Data-Driven Decision Making

Laboratory managers are charged with making many kinds of decisions on a near-daily basis. In fact, decision making is one of the key responsibilities of leadership in any organization.1,2 Even though most laboratory managers are leading organizations that create, evaluate, and report data as a consistent part of the delivery of their function, many decisions seem to be made from opinion and feeling rather than from data. As laboratory managers, we can greatly improve the quality of our decisions by being more aware of the data available and learning how to use it effectively in our decision making.

Available data

Fortunately, most laboratory managers have many kinds of data readily available:

  • Operational data that is often included in laboratory information management systems (LIMS)
  • People data that is often available from human resources (HR) systems
  • Financial data that is available from budgets and accounting reports
  • Quality assurance (QA) system data
  • Safety process data
  • External customer or internal client data
  • Vendor and supplier data

Examples of the various types of data lab managers have at their disposal.

Operational People Financial QA Safety Customer Vendor
Numbers of samples analyzed Staff performance Income Training records Safety incidents Number of customers Cost comparisons
Numbers of tests performed Previous experience Sources Audit outcomes Near misses New customer generation Repair and maintenance records
Time required to complete tests Levels of experience Costs Quality investigations   Customer management Delivery records
Test turnaround times
(TAT)
  Profit and loss, or delta to budget Instrument performance Customer feedback and satisfaction  
      Market data about trends and needs

These are sources of large and important data that can be used in decision making. There are many other small and isolated sources of data. One thing we can all do to become more data-driven decision makers is to become more aware of sources of data that are available to us, and to teach our staff where these data are located, and how to use them.

Making decisions

Laboratory managers are expected to make rapid, high quality decisions every day. Decision making is a skill at which we can all improve. One of the keys to making effective decisions is to find the objective foundation for the decision. There are two key processes to clearly establish that objective foundation, effective questions that sharpen and focus the actual decision to be made, and objective data that can be used to answer specific questions.

Our metaphor for using questions to establish the focus for a decision is “fencing the pasture.” We want to effectively establish the space in which the right decision lies—the fenced pasture needs to be big enough to contain the best decision, but not so large that we waste time exploring low value space. To find the optimum perimeter, we use a combination of open-ended and close-ended questions. The open-ended questions enable us to diverge and explore the space. Open-ended questions also invite other participants to share more freely and more broadly in the process, enabling the sharing of knowledge and information.3 The close-ended questions enable us to ensure understanding and have clarity about conclusions.

Related Article: 4 Ways to Improve Inter-Generational Decision Making

Communicating about how the decision will be made is also a key element of the process. Many employees aren’t aware of the different decision making options we have as laboratory managers. Communicating about these choices and being transparent about the decision making process helps to build trust in organizations.1 While there are many different decision making options, there are four that we use most often. The first is that the leader requests and receives input and then makes the decision—this is an important option for some of the most difficult decisions, like staff reductions and policy changes. The second is a group consensus decision. While not requiring full agreement from the participants, consensus requires that all participants support the decision moving forward. The third is group unanimity, which requires enough compromise and negotiation so that all participants agree with the decision. The fourth is majority rule.

Case studies

Data-driven decision making is perhaps best discussed through specific examples. The next three case studies will show how we can use objective data that most laboratory managers have available to improve the quality of the decisions we regularly face. The three examples will cover a wide range of common laboratory leadership decisions.

Promotions

Promotions are an exciting time for most people. However, too many promotion decisions are made for subjective reasons, such as: do I like the employee, is the employee liked by peers and/or is the employee a good worker? We can use a data-driven decision making process to take the mystery out of most promotions. There can be ample objective data available to use in the decisionmaking process, such as:

  • Has the employee consistently demonstrated an increase in responsibility?
    • Use QA or HR data to compare actual job delivery to the current and next job description or role document
    • Use operational data to measure the actual output of work
    • Compare the actual output of work to comparable peers
  • Has the employee consistently met or exceeded expectations?
    • Use HR data to compile performance history over the past couple of review cycles
  • Has the employee consistently demonstrated the required behaviors?
    • Use safety and QA data to look for trends
    • Use operational and HR data to see how consistently the individual has been performing

By using the available data in this decision, we can be confident that the employee is meeting our criteria for the promotion, and perhaps more importantly, we can clearly communicate why this individual was promoted. For staff who question the validity of a promotion, it is very important to point to objective data used in making the decision. Using these data and communicating their use will also inform peers how these decisions are made and encourage them to demonstrate the same level of performance to achieve their own promotions.

Capital Investment

In these days of do more with less, there can be significant competition between different labs or functions for limited capital investment resources. Using a data-driven decision making process can reduce the friction and discontent that arises from contested priority decisions like this. We can avoid feeling like the manager’s favorites get the capital or certain technology areas have important friends. Using data can also address the expectation for an even distribution of capital in an uneven world.

Laboratory managers have an abundance of data that can readily inform capital investment decisions:

  • Operational data that shows usage rates, growth, or decline of testing, average TATs, and delays
  • Vendor data that shows price and cost of ownership expectations
  • QA data that demonstrates reliability, uptime, and training issues
  • Financial data that works in cost and available funds
  • Strategic market assessments show trends and needs

Using these data together to build option prioritization and cost realities makes these decisions clearer and much easier to communicate to staff. Like the promotion example above, pointing to objective data reduces the complexity of personality in the decision-making process. It also makes it clearer to staff desiring more capital investment what is expected and what is required to improve their project in the organization’s prioritization.

Service Contract

No lab scientist welcomes instrument and equipment problems. Scientists desire robust instruments that enable their experiments without worry. To reduce the risk of equipment problems, many laboratories purchase service contracts to ensure that instrument problems are quickly addressed and repaired by qualified vendors. However, service contracts can have significant costs to the organization. There is sufficient data available to enable data-driven service contract decisions:

  • Operational data to indicate the demand and the average TAT for specific instruments
  • Customer/client data to indicate the importance of specific instruments and the TAT expectations
  • QA data to measure the reliability and downtime for specific instruments
  • Safety data to understand any broader issues related to instrument failures
  • Financial data that balances the cost of the service contracts with the benefits

One approach to using data-driven decision making in service contract options is to use data from the past 12 months and compare the calculated cost and impact of having a service against not having a service contract. The calculations can show which scenario generates greater value for the laboratory.

A data-driven decision making process uses data that is generally available to most laboratory managers. Using these data enable laboratory managers to be more objective in making important decisions. It also provides a significant benefit in enabling us to communicate clearly to the organization how the decisions are made. That transparency brings significant benefits in trust, engagement, and teaching staff what is expected to get the decision they desire.

References:

1. Daniel Coyle (2018) “The Culture Code” New York, Bantam Books
2. William A. Cohen (2008) “A Class with Drucker” New York,
AMACOM
3. Amy C. Edmondson (2019) “The Fearless Organization” Hoboken, NJ, Wiley