At DCP Midstream, Technology and Data Power-Up Talent Acquisition

Headquartered in Denver, Colorado, DCP Midstream is a Fortune 500 company with approximately 2,700 employees and operations in 16 states. The company is one of the largest producers of NGLs (natural gas liquids) and one of the largest natural gas processing companies in the U.S., transporting or processing nearly 25% of the country’s natural gas supply.  

DCP Midstream is not a typical oil and gas company. The company recognizes that people are its greatest asset, a team of progressive thinkers who are driven to challenge the traditional and identify revolutionary efficiencies through DCP 2.0, a company-wide technology and innovation initiative. 

This strategic focus on innovation enabled DCP Midstream’s Human Resources department to embrace new technology and incorporate real-time data to transform recruitment and hiring processes across the company. Specifically, the Talent Acquisition function implemented programming that uses real-time data to allocate recruiting resources. This allows the team to prioritize resources and fill requisitions more efficiently, and the company is seeing impressive results.

This case study represents DCP Midstream's submission to the i4cp Next Practice Awards. The awards will be presented at the i4cp 2019 Next Practices Now Conference.    

Business Challenge 

DCP Midstream works to operate reliably and efficiently, while reducing fiscal waste wherever possible. This includes functioning as a lean organization, meaning each employee’s contribution is extremely valuable. Therefore when a full-time equivalent (FTE) resource is needed, the focus is on both time and quality. The Talent Acquisition team’s challenge is to effectively utilize recruiting resources to fulfill the needs of the organization in a timely manner, resulting in quality hires. Historically, meeting the demands of the business has been hindered by limited resources and lack of real-time insight into the status of the recruitment process. 

From a technology standpoint, the challenge was to gather reliable data from multiple sources. Those include Taleo, DCP Midstream’s applicant tracking system (ATS), and Smartsheet—cloud-based software enabling collaboration between internal and external recruiters and enhanced by both real-time and historical data. Historical reporting solutions have been unreliable and inaccurate due to the functionality constraints of the ATS, and have been unable to meet requirements for the business.  

Solution 

In order to utilize in-house recruiting resources effectively, DCP Midstream’s corporate recruiting structure shifted to a full recruitment process outsourcing (RPO) model beginning in January 2018. Prior to adoption of RPO, there was no vendor benchmarking data to define “what good is.” In order to define what good looks like, both qualitative and quantitative benchmarks were established based on current vendor performance and industry recruitment benchmarks.  

Quantitative benchmarks include: time to first candidate, average time to screen, average time to select, average time to process, and total submittals per requisition. Qualitative measure is the conversation ratio of submittals to interviews conducted. However, implementing RPO and establishing benchmarks were not enough to drive the function forward into a data-driven center of excellence (COE). As a result, a data analytics solution was developed utilizing Alteryx, a robust data blending tool, to combine historical and real-time data to give recruiters a holistic view of a requisition’s lifecycle. Providing recruiters with a holistic view enables them to strategize resources based on data and take action. In turn, empowering effective support for multiple customer groups.  

Prior to implementing the data analytics solution, the Talent Acquisition team audited the capabilities of current systems and researched additional tools that might be needed to drive this solution. Next, the criteria of benchmarks to be measured had to be established. Then the team identified the best tool for reporting and the best source for real-time data. A solution was built, tested in iterations with the team, and then implemented.  

Along the way, the team encountered multiple barriers, including human error, limitations of current system capabilities, and multiple data sources (on-premise and cloud). It took roughly nine months prior to the successful implementation of the solution: three months to implement the RPO solution, six months to establish benchmarks (three months to identify the first iteration, which included gathering enough data to establish a baseline, and three more months to finalize the benchmarks), and 120 hours to develop the data analytics solution.  

Results

Since implementation of the data analytics solution, Talent Acquisition has been able to reduce average time to fill (TTF) by 14 days, which represents an 25% decrease. Currently, the team is conducting the company’s year-end talent review process. Based on the newly implemented recruiting efforts, it is expected that individuals hired in 2018 will have higher initial talent review scores than the comparable scores of 2017 hires. It is anticipated that those higher initial talent review scores are attributable to the 2018 hires having a better candidate experience and being more prepared for their roles because of the new qualitative focus on recruiting, effective allocation of resources and stronger relationships with hiring managers.  

Conclusion 

As the war on talent progresses, the HR and Talent Acquisition teams at DCP Midstream are committed to utilizing data analytics in new and innovative ways to ensure that resources are prioritized to Hire Great, Inspire Awesome. Currently underway is the iterative process of taking the analysis produced and integrating it with benchmark expectations to identify each individual recruiter’s weekly priorities. At this point, the benchmark expectations integration is produced manually.  

There were definitely some lessons learned from implementing this solution. The biggest was having a more robust process in place to reduce human error. Data integrity is a “garbage in, garbage out” process. As a result, reducing human error would have enabled faster achievement of the solution. 

Next steps will involve automating that process and producing interactive dashboards that allow for real-time feedback and adjustments. Future plans also include layering on estimates and producing predictive analytics to help project required time to produce results and approximate expectations for hiring managers. This will provide a more accurate reflection of the time required to fill each individual requisition based on historical, current, and predicted analysis.

Michael Chen and Kourtney O'Connor are the HR Data Analytics Lead and the Talent Acquisition Business Partner  respectively for DCP Midstream.