Gallup’s ‘State of the Global Workplace 2026’ report shows global employee engagement has dropped to 20 percent, its lowest point since 2020, costing the world economy approximately US$10 trillion in lost productivity last year.
These numbers exist alongside the largest AI investment cycle in business history. Research cited in the report found 95 percent of organizations had seen no measurable profit impact from recent AI investment.
The technology is not the problem – it’s the way change is being led.
I work with organizations across Australia and I can almost guarantee what happens when I ask to see their current strategy document. If it’s more than two years old, they hand it to me with the same heads-up: “A lot of this has changed now. It’s not really accurate anymore.”
The strategy is still in play, but the conditions it was written for have already changed.
Balancing the cost of waiting with the cost of trying
Technology engineer James Galdes led South Australia’s digital response to the COVID-19 pandemic. His team was tasked with building a functioning system to track the state’s movements and identify exposure chains across at least three government agencies that, in Galdes’ words, don’t typically work that well together.
Overthinking has diminishing returns. The best way to learn is to release, observe and adapt.
Under normal conditions, that project would take years. His team had three weeks. The usual machinery of government, the committees, the policies, the documentation, had to move aside.
“We didn’t have time to get it perfect,” Galdes told me. “We just had to get it working and then make it better every day.”
Once live, the team iterated fast. They added test bookings, connected pathology reports, then built a home quarantine app using geolocation and face verification to check compliance and free up police resources. The app evolved from a compliance tool into a platform supporting mental health and symptom monitoring.
“We shipped different versions of the app as we trialed concepts,” Galdes said. “It evolved a lot as we learned in real time.”
The entire pandemic tech response took around three months. South Australia’s approach was later adopted nationally.
The technology is not the problem – it’s the way change is being led.
Galdes reflected that none of it would have been possible under normal conditions. They stopped over-engineering. Overthinking has diminishing returns. The best way to learn is to release, observe and adapt. Risk tolerance expanded not because risk disappeared, but because the cost of waiting was higher than the cost of trying.
Once the crisis passed, Galdes noticed how quickly they reverted to plans, hierarchies and fixed processes. You’d think the learnings would stick, but they rarely do. Most organizations aren’t forced to strip the process back and make testing the only option. They default to the rollout. And it costs them.
The cost of skipping the test
In the 2010s, organizations across Australia and the world spent millions tearing down their walls. Architects and executives believed that open spaces would spark creativity, collaboration and culture. The logic seemed sound.
Sociology research had suggested that proximity drives interaction and surveys suggested people liked the idea of fewer barriers. The change was framed as evidence-based. What it lacked was behavioral data. No-one had measured what actually happened when real people sat in these environments day after day.
That’s what Harvard researchers Ethan Bernstein and Stephen Turban set out to do in 2018. They tracked two Fortune 500 companies as they shifted thousands of employees into open plan spaces. Using wearable devices to measure real-time interaction, they captured what people actually did, not what they said they’d do on a survey. The result was the opposite of what the broad studies had predicted.
The old playbooks are expiring faster than organizations can replace them.
Face-to-face interaction fell by around 70 percent. Communication via email and instant message spiked instead of dropping. Employees withdrew, put on headphones and did whatever they could to carve out privacy. Instead of rising, collaboration collapsed. The evidence wasn’t wrong. It was just incomplete. Surveys and theories told one story. Lived experience told another.
What if they’d tested it on one floor first? The data would have told a different story. The decision that followed would have been better and saved them significant costs.
This is the gap experiments close. Not guesswork or endless pilots, but a structured test. A clear hypothesis, a time frame, defined metrics and a decision point before you begin. The smallest version of a change you can run with real people, in real conditions, before you commit to scale.
Before the next change program, ask three things:
1. What is the smallest version of this we could run and with which team?
2. What would we measure to know whether it’s working?
3. What is the decision point where we adapt, anchor or abort?
The old playbooks are expiring faster than organizations can replace them. What differentiates the ones that thrive isn’t their strategy. It’s their capacity to learn as fast as the world changes. Run the experiment.
