Still using the staff engagement survey? It's time to get down into the data

Banking behemoth, Wells Fargo, made global headlines in 2020 when it was revealed that it had paid just shy of $3 billion to the US Department of Justice and the Security and Exchange Commission. The fine, one the biggest settlements in corporate history, marked the end of a drawn-out cross-selling scandal that saw rank-and-file employees create fake customer accounts to meet aggressive quotas set by the bank’s senior executives. By 2016, Chief Executive, John Stumpf, was obliged to stand down. Since 2020, it has been subject to a stringent asset cap imposed by the Fed to limit the fees it can charge its customers.

Meanwhile, 2024 saw aeoroplane manufacturer, Boeing, agree to pay new fines related to two fatal 737 Max 8 crashes in 2018 and 2019. The US Department of Justice reported that Boeing had failed to “design, implement and enforce” a compliance and ethics program to prevent and detect violations of U.S. fraud laws in the company’s operations. Rolling out that programme was a condition of Boeing’s original settlement, which also carried a $2.5 billion sanction.

The Boeing crashes were the result of a flawed manoeuvring system designed to help avert stalls in flight; something that Boeing is accused of deliberately covering up. In total, they claimed the lives of 346 people.

The Boeing and Wells Fargo cases provide something of a cautionary tale for organisations in every sector. And they beg a critical question: how do you gauge the health of your organisational culture? How do you know if core values and ethical principles are being upheld; or if there are tensions or problems gathering momentum beneath the radar that could put your business and your customers at risk?

Keeping track of how your workforce is feeling, how they are behaving, and if that behaviour aligns to desired culture norms is critical to the long-term wellbeing and sustainable growth of your organisation. All too often, leaders miss key cues and warning signs that something may be going wrong somewhere within the business.

The limitations of traditional surveys

Tried and tested means of taking the organisational pulse such as staff surveys or questionnaires can be limited in their effectiveness. The validity of survey answers assumes that the right questions have been asked. Then there’s the issue of anonymity. Even if employees are told that their responses will be treated in full confidence, all too often they will have some fear of consequence if their feedback is critical or negative in some way—especially if aggregate staff survey scores are openly tied to managerial bonuses.

An alternative and potentially more revealing way of gauging cultural health is to use unstructured textual data: to leverage the wealth of employee and stakeholder information, insights, perspectives, concerns and reviews that exist across the plethora of existing online platforms and sources within and beyond the organisation—and use all of this data to bring underlying issues to the surface.

In our research with organisations in healthcare and other sectors, we have been able to pinpoint times where bosses set “target pressure:” where they start squeezing employees to deliver excessively ambitious or risky targets, like the unrealistic sales goals set by Wells Fargo, or the oil production quotas instituted by BP prior to the Deepwater Horizon oil rig explosion in 2010. Using online employee feedback—unstructured textual data from social media, platforms like anonymous career intelligence website, Glassdoor, and others—we can apply natural language processing methods to collate candid, unvarnished employee insights into organisational culture in real time. This gives us a snapshot, warts and all, of the kinds of tensions at play deep within the business that are shaping employee behaviours and responses, as well as the risks that these might engender. And it’s not always black and white.

Crunching the data in many of our studies, we find that alongside negative associations—where analysis reveals fears over “shark tank mentality,” “cutthroat,” “competitive and fear-driven” organisational culture, for instance—there can also be plenty of enthusiasm and relish for target pressure that is “results-driven” or that provides “autonomy.”

In today’s disrupted and change-prone environment, leaders need to be able to get down into the weeds of organisational culture as part of their scenario-planning. The good news is that we access to the data and to new methodologies to empower leaders to do just that: to truly grasp the day-to-day reality of the people and the teams they manage.

In our new programme, Leading Organisational Culture, we walk participants through some of these methodologies along with a raft of real-world case studies and examples, to empower organisations in the use of big data and AI in diagnosing cultural risks and issues. The programme leverages our research to arm decision-makers with new tools, procedural frameworks and effective practices to shape behaviours and optimise outcomes. Our aim is to provide a new, data-powered and evidence-backed lens to understand organisational culture—to spot and act on issues before they mushroom into problems.

Of course, not all problems have the potential to become a scandal as acute as the crises that engulfed and continue to engulf Boeing and Wells Fargo. But experience and history tell us that problems do arise, that they arise frequently, and that they have a nasty tendency to proliferate before leaders even realise that they have an issue. Crises happen. And in a world where complexity and change are accelerating at scale, never was it more the case that prevention is better than cure.

 

 

You can read more about this topic in Dr Tom Reader and Professor Alex Gillespie's recent paper:

Reader, T. W., & Gillespie, A. (2024). Target pressure and corporate scandals: a natural language processing investigation of how organizational culture underlies institutional failures. European Journal of Work and Organizational Psychology, 33(6), 855-867.

https://www.tandfonline.com/doi/pdf/10.1080/1359432X.2024.2398181