The key to using data intelligently in HR is to start with the business numbers. This article is about how to structure a pay round driven by business results. It focusses on data rather than the normal subjective judgements and gaming that goes on in the vast majority of companies at pay increase time. The objective is to reward what matters to the business.
This is a long post; but it is not going to give the full detail of the approach. This will differ in each organisation. It is a new approach in thinking about the pay round process and the article gives the broad concepts and approaches to the subject rather than a detailed “Dummies guide”. (I can provide one of these for a reasonable fee).
Effective and efficient
This approach is based on good use of data and behavioural psychology. It generates rewards for behaviour that is important to the business. By doing this it sends out a clear message both on culture and on what behaviours are rewarded in the workplace. This creates the “virtuous circle” of reinforcing profitable behaviour leading ultimately to high performing teams.
You should use this approach. It gives hard statistical evidence as to why pay increases were, or were not given. It gives a better return on your pay increase investment. You are rewarding behaviour that benefits the organisation; not a managerial whim or some perception of employee merit based on last week’s conversation.
In the beginning
Like all good science, start with a test group. Select a discrete group of employees where business and HR data is available.
Key business success metrics
The key business success metrics for this group need to be clearly defined. As an example, if looking at sales staff then consider sales revenue, the conversion rate of sales calls to sales, repeat sales and so on. It is advisable to weight these business success metrics from the most important to the least important; always focussing on the bottom line impact of these factors.
Rank the employees
The next step is to rank the employees against the business metrics. This must be undertaken strictly against the business metrics. It is difficult, but is an essential part of the process. It may well be that your “best” employees are not those who score highest on the metrics. Stick to your original business metrics. Do not change them because the employees in the list do not fit your perception of “good”.
What makes these employees “good”?
This is the most difficult part of the process, but the most important. This is where the power of big data starts to prove itself. Now take the HR data on each of the top employees to see what common factors make these employees perform better against the business metrics than others. This could include:
- Time in role
- Education level
- Personality profile
- Training courses
- Previous roles
- Previous employer(s)
- Outside interests
- Social network size
- Email activity – internal and external
- Sales calls length and frequency
- Time and attendance data
- Daily newspaper and magazine reading
- Social profiling (you can use postcodes for this)
As long as you have the data and you should have the data, you can include it as a factor.
You will now need some strong statistical knowledge to undertake a regression analysis to identify the common factors for your high-ranking employees. I am aware there are a number of statistical techniques that can be used at this stage. You pay your money and you take your choice.
The outputs from this exercise will depend both on the richness of the data you hold on employees, the type and location of your organisation and your company culture.
It is important to note that this technique is not limited to revenue generating activities. We can build success factors for HR, cost of hire, attrition, benefit spend, payroll costs and so on. Or much the same in Compliance, for example. External audits passed, compliance costs, compliance checks carried out – you can fill in the blanks.
Results part one
What you will have, if the process has been carried out correctly, is a list of individual factors that predict behaviour that support business success. Some of the factors will appear not to be relevant; and I am aware that correlation does not imply causation. Some of the factors will be surprising, do not rule them out or ignore them. GO WHERE THE DATA TAKES YOU. Human beings are programmed to look for patterns where none exist and make choices based on often faulty heuristics. The data may not always take you in the right direction – but normally it will.
This is the easier part. You rank the employees by the factors. This process is already part carried out by the earlier steps. The exact nature of the ranking will depend on the analysis. One approach may be to rank the employees by the factors with the highest correlations to business metrics success.
The pay increase allocation process
In an ideal world you would allocate 80% of your budgeted increase to the top 20% of employees. That is because it is statistically likely that 80% of your revenue comes from this top 20% of employees.
This process largely removes the subjective elements and gaming that goes on around pay allocation in most organisations. Decisions can be justified and supported by the data. A clear signal is sent out to employees as to what is being rewarded.
Extra benefits to recruitment, training and development and succession planning
By having identified the factors that are correlated to business success (provided you have chosen the business metrics correctly) you have a powerful dataset to aid recruitment, training and succession planning.
You have a list of factors that predict business success and effective employees. Using these factors a template can be developed to quickly and factually identify those applicants who are most likely to do well in your organisation. It may not be the only measure; but it will provide an excellent screening tool.
Training and development
The factors that lead to success have been identified; thus you can train and develop employees based on those success factors. A provable bigger bang for the training buck.
From the analytical process you will have identified both the success factors and those supervisors who have the most successful teams. A variant on this exercise can be used to identify what factors make up the most successful supervisors and managers and build your succession plans accordingly.
Power of big data
The above discussion shows how HR data can be used to drive business success. One of the tenants of big data is to automate the analysis of the data. With a little work it is easy to automate the data scrapping processes to allow the identified factors to be ranked against employees and allocate the pay increases once the basic rules have been formulated. Having the data available and categorised allows for very powerful management information reports and data visualizations.
Warnings and alternatives
The above process is, for the majority of organisations, new and perhaps frightening. It will not work the first or second attempts. However, the very process of data scraping and analysis will yield a honey store of good things. The process can be changed and refined to fit the organisation.
This is different HR.
It is data driven and business focused. Some will argue it takes HR away from its traditional routes; why not? HR has not yet earned a full place at the board table with its current approach. Finance, IT and other support functions have a greater claim, because they have the data and facts to support cost activity.
This concept is fairly new for most organisations and will take:
- A change in mind-set
- A robust data store of employee and business data
- A strong understanding of the underlying statistical processes to carry out the appropriate analysis
- HR working with Finance, IT and data professionals, statisticians and the business to get the clear benefits from the approach.
When this approach is fully working it provides a rich and effective way of spending the salary budget as well as providing a firm “big data” base for HR strong analytics.
Working this way gives credibility to HR and builds up a subjective data bank of HR information with which to support business decision-making. Implemented appropriately it is a win win for all parties in the annual pay round process as well as for the wider HR community.