This is a companion piece to our earlier post on how habit science can build a culture of continuous improvement. That post covers the foundational concepts -- the habit loop, BJ Fogg's behavior model, identity vs. goals, persona-based design, and how interlocking loops create self-sustaining systems. If you haven't read it, start there.

This post skips the theory and goes straight to application. For each of five core CI behaviors, we'll walk through how to design the cue, the routine, and the reward so the behavior sticks -- drawing on what practitioners at Electrolux, Union Pacific Railroad, and other organizations have learned through years of trial and error.

1. The Daily Huddle

The huddle is the single most important CI habit for frontline teams and their leaders. It's also the one most likely to die quietly. Huddles get canceled when the week gets busy. They drift into status updates where everyone reports to the leader instead of solving problems together. Within a few months, the board on the wall is wallpaper.

Cue: A calendar invite is a weak cue -- it's buried in a screen full of other invitations. The strongest huddle cues are physical and environmental. Sandro Casagrande, who leads the Electrolux Manufacturing System methodology globally, describes designing visual management boards so that the signals in the environment -- a red indicator, a deviation from standard -- become the triggers for specific leadership routines. The board itself is the prompt. When people pass it on the way to start their shift and see something that needs discussion, the huddle becomes the obvious next step rather than an interruption.

Time anchoring helps too. The huddle happens at the same time, in the same place, every day. No decision required about when or where. The consistency is what allows the brain to automate. Habit stacking -- tying the huddle to something people already do reliably, like clocking in or pouring their first coffee -- gives it an existing neural pathway to attach to.

Routine: Keep it short and structured enough that there's no ambiguity about what happens. Review yesterday's key metrics. Surface any safety or quality issues. Note where improvements are in progress. Identify one thing that needs attention today. Ten to fifteen minutes. If it regularly runs longer, the scope has crept and the habit is at risk -- because the harder a routine is, the more motivation it requires, and motivation fluctuates.

Lynn Kelley, who led organizational change across Textron and Union Pacific Railroad, makes a point that applies directly here: when you're designing the change, get the bugs out before you ask people to adopt it. If the huddle format doesn't work well, the people who resist change will point to the problems as proof the whole thing is a waste of time. Pilot the format with a willing team. Refine it. Then spread what works. Don't debug in production.

Reward: This is where most organizations fail the huddle. It ends and... nothing. Nobody feels better. No dopamine fires. The reward needs to be designed in. End every huddle with a specific acknowledgment -- an improvement someone made, a problem someone raised, a follow-up that was completed. A sentence of recognition from the leader, directed at a specific person for a specific contribution, takes 10 seconds and changes the emotional residue of the entire meeting.

When middle managers run huddles consistently, that consistency itself becomes a cue for frontline workers ("the huddle is coming, I should have something to share") and a signal to executives that the management system is running. One persona's routine becomes another persona's cue -- the interlocking loop doing its work.

2. Idea Submission

Getting frontline workers to submit improvement ideas is one of the highest-leverage habits in a CI program. It's also one of the most fragile, because the failure mode is invisible: people simply stop bothering, and nobody notices the silence.

In traditional suggestion box systems, only 2-3% of ideas are ever implemented. The rest vanish without feedback. That's not an engagement problem. It's a habit destruction machine. Every idea that disappears teaches the submitter that the behavior isn't worth repeating. The reward component of the loop isn't just missing -- it's actively negative.

Cue: The huddle is the natural trigger point. If "does anyone have an observation or improvement to share?" is a standing part of the huddle routine, the huddle does double duty as the cue for idea submission. Other effective cues: a visible prompt at the workstation (a card, a QR code, a screen); a recurring notification from the improvement system; or the simple habit of leaders asking "what's making your work harder than it needs to be?" during walks through the work area.

Social cues matter too. People imitate the habits of the people around them -- their close team, the broader organization, and especially leaders. If a manager submits their own improvement ideas visibly, it normalizes the behavior. If nobody in the immediate peer group participates, you're fighting gravity.

Routine: Friction kills participation. Every additional field on a submission form decreases the chance someone will use it. Start with the minimum: a sentence describing the problem or opportunity. That's it. You can enrich the data later, after the habit is established. The two-minute rule from habit science says a new habit should take less than two minutes at first. The goal isn't to capture detailed A3s on day one -- it's to build the neural pathway of "I see a problem, I say something about it."

At KaiNexus, new employees are required to submit at least one improvement to the onboarding process as they go through it. Not asked. Required. And it starts small -- Keith, a recent hire, noticed an inconsistency in LinkedIn profile standards and automated a manual notification workflow. The improvement wasn't grand. But the habit of "I see something, I fix something" was established on day one, in a low-risk setting, before the person had even absorbed the full culture.

Reward: The most powerful reward for idea submission is a fast, visible response. Not implementation -- response. An acknowledgment that says "we received this, here's who owns it, here's what happens next." Speed matters. A response within hours reinforces the behavior. A response three weeks later, after a committee reviewed a batch, might as well not exist -- the brain can't connect the reward to the action across that gap.

Lynn made the point with a stat that sticks: 66% of employees say they'd consider leaving a job where they don't feel appreciated. That's for their regular work. Now ask people to go beyond their regular work, to take the risk of suggesting a change, and give them nothing back. Of course they stop.

Recognition beats incentives here. A leader saying "that idea about the supply room -- tell me more" right after a huddle fires dopamine in a way that a gift card distributed at the end of the quarter can't touch, because the brain connects the specific behavior to the specific feedback. The gift card is too delayed and too disconnected to build a loop.

3. Gemba Walks

Leader gemba walks appear in every Lean playbook. They're also the leadership habit most likely to be crowded out, because they don't have the structural protection of a meeting invite and they compete with everything else in a leader's day that feels urgent.

Cue: The most reliable cue is a block on the calendar treated as non-negotiable -- not "walk the floor if I have time" but "gemba walk, Tuesday and Thursday, 9:00 AM." Some organizations go further and make the walk the first activity of the day, before email, before meetings. The physical act of walking to the work area becomes the transition cue.

Sandro describes how leader standard work at Electrolux serves as the cue system. If the documented routine says "audit area B on Wednesday," the standard work is the prompt. And critically, Electrolux evolved their leader standard work to include a reflection step -- after each routine, leaders note whether their behavior during the session was effective and what they'd change next time. The standard work isn't just a checklist. It's a PDCA cycle applied to leadership habits themselves.

Routine: Go with a focus. Observe before asking. Ask before telling. And produce a specific output -- a note about what was observed, a follow-up item assigned, a coaching conversation documented. Without an output, the walk becomes a social visit and the habit loop has no anchor. A walk that generates nothing feels like it accomplished nothing, which undermines the reward.

A practice from the Electrolux approach: when doing an area audit, put yourself in the shoes of the person leading that team. Ask yourself -- if I were running this area, what would I need to know to manage well? Do visual aids make that information available simply? This reframe turns the gemba walk from an inspection into a service act. You're not checking up on people. You're evaluating whether the system gives them what they need to succeed.

Reward: For the leader, the reward is seeing something they wouldn't have seen from their desk. For the people on the floor, the leader's presence is itself a signal -- but only if the leader does something useful. Karyn Ross, drawing on Taiichi Ohno, describes the test: if it takes you hours to walk 100 meters because people are stopping you with problems and ideas, you're doing it right. If you walk through quickly, nobody sees you as helpful. The reward for the frontline is a leader who makes their work easier, not one who observes and leaves.

The follow-up completes the loop. If a leader learns about a problem during a gemba walk and nothing changes, they've taught people that speaking up is pointless. Track what surfaces during walks and close the loop. That follow-through is the reward that makes people bring problems to you next time.

4. Coaching Conversations

Coaching is the habit that multiplies every other habit. A leader who coaches develops problem-solvers. A leader who directs develops people who wait for direction. The difference compounds over years.

Cue: The Toyota Kata framework, which Electrolux adopted as a cornerstone of their evolved system, anchors coaching to a physical location -- the learner's improvement board or storyboard. The board is the cue. You don't schedule "coaching" as an abstract calendar event. You walk to the board and the conversation begins.

The cadence is daily. Not weekly, not "when there's a problem." Daily repetition is what builds the neural pathway for both coach and learner. Experienced pairs complete a full coaching cycle in under 10 minutes. That brevity matters -- it keeps the routine below the friction threshold where it starts requiring heroic motivation to maintain.

Routine: The Kata coaching questions provide the structure: What is your target condition? What is the actual condition now? What obstacles are you addressing? What was your last step and what did you learn? What is your next step? The questions are the same every time. The learner knows what the coach will ask. There are no surprises. But answering well requires genuine thinking, which is the development mechanism.

Sandro adds a layer that matters for sustainment: the coach doesn't stand in front of the learner, leading the way. The coach stands behind. The learner is in front, running the experiment. The coach observes and guides from behind -- "not because we want followers," as Sandro puts it, but because the long-term objective is building people who own the improvement of their processes. That positioning isn't symbolic. It changes the dynamic of every conversation.

Electrolux also applies coaching beyond the improvement kata. They coach strategy deployment (challenging whether the business plan is still the right one, whether the team has the skills to execute it), and they coach conditions (evaluating whether visual management in an area gives a team leader what they need to run their business). Each of these coaching contexts has its own cue built into the leader standard work cadence.

Reward: For the learner, the reward is visible progress toward the target condition, and the growing confidence that comes from practicing scientific thinking daily. For the coach, it's watching capability develop -- seeing a person who needed heavy guidance six months ago now running structured experiments independently. Sandro describes celebrating even failed experiments, because the team learned something about the process. The test failing but the team growing is a success worth recognizing.

Feedback research that Sandro shared from Aubrey Daniels adds precision: the most powerful feedback is positive, immediate, and certain -- meaning it happens every time, not intermittently. Leaders who deliver positive recognition four times more often than corrective feedback gradually shift their teams from compliance ("have to") to ownership ("want to"). Most leaders underdeliver on positive feedback because it doesn't feel as urgent as correcting problems. But the habit science is clear: positive reinforcement builds the loop. Correction alone doesn't.

5. Recognition

Recognition isn't usually listed alongside huddles and gemba walks as a "CI behavior." It should be. It's the element that fuels every other loop, and it's the one most often left to chance rather than designed as a deliberate practice.

Cue: For leaders, the cue for recognition should be embedded in existing routines -- not treated as a separate activity that requires its own motivation. At the end of every huddle, acknowledge one specific contribution. During every gemba walk, thank one person for something specific. When reviewing improvement status in the system, comment on at least one item. The existing routines become the trigger for the recognition behavior.

For executives, the designed cue might be a 15-minute calendar block -- say, Thursday at 4:00 PM -- to log into the improvement system and comment on three or four items. That single routine, executed weekly, sends recognition signals that cascade through every layer of the organization.

Routine: Specificity matters more than scale. "Great job, team" is wallpaper. "Maria, the way you redesigned the supply cart layout cut 12 minutes off the morning setup -- that's going to help everyone on second shift too" is a reward that reinforces a specific behavior and signals to everyone else what kinds of contributions get noticed.

Lynn made the case for planning recognition ahead of time rather than waiting until there's a critical mass of successes. Why wait? Start recognizing the first idea submitted, the first huddle where someone raised a problem, the first executive who showed up at the gemba for the third week running. Early recognition creates the momentum that makes later recognition feel natural instead of performative.

Reward: The reward for the recognizer is seeing the impact -- increased participation, people who light up when their work is acknowledged, a visible uptick in idea volume after a round of recognition. This creates a secondary habit loop: the leader recognizes contributions, sees the positive effect, and is more likely to recognize again. The loop feeds itself. But it needs to start. And it needs to start deliberately, not whenever someone happens to feel appreciative.

The System as Habit Infrastructure

When you look at these five behaviors together, a pattern emerges: the organizations that sustain them longest are the ones where the system -- not individual memory or willpower -- delivers the cues and rewards.

A notification that an idea needs review is a cue that doesn't depend on a manager remembering to check a spreadsheet. A workflow that routes an idea to the right person and sends the submitter a status update is a reward mechanism that works whether or not the manager remembered to say thank you in that moment. A dashboard that shows an executive how many improvements were completed this month is a cue to engage, and the engagement streak badge that follows is the reward.

This is why purpose-built improvement infrastructure matters to habit sustainment specifically. Spreadsheets can't send timely cues. Email threads can't make status visible. Shared drives can't deliver the instant feedback that reinforces participation. The technology doesn't replace the human behaviors -- the coaching, the gemba walks, the recognition at huddles still require a person showing up. But it creates the structural backbone that makes those behaviors triggerable, trackable, and rewarding at a scale no manual system can match.

See KaiNexus in action ->

The goal isn't to automate improvement. It's to automate the scaffolding -- the cues, the routing, the visibility, the feedback loops -- so that human energy goes where it matters most: solving problems, coaching people, and making the work better.

Related KaiNexus webinars referenced in this post:

• Unlock the Power of Leadership: The Electrolux Manufacturing System Way with Sandro Casagrande
• No More "Flavor of the Month" Change: How to Deliver Sustainable Improvement with Lynn Kelley
• Why Kindness Is the Key to Lean Leadership with Karyn Ross
• How to Sustain Continuous Improvement with Habit Science (Part 3) with Greg Jacobson and Morgan Wright