Skip to content

Posts from the ‘Analytics’ Category

Podcast: Creating HR Analytics That Matter

Kronos analyticsYesterday I chatted with our board members David Creelman and William Tincup about what organizations need to do to create HR analytics that matter to the business.  Big Data is one of the linchpins of the big 4 SMAC themes in technology today: Social-Mobile-Analytics-Cloud.  Deployed alone and in various combinations, these technologies continue to transform the way work gets done. For many people,  harnessing the power of big data remains a new frontier.  In this podcast, David and William share their thoughts on some of the following questions:

  • If you’re new to analytics, how do you get started?
  • What skills do you need to implement analytics projects?
  • How do you present the results of your data analysis to senior management?

You can listen to our conversation here: 

What have you done to embed more data-driven decision making into your organization?

A Conversation About Making Big Data Small


big data whisperer
I recently had the opportunity to interview  Aram Faghfouri, senior data scientist at Kronos; Holger Mueller, vice president and principal analyst at Constellation Research; and David Wright, vice president of architecture at Kronos, about the challenges and rewards associated with making big data small.  Everybody’s talking about Big Data, which is increasingly becoming a major strategic focus for firms that sell technology and consulting services.  The mountain of data that the former help organizations create provides an equally large opportunity for the latter to help interpret.

For many organizations, mining their transactional data for analytical “pots of gold” can seem both daunting and potentially dangerous when they begin to contemplate issues like data security and privacy.  Aram, Holger and Dave shared their views on the following topics that may help you transform your mountain of data into nuggets of insight:

  • What kinds of data are valuable to workforce management leaders;
  • Practical advice on how to get started;
  • Privacy concerns and projections for how they will be addressed; and
  • The skills gap surrounding big data and strategies for finding success.

 You can listen to our conversation here.

What are you doing to leverage data to transform your workplace?

The Scientific Method isn’t Just for Scientists

scientific_method_wordleToday’s guest post is by Sharlyn Lauby, the HR Bartender and a member of the Workforce Institute board of advisors.  Sharlyn writes about how the scientific method of investigation can be applied to solving problems in a business environment.  This topic is near and dear to my heart as I was a scientist and science teacher early in my career.  Sharlyn is right on in her analysis about how this method can help non-scientists to find the right solutions.

Companies face challenges on a regular basis. As such, employees need to know how to problem solve. A tried and true problem-solving process is the scientific method. I know many of us haven’t thought about the scientific method since our school days but it does provide a logical way of tackling business problems. As a reminder, here are the steps to the method:

1.  Identify the problem. The first step in the scientific method is to identify and analyze a problem. Data regarding the problem can be collected using a variety of methods. One way we’re all accustomed to is the classic: who, what, where, when, how, and to what extent? The scientific method works best when you have a problem that can be measured or quantified in some way.

2. Form a hypothesis. A hypothesis is a statement that provides an educated prediction or proposed solution. A good format for a hypothesis would be, “If we do XX, then YY will happen.” Remember, the hypothesis should be measurable so it can help you solve the business problem identified in step one.

3. Test the hypothesis by conducting an experiment. This is when an activity is created to confirm (or not confirm) the hypothesis. There have been entire books written about conducting experiments. We won’t be going into that kind of depth today but it’s important to keep in mind a few things when conducting your experiment:

    • The experiment must be fair and objective. Otherwise, it will skew the result.
    • It should include a significant number of participants or it will not be statistically representative of the whole.
    • Allow for ample time to collect the information.

4. Analyze the data. Once the experiment is complete, the results can be analyzed. The results should either confirm the hypothesis as true or false. If by chance, the results aren’t confirmed, this doesn’t mean the experiment was a failure. In fact, it might give you additional insight to form a new hypothesis. It reminds me of the famous Thomas Edison quote, “I have not failed. I’ve just found 10,000 ways that won’t work.”

5. Communicate the results. Whatever the result, the outcomes from the experiment should be communicated to the organization. This will help stakeholders understand which challenges have been resolved and which need further investigation. It will create buy-in for future experiments. Stakeholders might also be in a position to help develop a more focused hypothesis.

Now let’s use the scientific method in a business example:

Step 1 (identification): Human resources has noticed an increase in resignations over the past six months. Operational managers have said that the company isn’t paying employees enough. The company needs to figure out why employees are resigning?

Step 2 (hypothesis): If we increase employee pay, then fewer resignations will occur.

Step 3 (test): For the next three months, HR will have a third-party conduct exit interviews to determine the reason employees are resigning.

Step 4 (analysis): The third-party report shows that the primary reason employees are leaving is because health care premiums have increased and coverage has decreased. Employees have found new jobs with better benefits.

Step 5 (communication): After communicating the results, the company is examining their budget to determine if they should:

  1. Increase employee pay to cover the health insurance premium expense or
  2. Re-evaluate their health care benefits package.

I’ve found using the scientific method to be very helpful in situations like the example where a person or small group have a theory about how to solve a problem. But that theory hasn’t completely been bought into by everyone. Offering the option to test the proposed solution, without a full commitment, tells the group that their suggestion is being heard and that the numbers will ultimately provide insight – after the full scientific method has been followed.

Have you ever used the scientific method to solve a business problem? Share your experience in the comments.