Why AI is making your team members faster, but not your business

Many of your team members already work faster with AI. Here, we share how to carry that speed through to what your teams deliver, what we learned rebuilding our workflows, and three questions for leadership teams.

Most organisations have put AI tools in their team member‘s hands, and many of those people will tell you their own work is faster for it. Leaders are now rightly asking how that same speed can show up in the work their teams deliver together.

Australian businesses are telling the Reserve Bank of Australia they are “not seeing productivity yet”, as Governor Michele Bullock said in September. She noted that when the Bank of Korea asked workers instead, they reported saving up to an hour and a half, and taking that time as leisure. The time is being saved, and so far it is staying with the person who saved it.

Those gains come from tasks one person does alone, like drafting a document, summarising a meeting or analysing a dataset. Most work, though, passes through several hands before it is finished, waiting on briefs, reviews and sign-offs along the way. When one person speeds up while the team’s workflow stays the same, the work still moves at the team’s pace. 

Software development, one of the first fields to adopt AI widely and one where code is tracked at each stage on its way to release, shows how wide that gap can be. A US National Bureau of Economic Research working paper covering more than 500,000 developers found AI lifted code changes by as much as 240% but finished releases by only about 30%. The extra work was held up at review, testing and release.

Closing that gap means rebuilding how work moves between people, and teams rarely manage it alone. It takes time to look at how a workflow runs, and it takes people willing to share what they have built for themselves. Both tend to fall to the bottom of the priority list, and the teams with the most to gain are often too busy delivering to step back. Until a leader makes room for it, it does not get done.

That was the decision our own leadership team faced. Our team members had created thousands of AI skills, but most served only the person who created them, and several teams had built near-identical versions without knowing it. So, we invested in Shift Up – our internal AI capability program where every team member across every market steps away from day-to-day delivery to rebuild their entire workflows as a team with AI at the centre – beyond tasks. 

Three questions for leadership teams:

1. Who benefits from what your team members have built with AI? If the answer is mainly the people who built it, your AI investment is making individuals faster without changing what your teams deliver. The return grows when the strongest skills reach every team doing similar work.
2. Which of your key workflows would you design differently if you started today? Many were built around people handing work to one another, step by step. Challenging this assumption surfaces steps that require human judgement versus those that can be coordinated by an AI agent or automated entirely. Our article Augment, automate or agentify? sets out how to decide.
3. What should the time AI frees up be used for? Decide in advance what the time saved is for, redirecting teams towards higher-value work. Then track how long a piece of work takes from request to completion, because AI adoption figures alone are not reliable measures of productivity.

If you are working out how to move AI gains from individuals to teams, reach out to our team.