We Tell Leaders Culture Matters, But Then Reward Them for Something Else.
In discussions with HR leaders recently, I have been hearing a lot about culture being an issue. Not necessarily as part of an AI Transformation, but actually digging in to who owns culture when you get down to the team level. Policies, values and organizational systems sit mostly with HR, but a team experiences culture largely through what its leader rewards, challenges and tolerates.
One HR leader mentioned that she’s struggled with culture across a few teams, and when speaking to leaders about them being responsible for the shift, she was met with:
“Culture isn’t my responsibility. I’m here to drive results, not make sure everyone is friends.”
And honestly, I can see why they may think that.
In most cases, we say we need leaders to collaborate, develop capability, create trust and lead differently through AI, but then we still predominantly judge them on what they deliver.
If an organization rewards them for delivery and output, then in my opinion, that leader is behaving rationally to the standards and accountability they’re being held to.
So my question would be, what are leaders in your organization actually accountable for? If their performance reviews, bonuses, promotion criteria and exec conversations are overwhelmingly about revenue, delivery, margin, deadlines and team output, why are we surprised when culture, collaboration and capability-building come second?
And that was before AI came into the equation. Now, AI is making this mismatch even more consequential because we’re expecting leaders to get teams experimenting, working across functions, challenging how work gets done, and having more open conversations that could lead to disagreements on what should be automated, as well as developing people’s judgment and helping teams navigate uncertainty.
In order to do that, you need a very particular environment. Not necessarily “everyone being friends”, but can someone challenge how the team works? Can they admit they don’t understand the tech? Can they experiment and get something wrong? Can functions disagree constructively? Can a junior person challenge a senior person’s assumptions about where AI should be used? Can people say an AI-generated answer or product is rubbish even when someone senior is excited about it?
Those things can affect whether AI transformation works, and so if a leader says “culture isn’t my job; results are”, then the response should be – what if culture is now part of how you get the results?
I think this is really interesting, because when we typically look at leadership capability, we focus on training and development, but don’t always measure the same things. Leaders go on programs and to workshops about collaboration, developing others, decision-making, leading change etc., but what happens when their performance review comes around? Are they actually accountable for these things? Not vaguely through engagement scores, but:
Did they build capability below themselves?
Did they develop successors?
Did decision-making improve?
Did they reduce dependency on themselves?
Can they work effectively across functions?
Did their team become more capable of operating through change?
Did they create an environment where experimentation can happen?
If the answer to most of these questions isn’t a resounding yes, then I think there’s a bigger discussion to be had about what we’re actually rewarding leaders for.
We can absolutely keep investing in leadership programs that teach collaboration, developing others etc., but if none of those truly count when performance and progression are decided, why would we expect them to compete with delivery?
Maybe the question isn’t just whether leaders have the capability we need from them now, but whether we’ve actually made them accountable for using it.
If a leader delivers the results but leaves behind a team that can’t operate without them, hasn’t built capability, and isn’t able to challenge, collaborate or experiment, are we still calling that high performance and rewarding as such?