Education is step one. Changing behavior and unbiasing decisions is a very difficult step two.

Google is approaching this challenge in a few different ways, one of which relies on structure. Use a structured process to guide decision-making, as it can help reduce the potential impact of unconscious bias. In the workplace, a structured decision-making process could include:

• Clearly articulated success criteria: Determine the components of what an ideal decision should look like and write down examples, prior to making the decision.
• A shared understanding of success: Decision-makers should be calibrated on success criteria and use them to guide their decision.
• An opportunity for discussion: When possible, it’s great to allow decision-makers to discuss their recommendations and rationale, to avoid assumptions and allow for clarity.
• A predetermined decision scope: Decision-makers should know what’s on the table for discussion to avoid superfluous information from entering the conversation.

For example, in the case of hiring, research shows that in the absence of predetermined criteria unconscious biases can lead to shifting and unfair expectations when assessing potential hires. One study showed that clearly stating expectations and standards ahead of time can reduce the influence of gendered stereotypes on hiring decisions. Applying “structure” to hiring doesn’t need to be hard - write down the minimum qualifications needed to do the job, share them with all the interviewers, and ensure everyone is asking questions to assess those qualifications. Better yet, use structured interviews.

By structuring the processes by which people decisions are made, you can begin to mitigate the potential impact of unconscious bias.

Researchers found that judgments made in the first ten seconds of an interview could predict whether the interviewer would hire the candidate or not. But snap judgments are not necessarily predictive of a candidate’s future performance. What’s demonstrated here is confirmation bias - interviewers can make an initial judgment and spend the rest of the interview unconsciously looking for evidence to confirm what they already believe and dismissing contradictory evidence.

Research shows that using structure and consistent questions can help improve the validity of interviews. To address this, Google often uses structured interviews which use a predetermined rubric to assess candidate responses fairly and consistently, and the best hire is selected for the position.

Google often uses committees to make some of the most important people decisions. Making decisions by committee might be tough or uncomfortable at times, but research shows that teams with divergent opinions can make better, less-biased decisions.

For example, hiring committees are used to review and assess candidates in the final stages of the hiring process. While decision by committee may require extra time, the trade-off is often worth it as it’s far more productive to make the best hire the first time than deal with a bad hire over time.

Another example is Google's use of committees in the promotion process to discuss and make decisions. The committee discusses criteria for promotion first, and then collectively assesses all promotion candidates using that same criteria.

If you’re going to have multiple committees of any given type, it’s important to know that the committees are making decisions based on the same set of expectations and criteria. Calibration committees are designed to set and maintain the decision-making bar and are crucial for keeping processes fair and consistent. At Google, calibration committees help ensure consistency across performance management and promotion processes. Yes, committees on committees might seem overly bureaucratic, but it can be helpful to make sure you're achieving the best outcome.

Research suggests that checklists can help reduce the influence of unconscious bias in decision-making. Unbiasing checklists are integrated into some of Google's people processes in an effort to highlight common unconscious biases and provide employees with targeted unbiasing strategies. Googlers are encouraged to use these checklists to mitigate their own, and others’, unconscious biases.

Research shows that empowering employees to voice their opinions supports inclusion and perceptions of fairness. Google ran a controlled experiment to test whether giving managers an unbiasing checklist changed their perceptions of fairness and voice in calibration committee meetings, where managers discuss and evaluate the performance of their employees.

Managers were randomly assigned to one of three conditions:

1. Managers were emailed the unbiasing checklist in advance of the meeting.
2. Managers were emailed the unbiasing checklist in advance and meeting leaders gave out hard copies, encouraging participants to call out unconscious bias when they noticed it.
3. Managers experienced no changes (control).

Results were complex. When the checklist was simply emailed to managers prior to the meeting they actually perceived significantly less fairness than managers who did not receive the checklist. Why? The hypothesis is that the “ignorance is bliss” effect occurred - when educated about unconscious bias, managers became more likely to see it in themselves and others. However, when participants were given printed handouts of the checklist (condition two), they reported significantly greater perceptions of fairness, as they were given the “right” to call out unconscious bias and the tools to do so effectively.

Education, action, and a platform to call out unconscious bias are a powerful combination Google uses to inform unbiasing strategies.