Two experts predict that artificial intelligence will change the way companies see themselves

However, in the same study, an even greater number of respondents – 84% – said their company culture was not where it should be. Again, the data confirms their intuition. The average culture rating for gigantic employers in America on Glassdoor, a site that allows employees to rate their employers, is 3.6 out of 5. Few would be excited to eat at a restaurant or ride with an Uber driver with such a rating. Similarly, few employees will want to spend 40 hours a week in an average culture.

Building and maintaining a vigorous corporate culture can be even more challenging in organizations where employees work remotely. An ongoing study shows that companies where employees are most effusive about remote work perform worse than their peer groups in terms of corporate culture, especially opportunities for learning and development and sincere communication.

Leaders can’t improve what they can’t measure. Unfortunately, the most popular tool for measuring corporate culture – the engagement survey – has solemn limitations. When faced with a long list of questions, employees go on autopilot and assign each question the same or similar score. Employers who, like many, ask dozens of multiple-choice questions may only obtain a few reliable insights because of the scattered nature of respondents. Even if employees ask a question, the result gives no clues on how to improve the situation. What if a topic the employee really wanted to cover wasn’t covered?

Recent advances in artificial intelligence—particularly gigantic language models (LLM)—are enabling leaders for the first time to gain detailed insight into corporate culture based on how employees talk about their company in their own words. Instead of answering tons of questions on a five-point scale, employees can now simply explain what’s working and what’s not working in their organization and offer suggestions on how they can improve it. Artificial intelligence can do the weighty lifting, providing much more detailed comment classification and sentiment assessment.

Freed from the shackles of time-honored surveys, organizations can utilize artificial intelligence to collect and process employee feedback from multiple sources. The amount of feedback available is staggering. The combination of free text from internal surveys, performance feedback to managers, online employer reviews, and other sources adds up to tens of thousands of pages of data per year for a gigantic company. Until recently, organizations had to rely on primitive tools such as word clouds or search keywords to gain insight into this source of information.

Armed with more and more detailed measurements enabled by AI, executives can more quickly and easily assess whether their company is adhering to what they consider “core” values, identify key cultural elements influencing everything from employee attrition to innovation, and diagnose toxicity. subcultures in the organization and analyze progress over time. Once they spot significant patterns, leaders can review the raw feedback for more detailed context and employee recommendations for improving the culture.

Take Amazon, which aspires to be the best employer in the world. We used our AI platform to analyze tens of thousands of employee opinions about the e-commerce giant. This showed that Amazon was good at many of its leadership principles, such as “customer obsession” and “invent and simplify.” But company culture also contributes to burnout, especially among software engineers, who are twice as likely to report burnout than warehouse workers or drivers. Raw employee feedback highlights ways Amazon can reduce engineering stress, such as fixing a performance review process widely viewed as brutal or minimizing disruptions at night when tech workers are on duty.

Even the largest companies will not rush to adopt artificial intelligence. However, cultural analysis is one of the few areas where it can be applied right now because it leverages one of AI’s greatest strengths today: understanding natural language at scale.

This does not mean that leaders should blindly trust LLM results. Tools require safeguards to protect against vulnerabilities such as hallucinations with made-up answers. Models should measure elements of culture based on solid evidence, not the latest management fad. Leaders must take a broad view of culture, measuring not only the factors that influence employee satisfaction but also the topics that shape a company’s ability to adapt to market changes and avoid unethical or illegal behavior.

Leaders who actually utilize AI for cultural analysis can utilize it to make their employees happier, reduce the risk of a PR disaster, and ultimately raise their profits. Measurement is not the only piece of the “culture of success” puzzle, but it is crucial. Culture has always been the puzzle at the heart of organizational performance: undoubtedly significant, but impenetrable. Thanks to artificial intelligence, significant progress can be made in deciphering It.

Don Sull is a professor at the MIT Sloan School of Management and co-founder of CultureX, a research and AI company. Charlie Sull is the co-founder of CultureX.