Fifty-two percent of organizations named strengthening manager coaching and feedback skills their single highest priority this year. More than three times any other answer.
Fifty-six percent rate their managers at 5 out of 10 or lower for that specific skill.
Same dataset. Same organizations. That gap has been obvious for years, and L&D keeps answering it the same way. Better content. Tighter facilitation. A coaching model with a memorable name and a nice one-pager.
The content was never the thing holding it up.
Above is the 2026 Mid-Year Workplace Trends Report from LifeLabs Learning on how workplace complexity is reshaping manager effectiveness, focus, collaboration, and AI adoption. Below are my takeaways on what that means for how you design, sell, and reinforce learning.
TL;DR
- Quiet, chat-heavy workshops are a warning sign, not a facilitation problem.
- Manager coaching skill needs weekly reps, and most manager calendars have no room for them.
- Training and delegation both fail at follow-through.
- Focus training blames individuals for a problem the operating model created.
- The biggest AI capability gap is organizational context, not prompting skill.
1. A quiet workshop is a warning sign, not a facilitation problem
You can feel it in the first fifteen minutes. Breakout rooms that stay silent until someone volunteers to be the note-taker. Chat filling up while the mics stay muted. Three people carrying a session built for twenty.
LifeLabs Learning facilitators saw the same thing across roughly 400 workshops in the first half of 2026. Compared with previous years, they observed more hesitancy to contribute, more passive participation, and a clear drift toward chat over speaking up. KJ McConnell, one of their facilitators, thinks the reason is simple. "People don't take risks with those they don't trust."
Most L&D teams read a quiet room as a facilitation issue and go looking for better icebreakers.
It's a forecast.
Speaking up in a workshop costs almost nothing. No stakes, no audience beyond your peers, nobody remembers a half-formed thought by lunch. If someone won't take that risk in a room built to be safe, they won't try a new feedback technique on a direct report Tuesday afternoon, when the stakes are real and the person can get upset.
Your evaluation form will never catch this. Smile sheets measure whether people liked the experience, and quiet rooms produce perfectly pleasant ratings. A session can score 4.6 and change nothing.
The facilitators already know which rooms went dark. Nobody asks them, and nobody writes it down.
Ways to turn facilitator observations into real signals:
- Capture participation quality by adding three fields to your post-session facilitator debrief, covering how many people spoke unprompted, whether chat outpaced voice, and which segment of the room went quiet. Track it by cohort and department so you can see which teams are carrying the most pressure.
- Test psychological safety before the session through a two-question pre-work pulse asking how comfortable people feel admitting a mistake in front of their team. A low score tells you to build in reflection time before you introduce anything they have to practice.
- Rescue silent rooms in the moment by using paired practice instead of full-group discussion. Two people rehearsing a script is a far smaller risk than one person performing for twenty.
- Flag quiet cohorts to the sponsoring leader through a short note after the session, naming what you observed without diagnosing it. The conversation that follows is usually worth more than the workshop was.
2. Manager coaching is blocked by the calendar, not the curriculum
Manager coaching has been L&D's biggest line item for a decade. It's also the capability organizations rate lowest.
A senior director in one of LifeLabs' May coaching workshops named the reason: the business moves too fast, so managers skip the questions and jump straight into problem-solving.
Nobody forgets how to ask "what have you tried so far?" They skip it because the question costs eleven extra minutes and the answer creates work. Solving it yourself takes four minutes and ends the meeting.
Coaching is a skill that decays without reps, like a language. A manager leaves your workshop genuinely able to run a coaching conversation. Then she gets four straight weeks of back-to-back days, uses none of it, and by week six the model is a PDF in her downloads folder. Nothing about the curriculum caused that.
A senior leader in their April delegation workshop said the quiet part out loud. They're surviving, and there's so little time that delegating anything feels impossible. Which closes the loop. Managers can't hand work off, so the calendar never opens, so the coaching never happens, so the capability rating stays at 5 out of 10 and next year someone buys another workshop.
Selling coaching training into a calendar with no coaching time in it is how L&D spends a real budget and produces nothing. The workshop is fine. The environment eats it.
Steps to make the calendar part of the deal:
- Make protected time a condition of enrollment by requiring each participating manager to block two 30-minute coaching slots per month before they get a seat. If the sponsor won't agree to that, you have learned something useful about whether this program can work.
- Shrink the rep through teaching one question that fits inside an existing 1-on-1 rather than a full conversation model that needs its own meeting. "What have you tried so far?" costs 30 seconds and changes the shape of the next ten minutes.
- Audit real calendars before you design anything by asking three volunteers to screenshot a typical week. Design for the gaps that exist instead of the ones you hope exist.
- Report the calendar constraint upward through your program readout so the capability rating and the meeting load land on the same slide. Executives fund what they can see next to a number.
3. Follow-through is where both delegation and training fail
Seventy-two percent named monitoring and follow-through as the hardest part of delegation. Goal-setting and empowerment weren't close.
Fifteen pages later, the same report covers training transfer research from Burke and Hutchins showing that learning alone rarely changes behavior. Whether a skill sticks depends on practice opportunities, manager support, reinforcement, and an environment where experimenting feels safe.
Those are the same finding wearing different clothes. Handing the work over is easy. Staying with it is where everything falls apart.
L&D has a delegation problem and doesn't call it that. You hand a skill to a manager the way a manager hands a project to a direct report. Clear brief, good materials, real enthusiasm in the room, sincere intentions on both sides. Then you both go back to your regular jobs and hope.
Delegation data already told us how that ends.
Picture the manager six weeks after your feedback workshop. She remembers the model. She hasn't used it, because the one conversation that needed it happened on a Thursday when she was underwater, and she handled it the way she always has. Nobody was there in week two.
Reinforcement is expensive, unglamorous, and invisible in a budget review. A better deck photographs well. A check-in call in week three does not. That is most of why the follow-through stage keeps losing the argument internally.
Tactics for holding the follow-through stage:
- Budget reinforcement as a line item by carving 20% of every program's cost out of the delivery day and into the eight weeks after it. A shorter workshop with real follow-up beats a longer one that ends at the door.
- Give participants' managers a two-minute script through a single email naming exactly what their report practiced and what to ask about it in the next 1-on-1. Manager support is the strongest transfer variable and the one L&D most often leaves to chance.
- Schedule the second rep before people leave the room by using a calendar hold for a 20-minute practice check-in three weeks out. Skills die in week two, so put something in week two.
- Measure at 60 days instead of at the door through one behavioral question sent separately to the participant and their manager. Matching answers tell you it transferred, and mismatched ones tell you exactly where to look.
4. Focus training blames the individual for a systemic problem
Focus was the number one productivity complaint in the report at 41%, ahead of prioritization at 32% and staying organized at 20%.
The data underneath those numbers point somewhere else entirely. People weren't confused about what mattered. They couldn't find an uninterrupted hour to do it. One director summed up the whole environment in eight words. "When everything is a fire, it's a hot mess."
Katya Davydova, facilitating their May productivity session, saw managers conclude that protecting focus has to start from within. Don't check Slack first thing. Don't let meetings run over. Model time integrity so your team can copy it.
All true, and also the reason most focus training fails.
Individual habits are the only lever a 90-minute workshop can pull. The person who blows up your calendar isn't in the room. So you teach twelve people to defend their attention against an organization that hasn't agreed to stop attacking it, everyone leaves feeling briefly capable, and by Thursday the calendar has healed over.
Microsoft's own workday telemetry found the heaviest-hit fifth of employees taking a ping every two minutes across core hours, which works out to 275 interruptions a day. That is not a population you train your way out of.
The report's own conclusion gets this right, calling for attention to be treated as a shared organizational resource rather than a personal willpower project. A founder in their February webinar went further, describing fatigue as a systematic mistake and cultural debt, and arguing that executives would fix the traffic jams if they had the skills, incentives, and appetite to do it.
Which is where it gets awkward for us. When a request comes in for resilience training or time management, an operating model is asking you to help it avoid having a conversation. Running the workshop makes you part of the deferral.
Ways to shift focus from willpower to system:
- Reframe the intake conversation by asking the requesting leader what percentage of their team's calendar is currently unbooked. If the answer is under 20%, name that as the problem and offer to facilitate the norms conversation instead of the workshop.
- Teach urgency norms to the whole team at once through a session that produces a written response-time agreement rather than individual planning habits. Attention rules only hold if everyone who could interrupt is in the room when they get made.
- Design for interruption by building sessions and follow-up assets that survive being paused. Assume the learner gets pulled away twice and make re-entry easy.
- Name the AI wait as a focus risk using a team norm about what to do during the 30 seconds a tool spends generating something. Participants in these sessions described picking up something else and losing the thread completely.
5. Half of what looks like team conflict is a capability gap
A manager books time with you. Two people on the team can't work together; there's real tension, and maybe you could run something on communication styles.
You've had this conversation. Possibly this month.
In the LifeLabs data, 48% pointed to trust and capability concerns as the underlying cause of workplace conflict. That beat listening problems, personality differences, power dynamics, and unresolved tension. A Director of Talent and DEI at their May event described the pattern precisely, wondering whether the work had simply become complex enough that there was a skill gap in play, with a trust issue layered on top of it.
Half the time, one of those two people is underwater in work nobody trained them for, and the other has quietly stopped trusting them to deliver. A communication workshop produces a pleasant afternoon and changes nothing, because the tension is downstream of a capability gap nobody has been willing to say out loud.
The other common version is structural. Twenty-eight percent described collaboration breakdowns as ownership and accountability confusion, and one manager admitted to being months into a role and still unsure who was responsible for what. No amount of skilled listening resolves an ownership question. That's an org design conversation wearing a training request's clothes.
L&D is the only function in the building with real vocabulary for capability. We're also the function that takes conflict briefs at face value and starts building. Nobody asks the one question that would sort it: what specifically can this person not do yet?
Worth knowing before you ask it: naming a capability gap is politically loaded. Say it wrong, and a manager hears an accusation about a person, or the person hears it and starts job hunting. Frame it as the company failing to equip someone to do it, and it lands as support.
Steps to diagnose before you design:
- Separate skill from will by asking the requesting manager whether this person has ever done the task successfully before and what support they got. If the answer involves no training and no example, you have a capability gap dressed up as an attitude problem.
- Map ownership before you book a room through a 15-minute exercise where each person writes down what they believe they own and what they believe the other person owns. The mismatch is usually the entire conflict.
- Refuse the wrong brief by offering a two-week diagnostic in place of the requested workshop. Sponsors accept this more often than L&D teams expect, because they mostly want the problem gone.
- Frame gaps as organizational failures using language about what the company never equipped someone to do. People defend themselves against skill accusations and engage with offers of support.
6. AI enablement stalls without written organizational context
Fifty-eight percent described AI adoption at their organization as improvised. Fourteen percent called it thoughtful and coordinated.
When the same audience was asked what's hardest about leading a team through AI adoption, the answers split almost perfectly evenly. Leading through change took 21%. Keeping trust high, defining what humans own versus AI, and improving productivity each took 20%. Building critical thinking took 19%.
A flat distribution like that is the finding. There is no dominant pain point, which means there is no single course to buy. The problem is that a one-size-fits-all AI curriculum cannot help.
The most useful thing in the whole report is a story a global shared services director told at their May session. Their company rolled out Claude with an encouraging note to jump in and learn together. He connected his calendar, email, and Slack, then asked it about his work and his relationships. His word for the output was "UNHINGED." It told him his boss hated him, said he should be earning $400K in a role well beyond his experience, suggested he either confront the CEO or start job hunting, and then began drafting his résumé.
The team then gave the tool actual company information, values, and org structure. Same questions. The answers turned out useful.
The variable was context, not the model and not the prompt.
That puts most AI enablement programs in an awkward position. Prompt training teaches people to ask sharper questions of a system that doesn't know where they work, who decides what, or what the company claims to value. Sharpening the question does not fix a missing world.
Which quietly turns AI enablement into a knowledge management project, and almost nobody budgeted for that. The context that rescued the answers in that story lives, in most organizations, inside four people's heads and a Slack channel from 2023.
Ways to build the context layer before the prompt layer:
- Audit what your tools have access to by listing the five questions employees most want to ask an AI assistant and checking whether the answer exists anywhere in writing. The gaps become your first content project.
- Convert tribal knowledge into retrievable documents through short capture sessions with the people everyone already asks. Two hours with the person who knows how decisions really get made is worth more than a prompting course.
- Teach context-setting as the core skill using a repeatable habit of stating role, goal, constraints, and audience before any request. It generalizes across every tool and survives the next model release.
- Define where AI is expected, optional, and off-limits by publishing one page per function. Ambiguity is what turned curiosity into fear, role confusion, and what one director called AI brain fry that facilitators saw all spring.
7. Every session splits its minutes between processing and practice
Two findings in the report's learning section pull in opposite directions, and the report presents them as compatible.
Participants wanted immediate usefulness. The new bar, in their words, was "Can I use this tomorrow?"
Facilitators saw something different in the same rooms. People walked in already carrying change, uncertainty, and competing priorities, and they couldn't absorb anything new until they'd had room to process what they were already holding.
LifeLabs concludes that the strongest organizations design for both at once.
Both at once is a lovely sentence and a difficult 90 minutes.
Processing time and practice time are covered by the same budget. Every ten minutes you give people to talk about what they're carrying is ten minutes not spent rehearsing the thing you were hired to teach. And the person who approved the session bought the thing you were hired to teach.
Split the difference and you get the worst version of each. Too rushed for anyone to decompress, too shallow for anyone to build a skill. Everyone leaves with a nice feeling and a half-remembered framework.
The sequencing solves more of this than the ratio does. Processing goes first, always, because the report's own psychological safety finding says people won't practice until contributing feels low-risk. The reflection time is the precondition for the practice block working at all, which is also the argument you make to the sponsor. Cut content, not connection.
This does break down in places. A room of cross-functional strangers learning a discrete tool doesn't need a safe harbor. Neither does anything compliance-shaped. The need scales with how much the group is carrying together.
Tactics for spending session minutes on purpose:
- Cut content rather than connection by removing one full module from a 90-minute design and giving the first 15 minutes to structured processing. Two techniques people practice beat five techniques they watch.
- Sequence processing first through a short paired exchange about what's hardest right now, before you introduce any framework. People who have said one true thing out loud will practice, and people who haven't will spectate.
- Tell the sponsor what you traded by naming the module you cut and the reason in your program summary. This protects you when someone asks why the agenda got shorter.
- Match design to the group's history by using more processing time with intact teams under pressure and less with assembled groups learning a single tool. Read the room's load before you set the ratio.
The through line across all four trends is that organizations aren't struggling with the work. They're struggling with the noise around it.
That's a hard sell for L&D, because noise isn't a curriculum problem, and most of what generates it sits well above the training budget.
It's also the opening. A function that can sit in an intake meeting, say, "this isn't a training need," and then name what is becomes something more valuable than a content vendor. Start with the next request that lands on your desk. Ask one more question than you normally would before you agree to build anything.
