when key personnel may be likely to leave. Google, for example, used this technique to work out that new sales people who do not get a promotion within four years are much more likely to leave the company.3 As the volume of data we have and the computing power available are increasing all the time, our ability to predict future outcomes is improving too. Predictive analytics and tools like IBM Watson are therefore becoming increasingly popular with businesses of all sizes.
Looking at critical HR analytics
Having just looked at the general categories of data analytics being used today, it is worth us spending some time on the specific types of people analytics available to HR teams. Usually, these people analytics options make use of the techniques already outlined in this chapter, eg capability analytics (the first option we look at below) might make use of text analytics to analyse questionnaire or interview answers. It should be said that there are very many different types of people analytics, but some add more value than others. Based on my experience of working with HR teams, the following are the most important and useful people-related analytics being used by HR teams and managers to better understand their people (see Figure 5.2).
FIGURE 5.2 Critical HR analytics
Capability analytics
The success of any business depends on the level of expertise and skill of its workforce. Capability analytics is a talent management process that allows you to identify the capabilities or core competencies you want and need in your business. Once you know what those capabilities are you can then analyse your current staff members to see if you have any capabilities gaps. Capabilities, of course, do not just mean qualifications and skills – they also include capabilities that may not be formally recognized, such as the ability to develop and maintain relationships.
Why use capability analytics?
Knowing what skills you need and what you already have in your business can alert you to issues you may not have been aware of, allowing you to retrain or support individuals to close those gaps more effectively. Too often we hire new employees without really knowing exactly what skills we already have and what additional skills we need. As a result, we hire based on gut instinct, what is written on a CV or how well or otherwise someone comes across in an interview. There may be a laundry list of ‘ideal candidate characteristics’ but they are often generic personality issues such as ‘honesty’ and ‘integrity’ rather than ‘capability in X software’ etc. When these new recruits join the business, you may hope that they will fit in and work well with the existing team, but if you are unsure exactly what capability you expect them to bring to the table beyond being another pair of hands, then you end up being disappointed with the appointment. Capability analytics helps to avoid this scenario so that you know exactly what you need, what you have and what additional capabilities you may need to recruit (or what additional training you need to provide for those that are already in the business) to close that gap. I would say capability analytics is especially important if your business, industry or market is changing quickly.
It is always wise to conduct capability analytics at least once a year and certainly before every significant or important appointment. For existing employees, capability analytics can slot into the performance review process to inform ongoing training and improvement initiatives. It is also sensible to run capability analytics if your business is changing and moving into a new area or slightly different direction that will require additional or different capabilities. Knowing who can do what will also help you to move people around the business as appropriate, and provide additional training or support to those who could adapt to a new role or position. Just because someone is doing one job now does not mean they do not have the capability to do another. But you will not know what someone is fully capable of unless you conduct capability analytics.
How to use capability analytics
Capability analytics can be conducted via questionnaires, as well as by interviewing the individual being assessed and the people who work with them closely. Say, for example, you are an IT manufacturer. The speed of change in the IT industry, in terms of both the technological capabilities of the machines we use and buying behaviour, means that it is a highly volatile industry. Just a decade ago the industry was riding high, manufacturing large mainframe computers as well as smaller machines for the personal-use market. But the advent of cloud computing has massively changed the market. By conducting detailed capability analytics, you realize the capabilities that made you a dominant force in the last decade will render you obsolete in the next. To stem this potentially disastrous outcome, you create a competency framework for your business outlining specific and generic capabilities. For example, you might appreciate that everyone in the business needs to improve their ‘customer focus’ competency. In addition, you might appreciate that, within your specialist IT centres, you need to focus on big data competencies, including Hadoop skills and cloud computing skills that do not currently exist in the business. This competency framework helps the HR team to source appropriate training and/or recruit new employees to kick-start the skill shift.
While skills and capabilities are critical to success, remember that team fit, cultural fit and existing relationships are also very important. It is therefore often much easier to keep all that and retrain personnel to create the capability needed, rather than find the capability and hope that individual then fits into the culture or team.
Competency acquisition analytics
Competency acquisition analytics is the process of assessing how well or otherwise your business acquires talent. This is done by identifying key competencies that are vital to your organization’s success and then measuring how effective you are at attracting those competencies. Competencies may refer to specific skills or knowledge (like data analytics skills) or certain attributes or behaviours (such as leadership qualities or the ability to work well with others).
Why use competency acquisition analytics?
Talent recruitment and management are critically important for the growth of most businesses. The competition for talent is fierce and talented people can be very expensive to recruit and keep. Competency acquisition analytics helps you to assess how successful your talent strategy is and how well (or otherwise) your business acquires talent. Competency acquisition analytics should be something you assess at least every year to see how well your business is doing at: 1) identifying the competencies you need and want; and 2) finding those competencies cost-effectively.
It can be relatively easy to identify key players in any industry and, if your business has deep pockets, those individuals often can be attracted to your business. But individuals who only take a role because of the money will probably not stay that long or fully engage in the vision of the business. Finding ways to identify talent before it is fully fledged is key and competency acquisition analytics allows you to do that. It is therefore important to know how well your business is performing in the task of identifying talent early before stepping in to secure it at a reasonable price.
How to use competency acquisition analytics
A good starting point is to identify the key competencies your business requires now and in the future in order to stay competitive. This identification can be achieved using a number of tools and techniques such as text analytics, focus groups, interviews and surveys. Remember, as well as job-specific skills and knowledge, behaviours and attributes are just as important (if not more so). Depending on specific jobs and the organization’s goals, these attributes may include:
· communication skills;
· teamwork;
· coping with change;
· analytical thinking;
· conceptual thinking;
· managerial qualities like the ability to motivate and support others.
The next step is to assess the current levels of these competencies within your business and the gap between what you would like to have in terms of competencies and what you actually have at present. Then it is a good idea to create regular assessments so that you can track your progress in acquiring these key competencies over time, such as how effective you are at spotting and recruiting candidates with those competencies, how well you are able to close the competency gaps and which competencies are proving most difficult to acquire.
A great example of this in action comes from the world of baseball. You are probably familiar with the story from the book and film Moneyball. As in many sports, historically new baseball talent was ‘spotted’ by experts and talent scouts who would travel around the country watching baseball games in the hope they would be able to identify an up-and-coming star. The process was very subjective and, for the most part, it came down to experience and luck. Baseball advisor Bill James changed all that. He developed a scientific evidence-based approach to ‘spotting’ new baseball talent that broke a player’s behaviour and actions down into multiple measurable elements.4 Billy Beane, the general manager of the Oakland Athletics (aka Oakland A’s) baseball team heard about James’s theory and decided to work with him to acquire competency or talent. James’s hypothesis worked. Despite having the third-lowest payroll in the league, the Oakland A’s team was able to buy under-valued talent that took the club all the way to the playoffs in 2002 and 2003.5 Prior to this data-driven approach, the club simply was unable to successfully compete with deep-pocketed baseball clubs like the New York Yankees. Competency acquisition analytics changed all that, and changed the club’s fortune.
A final word of warning: competency acquisition analytics is only going to be successful if you are able to effectively identify and track competencies in your organization. Many companies do not concentrate on the vital (or difficult to get) competencies and instead produce generic competency frameworks that make the process of tracking and assessing competencies very complex and cumbersome. Key to effective competency acquisition analytics is focusing on a small set of absolutely key competencies.
Capacity analytics
Capacity affects revenue, which should make it a key focal point of intelligent HR. Capacity analytics seeks to establish how operationally efficient individual employees are in a business, eg are people spending too much time on administration and not enough on more profitable work, or are individuals stretched far too thinly? It also allows businesses to establish how much capacity they have to grow, allowing HR teams to identify patterns and trends in employee performance that then can be used to improve recruitment or training and development.
Why use capacity analytics?
If you do not know what your people are doing, you cannot manage those employees’ capacity appropriately. A consultant operating at full capacity, who is asked to pick up the slack with a new client, may end up stressed, unhappy and potentially look to leave the company. Meanwhile, another consultant who is spending too much time on administration tasks is not being as productive or profitable as you need them to be. Obviously, people are people, not machines, so an individual’s capacity will fluctuate throughout the year based on a variety of factors. These peaks and troughs of productivity are normal; however, capacity analytics can help to alert you to negative or worrying productivity trends. This analysis allows you to then step in with additional training or support to help the individual get back on track before they become too demoralized or negative.
How to use capacity analytics
As long as you have a system that tracks data on how people spend their time, you can use these data to establish capacity levels. The data can come from time-tracking systems (where people clock in and out) or from sensors (such as sensors in name badges, as mentioned in Chapter 4). Say you are a software engineering company and you have 20 software engineers working in your business. Capacity analytics allows you to track how much time they actually spend coding and how much time they do other work. This ratio then can be tracked over time to ensure the actual time spent (obviously relative to the billable output) on programming is not going down. It also allows the company to understand how much capacity it has to take on new projects. If everyone is at 100 per cent capacity then taking on any more work is not advisable unless capacity can be increased by recruiting new staff.
Fortune cited one real-life example of a manufacturing company that discovered some of its junior managers were spending upwards of 30 hours a week ‘managing up’, ie attending status meetings or giving reports to senior executives.6 Clearly, that left them only around 10 hours a week for getting on with revenue-generating work. Using this knowledge, the company has now implemented a strict policy of fewer meetings. The tricky part is establishing these systems without creating huge administrative burdens and without alienating employees with a ‘Big Brother’ approach. Capacity analytics can make people very nervous, so be careful how you represent capacity analytics to your people. The idea is not to find out who needs to be whipped into doing more work but rather to identify gaps in capacity that then can be closed to increase profit.
Employee churn analytics
Your employees are your most important and often most expensive asset. Using analytics to assess capability and hiring the people with the right capabilities is just part of the process. You also have got to keep them. Hiring employees, training them and then integrating them into the business costs time and money. When that investment is lost because too many employees are leaving, this can have a detrimental impact on the business. Plus, high staff turnover levels can be extremely disruptive to the remaining team members and lead to a decrease in morale and staff productivity.
Why use employee churn analytics?
Employee churn analytics is the process of assessing your historic staff turnover rates in an attempt to predict the future, so that you can intervene earlier and reduce churn. While some employee churn can be desirable to prevent stagnation, it is important to identify a healthy level of churn and develop a system to pinpoint the ‘regrettable’ churn. Some businesses have higher staff turnover than others. Call centres, for example, have notoriously high staff turnover, especially compared to a more traditional industry like manufacturing. In fact, some estimates put call-centre turnover as high as 30–45 per cent.7 Depending on the volatility of your business, you should track employee churn analytics every six months or annually. Essentially, you need to know the trend, ie is there more, less or stable employee churn in your business? If the trend is heading up then this can provide a red flag for further investigation to stabilize turnover or even reduce it further.
How to use employee churn analytics
Historical employee churn can be identified through traditional key performance indicators (KPIs) such as the employee satisfaction index (ESI), employee engagement level and staff advocacy score. In addition, surveys, exit interviews, performance reviews and social-media data can help to gather further information that then can be mined (perhaps using text analytics) for greater insights. Historical employee churn rates can be useful as a benchmark but the real value lies in comparing your business against industry averages, seeking to identify patterns in employee churn in your business and, most useful of all, applying different analytics techniques to understand why people are leaving. Once you know why, you can predict employee churn in the future and, most importantly, take any necessary internal action that could solve the problem and keep employees engaged.
One example of this in action can be seen on the Watson Analytics blog.8 The use case describes how a simple analysis by IBM’s Watson platform of data on past and current employees can be used to identify factors that are related to employee attrition and predict how job role and performance evaluation relate to employee churn, based on the data from employees who have left the company. Overtime, job level, number of years with the current manager and employee age were all significant drivers of employee churn. It also broke the data down by job role, showing that people in an HR role or management position were more likely to stay with the company than people who worked in sales or quality control. And, in this case, employees who worked more than 15 hours of overtime a week were most likely to leave the company. This may seem like a reasonable assumption for any HR professional to make: people who feel overworked are more likely to leave. But one of the critical things about data-driven HR is basing decisions on data rather than gut reactions. In this case, the data proved beyond doubt that the number of overtime hours was a significant driver of employee churn, allowing the HR team to make intelligent, evidence-based decisions to tackle this and prevent future employee churn.
You may have guessed by now that I am not the biggest fan of annual employee surveys. If you are using annual surveys to measure your employees, I believe you are almost certainly missing valuable data. A better way to use employee surveys, instead of getting all your employees to complete a survey once a year, is to invite one-tenth of the workforce to complete the survey every month for 10 months. In this way, everyone still only completes the survey once a year but you have 10 data points not one, giving you a picture of an employee churn trend, which can allow you to make in-time corrections to minimize employee churn.
Corporate culture analytics
Culture is notoriously difficult to pinpoint and even harder to change. Culture is not something that can be hung on the wall like a values statement, it shows up as the collective actions of the people in the business. Corporate culture analytics helps you to assess and understand more about your corporate culture, or the different cultures that exist across your organization, which then allows you to:
· track changes in culture that you would like to make;
· understand how the culture is changing;
· create early warning systems to detect toxic cultures in their development;
· ensure you are recruiting people who do not clash with the corporate culture.
Why use corporate culture analytics?
Essentially, corporate culture analytics can allow you to uncover the genuine culture of the business in order to amplify the good bits and help change the unhelpful parts. Part of the reason why culture is so difficult to change is that the people in the business do not fully appreciate what it is to start with. This type of analytics can lift the lid on culture, which, in turn, can influence strategy. Corporate culture is usually fairly stable. Once you have assessed it initially, you can then put systems in place to track key elements of that culture in an ongoing fashion, using data collections to create early warning systems of a mismatch between the culture you would like to have and what the data are showing you.
How to use corporate culture analytics
Perhaps the most common tools for culture analytics are surveys, focus-group research and employee interviews. The challenge with these approaches is that people can tell you what they think you want to hear; plus they can be expensive. There are now many more analytics tools that can be used to give a better and more accurate insight into corporate culture. You can, for example, collect data from internal intranet sites, social media and internal written communication, and analyse them using text analytics and sentiment analysis. Customer-service conversations also can provide a rich vein of data for assessing corporate culture. If you record customer-service conversations, or interactions between employees, these data can give some very useful insights into corporate culture. You can apply voice analytics to these data or text analytics and sentiment analysis tools. Say, for example, you believe that your corporate culture is efficient but fun. You may think that your business operates like a family, with a strong focus on excellent customer service, and those are values that have been driven home to employees. Your orientation for new recruits draws their attention to these values and the corporate culture that you believe exists. But what happens after six months, are those employees embodying these values or is something else calling the shots? You could ask your new employee or you could conduct annual surveys, but the quickest way to get to the truth is to assess what your employees are saying and doing as part of their day-to-day life, perhaps through activity data or conversation data: this is your culture. If you believe you are driven by high-quality customer service and yet no one answers the phone after 4.45 pm, chances are that culture is not quite as real as you would like it to be.
Recruitment channel analytics
As an HR professional, you will know that employees represent both the greatest cost and the greatest opportunity in most businesses. Plus, getting recruitment wrong can be really problematic for a business. A poor employee can disrupt a team and cause upset as others need to cover for their poor performance, which of course can increase employee churn. And, sadly, it is usually not the person you want to get rid of that leaves. Recruitment channel analytics is the process of working out where your best employees come from and what recruitment channels are most effective. It can help to ensure you recruit the right people from the start.
Why use recruitment channel analytics?
There are many ways to recruit employees, such as print advertising, advertisements in specialist journals or magazines, online recruitment websites and recruitment consultants. Their costs vary widely and the time required to recruit through these various channels also varies significantly. Knowing which ones are working and which channels are most cost-effective is therefore important for ongoing recruitment. The purpose of recruitment channel analytics is to allow you to use only those channels that yield high-value candidates. Traditionally, recruitment success has been measured by simply counting the number of applications delivered or the number of positions filled. Modern recruitment, however, is full of data and allows us to track reach, engagement, costs per appropriate candidate etc. But the ultimate measure is how many people are successfully recruited and actually stay with your organization.
How to use recruitment channel analytics
Recruitment analytics will involve some historical assessment of employee value using standard KPIs such as return per employee (RPE). These KPIs will help you to identify who your most productive and valuable employees are. Surveys, entry interviews, aggregator sites like Glassdoor and social-media sites also can be used to gather more data. You can then use these data to identify patterns or connections between high-value recruits and recruitment channels. The best results tend to come from mixing qualitative and quantitative insights, eg mixing referral rates, quality of candidate, quality of hire, and candidate and manager satisfaction with measures like cost to hire and time to hire. Say, for example, you use a recruitment consultant to help you hire for more senior positions. While they do help you to find suitable candidates and they pre-screen those candidates, saving you valuable time, the costs can really add up. Using recruitment channel analytics, you may identify that online recruitment overall is more effective. Further analysis also may show that your most high-value candidates always held positions prior to starting with you for three years or longer. These insights then can be used to fine-tune the assessment process and discount candidates who do not meet that criterion.
Leadership analytics
Poor leadership, whether of the whole organization or a specific team, costs money and holds the company back from fulfilling its potential. If a leader is not great at empowering and engaging their employees then this will impact on results, productivity and profit.
Why use leadership analytics?
Leadership analytics seeks to uncover how good leadership is in your business. So much of leadership is subjective. We are told that great leaders are born not made but is that really true? Leadership analytics unpacks the various dimensions of leadership performance via data to uncover the good, the bad and the ugly. Leadership is best assessed on an ongoing basis but, if that is not possible, then you can assess it at regular intervals, eg every six months or so. If someone is new to a leadership role, then it is probably wise to perform leadership analytics more frequently to track their early progress, which will allow you to pick up any failings early so as to get the individual back on track.
How to use leadership analytics
Data about leadership performance can be gained through the use of surveys, focus groups, employee interviews and possibly employee conversation data. Where you are directly asking employees for input, it is advisable to make the data collection anonymous so that employees can really open up. Few employees would feel confident or safe talking about their leader or manager if they knew that person may have access to their opinion. It is also possible to conduct behaviour profiling of leaders. Really good leaders tend to demonstrate certain personality traits or characteristics. These can be generic attributes or you can analyse your existing leadership to identify what the really good ones have that the less successful ones do not. These insights can be used to direct training and support programmes, as well as the recruitment process. Text analytics is a very powerful way of extracting key leadership characteristics, both of good and not so good leaders. You also can use financial metrics (ie turnover and profit demonstrate how the company is performing financially under the leadership), as well as data such as employee satisfaction or churn in assessing leadership.
In Chapter 2 we briefly looked at how Google used data and analytics to assess and increase the value of managers in its organization. Here we revisit that example by looking in detail at how it worked in practice. In an effort to raise leadership performance, Google set out to answer two questions: 1) What is it that makes a great manager? and 2) What are the behaviours that make managers struggle?9 Based on some extensive leadership performance analytics, including interviews with its managers, 360-degree feedback surveys of its employees and regression analysis of things such as job performance and employee satisfaction, Google was able to identify eight behaviours that make a great manager:
1. is a good coach;
2. empowers the team and does not micromanage;
3. expresses interest/concern for team members’ success and personal wellbeing;
4. is productive and results-oriented;
5. is a good communicator: listens and shares information;
6. helps with career development;
7. has a clear vision/strategy for the team;
8. has important technical skills that help them to advise the team.
In addition, the research alerted Google to the top three reasons why managers were struggling in their role:
1. having a tough transition (eg suddenly promoted or hired from outside with little training);
2. lacking a consistent philosophy/approach to performance management and career development;
3. spending too little time on managing and communicating.
Acting on these valuable insights, Google now gears the 360-degree feedback surveys for managers around these aspects and conducts them twice a year, thereby instigating an early warning system to detect both great and struggling managers. In addition, Google has revised its management training and recruitment in light of its findings.
While there are some generic leadership assessment models that you could use in your business, it is always better to create your own model based on which leadership characteristics you value in your particular corporate culture. Draw on data and insights from other analytics tools, such as employee performance analytics (which I will look at next) and corporate culture analytics, to help you establish what makes a great leader in your business. This was why Google was so successful in its quest to identify leadership excellence: it took the time to figure out what leadership excellence looked like in its unique culture first.
Employee performance analytics
Your business needs capable, high-performing employees in order to survive and thrive. Unless you measure performance, it easily can get lost in the day-to-day operations of the business. A poor employee can effectively be carried by a productive one, which will eventually irritate the productive employee. Your job is to know who is doing what and who needs support so that you can provide that support and lift performance across the board, which is where employee performance analytics can help you.
Why use employee performance analytics?
Employee performance analytics seeks to assess individual employee performance. The resulting insights can identify who is performing well and who may need some additional training or support in order to raise their game. An understanding of employee performance also can feed into the recruitment process so more of the right types of employees are recruited and costly mistakes are avoided. Most companies assess employee performance annually, but, in this world of big data, just once a year is not enough. In order to be effective, performance should be assessed on a regular and less formal basis, and modern data collection methods allow us to collect data from many different sources to aid in the assessment.
How to use employee performance analytics
Today, we have many innovative ways of collecting and analysing performance, ranging from crowdsourced performance assessments to sensors in employee badges, like the Sociometric Solutions badges we saw in Chapter 4. These data can be analysed in a number of ways, including text analytics, sentiment analysis and voice analytics. Such analyses may help to identify any patterns that you may not have been conscious of, which can be used to improve ongoing performance. For example, one of Sociometric Solutions’ clients, a major bank, noticed that its top-performing employees at call centres were those who took breaks together. Based on this knowledge, the bank instituted group break policies and performance improved by 23 per cent.10
Employee performance analytics can be particularly useful in businesses that traditionally have a high staff turnover, such as call centres. It is important to understand the different call lengths for each operative, how many calls they get through per hour, how many of their calls escalate into issues and how many end in resolution and a happy customer. These and countless other data points also allow you to detect patterns, and identify your star performers so that what they do can be replicated by others. These insights also can be used to fine-tune customer processes, recruitment, and training and development initiatives, so that you get more great employees and fewer poor ones. Not only does this improve results, but it also can significantly reduce staff turnover and recruitment costs.
If done well, performance analytics provides a positive experience that contributes to the overall employment and career development experience and helps to strengthen the relationship between line managers and their staff; however, be aware that, whenever you monitor the performance of employees, the monitoring itself will affect performance. Usually when people know that specific elements of their job are monitored, they make sure they perform particularly well at them. This can skew their attention away from simply doing a good job to simply focusing on the things that are being monitored and analysed, which is why modern data capture techniques such as video and sensor data are so helpful. Using these techniques, it is possible to analyse performance more holistically, being less focused on specific parts of a job that might cause the employee to skew their behaviour.
Combining analytics to get the best results
Often, to get the most out of data-driven HR, you will not be able to rely on one analytics tool alone. Just as we saw in Chapter 4, while it is usually a good idea to combine different data sets to get a fuller picture, the value of HR analytics lies in the insights that can be gained from combining different types of analytics. For example, corporate culture analytics may tell you that your culture is moving away from the values you have prioritized, but you may need text and sentiment analysis to tell you why that is. The idea behind combining analytics is to base your decision making and HR operations not just on what one set of analysis is telling you. Combining information from more than one source and using different analytics approaches allow you to verify insights from more than one angle.
All the approaches I have outlined in this chapter show only some of the analytics possibilities available to HR teams today. Just a few years ago, much of this was not possible; we could not do sentiment analysis on text, for example. Analytics in particular has made such huge leaps that no one knows for sure what is going to be possible in 10 or even 5 years’ time. Therefore, an important part of data-driven or intelligent HR is staying open to new opportunities that data and analytics may provide further down the road.
Turning data and analytics into insights
Data are only really valuable if you can turn them into insights and actionable knowledge. By analysing your people-related data using some of the analytics methods outlined in this chapter, you should arrive at various insights. Presenting these insights in a helpful way to the people that need them is a key step in turning insights into actionable knowledge. After all, businesses gain competitive advantage when the right information is delivered to the right people at the right time.
Who needs access to the insights you uncover?
For each objective outlined in your strategic plan and each data set related to those objectives, you will need to ask yourself: who are the decision makers who require access to the insights from those data? In some cases, it will be just those within the HR team, while in others it will be the leadership team or managers across the business. It is important to involve all the key players that relate to the business’s goals and strategic questions. What is the best way to disseminate insights to the people that need them? It depends on what you are measuring, who needs to know about it and how you usually communicate across the company. You could, for instance, have an indicator (such as employee performance) included in a monthly report that is distributed to the people who need it. Or, you may need more sophisticated, real-time information in the form of a dashboard that allows decision makers to access information whenever and wherever they want; indeed, this kind of democratized approach to data is becoming increasingly more common. According to a survey of 2,000 employees in the United Kingdom and United States,11 more than half of respondents said that knowing company performance data contributed significantly to their own positive performance. In other words, employees want to be included in discussions about overall business performance and that means, in an ideal world, key data would be communicated across every level of the business.
How best to disseminate insights
But even when people have widespread access to data and the insights generated from them, they do not always interpret those data in the same way, and this means they may need help extracting key messages. A blended approach therefore may work best for your company, combining widespread access to data across the company, where people are encouraged to use data as the basis of future business decisions, as well as a strong overarching narrative that sets out key insights and trends, just to be sure that the most critical messages are understood by everyone. However you decide to disseminate the information, keep in mind that the format in which it is presented plays a big role in how useful that information will be. People are less likely to act if they have to work hard to understand what the information is telling them. It is therefore vital that insights are presented in a clear, concise and interesting way.
Communicating and visualizing insights from data
Data-driven HR is about turning people-relevant data into insights and actions that add value to the business. In order to do this successfully, you need to ensure it is easy for the various decision makers, whoever they may be, to extract insights from the data. The easier it is to understand the data and pull out key insights, the easier it is for people both within the HR team and beyond to make decisions and act on those data. This is why data communication and specifically data visualization have become such big topics in recent years.
There are many different options for communicating data, from simple graphics and written reports, through commercial data visualization platforms that make the data attractive and easy to understand, to management dashboards that provide your people with the information they need whenever they need it. Different audiences have different needs, in terms of both the types of data they need and how they will use them. Therefore, when thinking about disseminating and communicating data, it is important to define who will have access to those data (or the insights from those data) and what their needs are. For example, what format works best for your data consumers? How will they access the information (web interface, reports, dashboards etc) and how often? Knowing the answers to these questions will help you to decide on the right visualization/communication tools for your needs. Data visualization tools are particularly helpful because they can very clearly highlight the most important data or results and help to identify trends in the data. There are now many excellent and inexpensive data visualization tools, like Tableau, Qlik or Google’s Analytics 360 suite. In addition, many commercial analytics platforms come with their own built-in visualization tools.
Visuals are great for conveying information because they are quick and direct, and they are far more interesting to look at than a page of text. But, unless we know how to decode its message, a picture also can be difficult to read. Words, on the other hand, usually have a very direct meaning and are simple to understand. With a short narrative you can ensure everyone understands the data in the same way. This is why using visuals and narratives together is much more powerful than using either on their own. For instance, a graph may be a good way of showing employee churn trends over time, but a simple narrative alongside it can pull out the key messages and put that information into context, explaining what might be behind those trends and why there was a spike in churn in late 2016, for example.
Key takeaways
I am aware that this has been a pretty chunky chapter, with lots of new information to process. The following is a quick rundown of the key points on data analytics:
· The key analytics techniques can be broadly categorized as:
1. – text analytics;
2. – sentiment analysis;
3. – image analytics;
4. – video analytics;
5. – voice or speech analytics;
6. – predictive analytics.
· In my experience, some of the most useful and valuable HR-specific analytics are:
1. – capability analytics;
2. – competency acquisition analytics;
3. – capacity analytics;
4. – employee churn analytics;
5. – corporate culture analytics;
6. – recruitment channel analytics;
7. – leadership analytics;
8. – employee performance analytics.
· To get the most out of data-driven HR, you probably will not be able to rely on one analytics tool alone. Often, the value of HR analytics lies in the insights that can be gained from combining different types of analytics.
· Data are only really valuable if you can turn them into insights and actionable knowledge. Businesses gain competitive advantage when the right information is delivered to the right people at the right time.
· The easier it is to understand the data and pull out key insights, the easier it is for people both within the HR team and beyond to make decisions and act on those data.
· Data visualization tools are particularly helpful because they can very clearly highlight the most important data or results and help to identify trends in the data.
Clearly, there are many new and exciting ways to use data and analytics across HR functions. But with these new and exciting methods comes increased risks. Part of using data intelligently is ensuring that they are properly protected and that employees’ privacy is not violated. Transparency, data governance and data protection are therefore vital things every HR team needs to consider. In the next chapter I explore the various pitfalls and risks concerning data usage, and set out good practice for avoiding those pitfalls.
Endnotes
1 US Department of Homeland Security (2014) [accessed 23 October 2017] Rapid Screening Tool: the Avatar [Online] https://www.dhs.gov/sites/default/files/publications/Rapid%20Screening%20Tool-NCBSI-AVATAR-Jan2014.pdf
2 Shrestha, K (2015) [accessed 01 February 2018] Hard hat detection for construction safety visualization, Journal of Construction Engineering, [Online] http://dx.doi.org/10.1155/2015/721380
3 Van Vulpen, E [accessed 23 October 2017] Predictive Analytics in Human Resources [Online] https://www.analyticsinhr.com/blog/predictive-analytics-human-resources
4 Adams, M (2015) [accessed 23 October 2017] The Man behind Moneyball: the Billy Beane Story [Online] https://www.domo.com/blog/the-man-behind-moneyball-the-billy-beane-story
5 Lewis, M (2004) [accessed 23 October 2017] Moneyball [Online] http://michaellewiswrites.com/index.html#moneyball
6 Clancy, H (2015) [accessed 23 October 2017] What Can Big Data Reveal about Corporate Culture? Get Ready for ‘People Analytics’ [Online] http://fortune.com/2015/03/20/analytics-corporate-culture
7 Reynolds, P (2015) [accessed 23 October 2017] Exploring Call Center Turnover Numbers [Online] http://www.qatc.org/winter-2015-connection/exploring-call-center-turnover-numbers
8 Alexander, F (2015) [accessed 23 October 2017] Watson Analytics Use Case for HR: Retaining Valuable Employees [Online] https://www.ibm.com/communities/analytics/watson-analytics-blog/watson-analytics-use-case-for-hr-retaining-valuable-employees
9 Blodget, H (2011) [accessed 23 October 2017] 8 Habits of Highly Effective Google Managers [Online] http://www.businessinsider.com/8-habits-of-highly-effective-google-managers-2011–3
10 Kuchler, H (2014) [accessed 23 October 2017] Data Pioneers Watching Us Work [Online] https://www.ft.com/content/d56004b0-9581-11e3-9fd6-00144feab7de?mhq5j=e6
11 Whittick, S (2015) [accessed 23 October 2017] Research Report: One in Four Employees Leave due to Mushroom Management [Online] https://www.geckoboard.com/blog/research-report-one-in-four-employees-leave-due-to-mushroom-management/#.V2GV1sdcJ6A