You can have the most accurate data in the world and best in breed technology but without the skills to apply these advantages you won’t get any benefit.

Of course, this should be a blindingly obvious statement but it is surprising if you talk to organisations, as we have done, about what drives innovation you find that employee capability can come low on the list.

So, we must not see HR analytics as the preserve of geeks, be they technological experts or data gurus: the key point of this article is that you need a range of skills to be successful.

The role of the HR business partners

The figure represents the types of skills required in the HR analytics process, but we should start with the least technical role in the range as it is the most important: the consultancy engagement with the customer to define the problem under investigation.

One of the ways proper analytics distinguishes itself from previous efforts in data management and reporting is that, rather than push information to the customer in the hope that something useful lands, a good analytics method begins with a business problem to solve.

This might be a poor safety record, falling productivity, unacceptable quality performance; whatever it is, the issue is defined in business not HR terms.

The value in HR analytics is that the people contribution to safety, productivity or quality failings is specifically examined ideally alongside other possible causes such as faulty equipment or cumbersome logistics.

Leading HR analytics practitioners liaise closely with other functions to offer integrated solutions to business problems.

The ability to work well with other colleagues is therefore a critical skill in identifying and prioritising these challenges.

This means the HR team must have communication, influencing and relationship-building capability.

Often it is the HR business partners who are best placed to perform the task as they should know both the leaders of their business unit and the nature of its activities.

The HRBPs must also have a reasonable grasp of the analytics process to deal with simple issues themselves (often what are essentially reporting points – is absence rising? Is turnover affecting critical groups?) or to weed out questions which are unanswerable (too complex or vague to be susceptible to analysis), as well as get the questions into shape such that they can be turned into hypotheses to be answered:

• Safety is suffering because it is given insufficient priority by managers?
• Productivity is failing because of poor basic skills?
• Quality is substandard due to poor communication between links in the process chain?

HRBPs also need the skills to perform a role at the back end of the HR analytics process in interpreting and explaining the results of an investigation.

This is an especially vital task if their analytics colleagues are not as familiar with the business context as they might be.

Whether the HR analysts have ‘domain’ knowledge of the function is a secondary matter if they have good HRBPs in place.

But if even where the analytical outcome is clear there is still a necessary task to discuss what to do next. HR analytics is action oriented; it is not a research exercise for the sake of it but a means to improve organisational performance.

Finally, HRBPs can help define the routine reports needed by their business unit to track critical human capital trends.

This should be specific to what the business must know – and it is not always absence and turnover; it might be rate of employee suggestions or proportion trained to a requite standard.

The better this task is done the fewer ad hoc reports have to be produced by the analytics team so that they can concentrate on trickier, more skilled activities.

The criticality of the HRBP contribution is such that some HR analytics leaders see the business partners as an integral part of the (extended) team and they will invest a lot of time and trouble in ‘training’ them in the work they do and in question specification, building hypotheses, knowing what can be answered and what not, communication of research outcomes, etc.

What the HR analytics team needs to know and do

Whether the HR analysts have ‘domain’ knowledge of the function is a secondary matter if they have good HRBPs in place.

Many of the leading lights in HR analytics have come into their jobs from outside the HR function. They have learned sufficient about HR to be effective and they can rely on their HRBPs if necessary for professional detail.

What is vital is knowledge of how the organisation operates – what are its strategic priorities, its business model including where value is generated (be it profit or social worth) and where its costs lie.

Knowledge of finance, supply chain, sales and marketing are also helpful to allow the analysts to have a rounded picture of how the company makes money and/or delivers services.

Analysis takes up much of the rest of the team’s time and necessitates having obvious technical skills like statistical modelling and data mining.

Another skill set which closely links to the HR business partner contribution is the communication of the results of the analytical exercise.

This once might have been the ability to produce a cogent report, whereas now it might be to generate a slide deck. What is now demanded is a concise (‘crisp’ is the word you sometimes hear) summary of what was found and what action is proposed.

However, technology increasingly allows fancier visualisation of results – infographics and the like. This technology is becoming accessible to the average user but it is helpful to have in the team someone with a good eye for a picture.

This should not go too far as you do not want the medium to obscure the message. So, the aim of the presentation should always be met, yet done in an attractive and well communicated manner.

Technical contracting skills may be required if third party suppliers are used.

These may be providing technology or data management and the IT department may have the principal interface with the analytical team giving the user requirement.

There may also be suppliers of analytical services ranging from conducting the employee survey to offering deep statistical expertise.

We have come across high-powered HR analytics units which contract out some of the most complex work to universities because it is uneconomic to retain such top end capability in house – it just would not be used enough.

Core HRA skills

This brings us to the core areas of work for HR analytics in data management, programming, statistical analysis.

Andrew Marritt claims data analysts spend nearly three-quarters of their time on ‘data munging’, ie ‘transforming, cleaning and manipulating data to get it from the state it is being stored, or was captured, and the state it needs to be in to do the analysis.’

As he points out, this requires programming skills in a language appropriate for your organisation.

Analysis takes up much of the rest of the team’s time and necessitates having obvious technical skills like statistical modelling and data mining, but insight also comes from curiosity and discovery.

This has led some organisations to talk about developing a sense of ‘playfulness’ and to trying to carve out time in the work schedule for exploratory activities; for example, what is the organisational value and ethics of internal emails?

Some organisations give particular attention to data governance ensuring definitions are agreed, data quality is maintained, reporting schedules are adhered to and HR KPIs are well chosen and populated.

This role demands high-quality administrative skills – attention to detail, thoroughness and determination to ensure compliance through persuasion, if not edict.

Organising the team

As the above account illustrates there are a variety of roles to perform using the range of skills we suggested.

This then begs the question of how best to staff and organise the HR analytics function. Most organisations tend to centralise the core activities, via a centre of expertise, to make what are usually scarce skills available to the whole organisation.

Of course, there are also devolved responsibilities vested in business partners who usually act as the conduit for the ultimate customers.

Within the HR analytics team itself, one approach (truer in smaller organisations) is to employ a group of all-rounders who are good at a variety of tasks with a broad understanding of analytical methods (eg, economists) who can undertake all the basic statistical tasks, but are happy to bring in more specialist expertise from outside as needed.

The contrasting resourcing model is to segment the team and use your own specialists in the key work areas deploying their skills across projects based on say owning a particular statistical approach (eg probabilistic causation) or data collection methodology (eg employee survey).

Within the HR analytics team itself, one approach (truer in smaller organisations) is to employ a group of all-rounders who are good at a variety of tasks with a broad understanding of analytical methods.

Conclusion

Given that the skills described here are so important to the success of the HR analytics operation there is an argument that the recruitment of the right team is the place to start when investing in analytics capability rather than spending money on the technology or on getting the data in order.

Indeed, this argument is consistent with the view that you can wait for ever to get the perfect set-up (especially with respect to data completeness) when in fact it is better to find an important business problem and use your skills to get an 80% solution.

This then gives you the justification to spend further money on buying extra kit and improving the data, but the skills come first.