The Status of Metrics: A Look at Top Issues Facing Retail Organizations

During a recent best practices conference covering the restaurant industry, I was given the opportunity to sit down with the leadership of some of America’s largest restaurant chains to discuss their top issues with regards to using metrics for talent management. What follows is an overview of some of the issues that were discussed and tips that emerged. While this event was targeted to the restaurant industry, many of the issues that were brought to the table are similar if not identical to those that any firm approaching HR metrics might encounter, especially those firms with multiple product lines and units of operation. The issues that quickly rose to the top as challenges for nearly all firms attending were centered around four key areas: including selection, distribution, getting benchmark data, and creating accountability or visibility for specific measures. Each of these four key areas is highlighted below. Selection of Appropriate Measures This issue was shared by firms who were just getting ready to start metrics initiatives, as well as by firms that had fairly established measurement programs. Some of the key questions that arose dealt with:

• How to select measures to support decision making at multiple levels

 

• How many measures should be tracked
• To what degree of detail should processes be monitored
• What the appropriate mix of measures focusing on effectiveness versus efficiency is

When selecting measures, it is important to choose metrics that reflect the goals or objectives of the organization, and more specifically the primary interests of the individual or individuals that will most benefit from the information metrics provide. It is also important to limit the number of metrics to those that help power decision-making. All to often, firms select too many measures, either because their technology infrastructure makes them possible, or because they try to satisfy each and every person’s personal interests. Most of the metrics in use today focus on transactional efficiency or transaction counting. Such metrics are easy to implement and maintain, but typically offer little value to managers outside the human resource function. Some of the tips that emerged during the conversation on this topic included:

• Limit the metrics selected to those that directly support the objectives that managers are compensated for. If objectives are not related to pay or rewards, they are perceived as not being as important as others that are, and will correspondingly receive less attention.
• Limit the number of metrics selected to support any one person or function to five or less. Selecting a large number of metrics will eventually barrage managers with too much for them to make use of. Management time is limited, so you will want to select just a few measures that at a glance help managers react to emerging issues.
• Drop metrics that show little variance from period to period, or lengthen the period between which you calculate the metrics. In other words, if you are reporting something such as “high-performance turnover” on a weekly basis and the metric fluctuates from week to week by just a few tenths of one percent, consider changing the frequency of the metric to focus on monthly or possibly even quarterly “high-performance” turnover.

It is important to note that metrics really fall into two categories, those that report historical performance and those that help power decision-making. Metrics that focus on historical reporting are executed and tracked over a long period of time, while metrics that power decision-making may be selected periodically and executed for a short period of time. Distribution to Multiple AudiencesDistributing metrics has long been an issue for organizations large and small. Some practitioners argue that managers should only have access to the information that pertains to their organization, while others argue that all managers should have broad access. Some feel that metrics should be distributed weekly, some monthly, and some on a virtual basis. One trend that many firms present seemed to have adopted is the regular production of a “scorecard” that reports numerous standardized metrics to multiple layers of the organization. While this approach seems like it would provide a great deal of useful information, it unfortunately leads to a lack of adoption and utilization. One CEO noted, “I simply don’t look at the report anymore; it’s too long” (the report in question was three pages). Another COO asked what others do with their data, most of which doesn’t pertain to decision-making. Some of the key questions that arose during conversation on this topic included:

• How often should managers receive reports?
• What format should reports be provided in?
• Who should get access to what?

It was overwhelmingly clear that few firms were satisfied with how they were reporting metrics. In response to the discussion, some of the key points of consensus were:

• Provide managers with only those metrics that help them make decisions relevant to their span of control and that pertain to the near future, i.e. just in time.
• Allow managers to view performance summaries for units similar to theirs, so that they can identify who in another unit might be doing something better than they are. Distributing performance measures of like units creates a performance culture. With the exception of lists such as reduction-in-force rankings and the draft, few people like to be at the bottom of a list.
• Present metrics in a format that is easy for managers to glean information at a glance from, i.e. graphically with short narrative summaries. Presentation of statistical tables requires more concentration to extract information relevant to making decisions on the fly.

Spotlight example: One leading specialty beverage retailer has correlated in-store retention/turnover with unit sales performance. Instead of presenting historical turnover numbers, they provide managers with a projection of lost sales or increased sales that result from talent turn. Getting Benchmark Data Many of the firms represented participate in multiple benchmarking efforts, including those provided by the Saratoga Institute, Corporate Leadership Council, and the event host People Report. However, nearly every participant indicated a need for more specific highly specialized benchmark information. Some of the questions that steered this course of discussion included:

• How can you find valuable sources of external comparison data, when so many firms execute practices differently and benchmark surveys do not account for the variations? (The use of outsourcing was mentioned several times because it alters headcount numbers and can impact metrics such as cost per hire and revenue per employee.
• How can you benchmark with competitors in an industry that historically does not share information, and which regards numerous practices as trade secrets?
• How can you get benchmark information without breaking the bank?

The conversation on this topic was relatively short. Apart from benchmarking internally, and very limited sharing of information between firms, most participants were still developing methods to benchmark externally. To help develop strength in this area, we discussed two tools that have been proven in other areas, developing practices for trade and creating learning networks. The first practice we discussed, developing practices for trade, basically means that you identify what practices you currently have that could be considered best practices and that competitors might want to learn from, and then propose that you trade your knowledge and learning in exchange for theirs on a practice you have interest in. The second practice provides all of the benefits of the first, but helps each firm retain most if not all of its confidentiality. Learning networks are informal grouping of leaders from various firms inclu,ing competitors and non-competitors, that have come together to learn more about a particular topic by sharing information and learning. The groups can share information via email chains, online discussion forums, teleconferences, and face-to-face meetings. For learning networks that want to retain the greatest degree of confidentiality, companies can get together and hire a local professor or consulting firm to coordinate the group, extract company-specific information, and produce results that are scrubbed of identifying or private data. Creating Accountability or Visibility for Specific Measures The last topic, and by far the grandfather of all issues with regards to metrics, deals with creating accountability or visibility for specific measures. Much like the scenario mentioned earlier where the CEO stopped reading reports, many organizations have mangers who simply don’t pay attention to or make few attempts to use the information HR can provide. In response to this issue, several points emerged including:

• Translation of reports that present information in “HR speak” to reports that present information in standard business terminology. All too often, reports from HR present information in the language of HR, despite the fact that line managers and managers from other functions don’t speak or particularly trust HR. To have true value to other managers, information from HR must be presented in standard business terminology and translated into terms they understand. For example, HR often reports on turnover; however, to the average line manager an increase in turnover from 4% to 6.2% means very little. If the turnover increase were translated into dollar impact though, and the manager was informed that the increase in turnover would result in a $486,000 increase in operational costs for her unit, the information suddenly becomes more valuable.
• Elimination of non-essential metrics. Barraging people with information only forces them either to dedicate more time to reading it or to ignore it, which happens more often than not. By eliminating standardized reports that deliver lots of historical performance numbers to managers and focusing on only providing those metrics that managers want and deem necessary to support decision-making, the visibility of HR to provide information of value will increase.
• Incentives. Simply reporting information will not necessarily increase the accountability of managers to talent management issues. Objectives with both positive and negative rewards/consequences must be introduced and formalized.

Conclusion By the time this event ended, it was clear to all who participated that many organizations are still in the infancy stage with regards to designing, implementing, and using metrics as a foundation for decision science. What was also very clear is that metrics will be a critical component of future management systems in HR, and that their importance will only increase as time passes.