October 17 , 2017

Top 15 Design Principles for Talent Metrics and Analytics

Published at ATC Hub February 4, 2014

Whether you work in Australia, the USA or Europe, the hottest topic relating to how the Talent Management function operates is how to shift to a “metric based” decision-making model using advanced and predictive analytics (Note: the #2 topic is often making all talent applications fully accessible from the mobile phone).

The fact that metrics remains a hot internal topic may be a bit surprising to many, given all of the time and resources that talent leaders have invested in metrics efforts over the last decade. However, as an expert with over 30 years in the field, I’m not surprised at all because most HR metric efforts are put together by an “ad hoc committee” of HR professionals that quite frankly, are merely learning as they go. Other talent leaders take the easy way out and simply utilize the “standard metrics” that come automatically supplied as part of vendor software that they purchased. Unfortunately, these vendor supplied metrics are not designed for day-to-day decision-making or predicting future problems. The resulting set of metrics has in many cases merely allowed talent leaders to say “yes, we have metrics”. But no one can prove that their metrics have actually improved talent decision-making, and certainly not talent results. So if you want to break away from the pack and adopt the goal of making your metrics so effective that they indisputably directly improve the performance of managers and HR professionals, you need to be scientific in your approach.

After 30 years of designing talent metrics, I have found that any effort is doomed to failure if it doesn’t start with what I call “design principles” to guide metric selection. Whether you call them laws, rules, principles or critical success factors, I have found that following them is essential if you are to successfully shift to a data-driven decision-making model in Talent Management. Shifting to a data-driven model brings your function into alignment with almost every business function, like supply chain, finance and CRM. All of them long ago shifted to metric driven decision-making because it allows you to demonstrate your business impacts. My design principles are listed below with the most impactful ones appearing first.

The top 15 design principles for high impact metrics

 

  1. Make the business case for adopting a data based decision model
    Begin the process by putting together a complete business case that demonstrates the negative business impacts and dollar costs as a result of having ineffective talent metrics. In it, show the costs of relying on historical metrics and emotional decision-making in order to get support and funding for your new metrics initiative. If you don’t demonstrate the potential ROI of a comprehensive metric effort, you are already getting off on the wrong foot.
  2. Show the positive impact of metrics after they are implemented
    After making the business case, it’s important that you also show your managers that when the use of metrics for decision making significantly increases in a business unit or team, their revenue and business results will soon measurably increase (i.e. demonstrate a positive correlation between business results and talent metrics usage). Also be able to show that adopting talent metrics can start a “business turnaround” in a struggling business unit or team.
  3. Convert talent metric results into dollars
    If you want to continue to get everyone’s attention, you must make it a standard practice to convert talent management results into their dollar impact on organizational revenue and on other corporate strategic goals. Converting to revenue impacts makes HR metric results comparable to the results in the business impacts of other business functions. The language of business is “dollars”, so convert your talent results into a language that everyone understands.
  4. Shift to a data driven decision-making approach for talent decisions
    Set the goal to increase the percentage of talent decisions that are made based on data, rather than emotion, experience or intuition. Raise the usage of metrics and the accuracy level of talent decisions to the accuracy level of finance and supply chain. Also measure the error rate of your talent decisions and attempt to get it down to the 6 Sigma level that is the target in other business processes.
  5. Focus on forward-looking predictive metrics.
    Almost without exception, traditional HR metrics tells you what happened “last year”. Forward-looking predictive metrics have twice the value to managers, because they provide enough warning so that upcoming problems can be mitigated. Include risk analysis in your forward-looking metrics, so that decision-makers can see both the probability and the likely costs of not properly addressing an upcoming talent problem.
  6. Show a visual trend line
    A number or metric that reflects a single point in time may not drive action. Some simply need a visual representation in order to recognize a trend and to get them excited. So add a visual trend line (graph line), so that everyone sees the historical direction, the current direction and the future duration of the metric.
  7. Include the cost of delaying action
    Because many will simply be satisfied with doing nothing, don’t forget to also calculate the dollar cost of delaying or doing nothing when you are facing a talent management problem. Your metrics should show that acting quickly when a problem or opportunity is identified has a high economic value and delaying action can result in the negative business impacts going up exponentially. Also recommend working with the CFO’s office to demonstrate the hidden or unintended costs resulting from excessive cost cutting in talent management programs.
  8. Managers must have easy access to “real-time” metrics
    Managers are the individuals that make most people management decisions. As a result, they must have direct access to “real-time” people management metrics, which cover what is happening today. These current metrics in most cases must be readily available on their mobile phone, so that they can be used for day-to-day decision-making.
  9. Make sure that your metrics are actually used for talent decisions
    Talent Management metrics needs to drive action and change the way talent decisions are made. So first there must be evidence that talent metrics are seen and examined by decision-makers (Note: the best way to improve visibility is to include talent metrics in standard business reports). HR must also be able to show that its metrics are actually being used by managers for people management decision-making. . And when they are utilized, that both people management and business results improve.
  10. You also need “why metrics” that reveal the causes of problems
    Most metrics simply make decision-makers aware that there is a problem. But if you want to actually improve results, you must go the next step and identify the root cause of problems (i.e. why things are happening). Identifying the root cause is an essential component of continuous improvement. For example, knowing that your turnover rate is 20% has little value and it won’t drive action because they don’t know the cause. However, if you reveal that you know the reason why people are quitting (i.e. a lack of training), and that that reason is preventable in 90% of the cases, you can convince managers that you can actually reduce turnover.
  11. Also include recommend actions for fixing the problem
    Metrics won’t result in improvement unless managers implement the right solution to the problem that the metrics point out. You can better drive managers to act if you provide decision-makers with action suggestions or recommendations. You will get a higher rate of acceptance to your recommendations if you include their probability of success and the “how-to” steps to implement them.
  12. Split sample proof is the most convincing
    The best way to convince cynical executives that a talent management program will work is to conduct a split sample. I generally recommend that you start with jobs that are already measured and quantified and that show immediate results if something new works (i.e. sales or call center jobs). Then you apply your new talent approach to half of the group and do nothing different in the other half, known as the control group. Then see how much your solution improves performance compared to the control group. That’s the way that product and drug managers prove their new product’s effectiveness.
  13. Show a correlation between using a talent program and improved business results
    Use correlations and other statistical tools to prove how the usage of a talent management approach positively correlates with business success. For example, as training hours go up 6%, the performance of the trained employee goes up 8%. Also always use statistical methods to identify the critical success factors that are required to make talent management programs successful.
  14. Calculate the improvement in revenue per employee.
    The single most important indicator of improved workforce productivity is showing the yearly improvement in the revenue per employee metric (corporate revenue divided by the number of FTE’s), or alternatively the ratio between labor costs and profit. This ratio has great value because it can easily be compared between different companies.
  15. Calculate the top employee performance differential.
    It makes no sense to spend a lot of time, effort and resources to hire, retain and develop top performers unless they perform significantly better than an average employee in the same job. Show the percentage increase in performance between a top-performing versus an average employee. Showing a large multiple helps to convince senior management to focus on top performers.

 

About Dr John Sullivan

Dr John Sullivan is an internationally known HR thought-leader from the Silicon Valley who specializes in providing bold and high business impact; strategic Talent Management solutions to large corporations.