Your last retrospective produced four action items. Three of them were dead by Thursday. Nobody decided to kill them. The sprint simply filled up, the way sprints do, and improvement went back to being something the team believes in rather than something it does.
That gap has a name. Carol Dweck, whose research created the growth mindset concept, spent years warning about what she calls the false growth mindset: organisations that declare the belief, print it on the wall, and change nothing about how work actually runs. In her follow-up work she was blunt that claiming the mindset is the easy part. The behaviour is the mindset.
"The passion for stretching yourself and sticking to it, even (or especially) when it's not going well, is the hallmark of the growth mindset."Carol Dweck · Mindset
Belief is a noun. A mindset is a verb.
A team has a growth mindset when improvement is normal work, not extra work. Not a value statement, not an offsite theme. A loop that runs every sprint: notice a signal, run a small experiment, look at the evidence, keep what worked. Then again. The loop is boring by design. Boring is what survives sprint pressure.
Watch what happens when the loop is missing. A retro surfaces a real problem, say the team keeps discovering blocked work two days before the sprint ends. Someone writes "improve blocker visibility" on a sticky note. It is too big to finish, has no owner, and no one is scheduled to ask about it again. That action item was dead the moment it was written. The team didn't lack belief. It lacked a loop with a next step small enough to take.
What the loop needs to run
An honest signal comes first, and honest signals are rare in rooms with hierarchy in them. This is why anonymous channels matter: across Performalise teams, Team Voice has surfaced more than 26,000 improvement ideas from people who had been sitting in retros for years saying very little. The appetite for growth was always there. The channel wasn't.
Then the experiment has to be small. Small enough to finish inside one sprint, owned by one name, checked at the next retro in front of everyone. Public follow-through is the difference between a loop and a wish. Teams that lose their retro actions almost never lose them to disagreement. They lose them to size and silence.
Loops compound. Beliefs don't.
Across our customer base, teams running this loop improve measured sprint outcomes by 4.2% sprint over sprint. That number looks modest until you let it run: at that rate a team roughly doubles in about seventeen sprints. Eight months, give or take. No transformation programme, no consultancy, no new vocabulary. Just a loop that never stopped turning, which is exactly what the Kaizen Engine exists to keep turning when attention drifts.
Here is the Monday morning version. Take the one retro action your team most recently let die. Shrink it until it fits inside a single week, give it one owner, and open your next retrospective by asking what happened to it. Do that for three sprints and you will not need a poster about growth mindset. You'll have the loop, and the loop is the mindset.