You had a coach. Your managers and high-potential ICs didn't. Here's what scaled AI coaching does for the layer running your org tomorrow.
.png)
Last updated: 2026-05-06
You had a coach early in your career. Or you have one now. Either way, you know what it did. The hard call you handled differently because someone helped you think it through. The decision you reversed before it shipped because the right question came at the right time.
You also know who got coaching and who didn't. The senior leaders did. The board members did. You probably did, eventually. Everyone below the line where coaching budgets stop has been figuring it out on their own. The ICF/HCI research is blunt about it: only 27% of organizations make professional coaching accessible to all employees. The other 73% are running on the cohort split you grew up inside.
Your managers and your high-potential ICs are working through the same moments you did, with none of the support you had. The leadership training they completed last year is months behind the conversation they're having on Tuesday. The 1:1 that decides whether someone stays, the feedback that changes a career trajectory, the team plan you keep meaning to help your manager build, those moments are happening with no coach in the room.
This is the case for AI coaching at the layer below you. Not as your replacement. As the multiplier of what worked for you, at the cohort that has never had access to it.
The bench you're counting on to carry the org through its next stage is largely uncoached. The middle managers who'll become your VPs. The high-potential ICs who'll become your managers. The team leads quietly figuring out whether to stay or go. They're getting their development from books, from peers, from a manager who is also uncoached, and increasingly from ChatGPT.
The gap shows up everywhere if you know where to look. Quiet attrition of people you wanted to keep. Manager promotions that don't quite work, but you can't say why. High-potential ICs who plateau without knowing they were on a path. Hard conversations that didn't land, then escalated, then turned into exits.
Gallup's 2026 State of the Global Workplace puts manager attribution at 70% of the variance in team-level engagement, on the back of 27M+ respondents across multiple editions. The same report shows global manager engagement has dropped to 27%, with the steepest declines among managers under 35. The variance is the cohort that never had a coach. The bench that doesn't build itself is the bench that no one is developing.
When you got coaching, the value was in your decisions, your conversations, your team. The same coaching at the manager layer of your org compounds in a way yours alone never could.
A single one of your VPs has roughly 50 to 200 conversations per quarter that materially affect retention, performance, or trajectory. With AI coaching available before each of those conversations, the variance in how those conversations land compresses. Multiply across the manager layer and the IC layer below, and you have a development engine running on every meaningful work moment in your org.
What scaled coaching produces is specific. The conversations you can't be in get the support they need, in the moment they need it. The org learns at the layer where the work actually happens.
Your team is using ChatGPT, Claude, Gemini, and Copilot for the questions they can't ask their manager or HR. Career questions, conflict questions, and the 11pm leadership second-guessing they can't bring to anyone. The conversations are happening, just not in spaces you'd want them happening.
On enterprise versions of those tools, those prompts are accessible to your IT admin via the standard compliance and audit paths these platforms expose by design. (We covered the architectural risk in detail in Why generic AI assistants aren't safe for employee coaching.) Your high-potential IC venting about the manager you're going to promote next quarter is on the record in your tenant. Your manager rehearsing how to handle a performance issue is on the record in your tenant. The exposure is structural, and it's there today.
Architectural privacy at the coaching layer is what lets you offer your managers and ICs somewhere to be honest. No compliance API. No admin override. The aggregate signal at the org level is real, and the individual conversations belong to the individuals having them. That trade is what makes scaled coaching deployable, and also what makes it work. Coaching only delivers when the person being coached can be honest.
You couldn't afford executive coaching for 200 people. You can't afford it for 50. The traditional model of human coaching was always going to stop above the manager line, which is exactly the line you most need to develop through.
The pricing comparison is what changed. Traditional human executive coaching is published at hundreds of dollars per hour, generally reaching five to ten people in a typical mid-market budget. Huckleberry is $20 per seat per month for teams. For roughly the cost of one human-coached executive, you can give every manager and IC in your org their own coach. The cost gap closes the access gap. The 95% becomes coverable.
The ROI proof for coaching at scale already exists. ICF/PwC's research shows 86% of organizations recoup their coaching investment, with an average return of 7x cost, and lifts to individual performance of 70% and team performance of 50%. Microsoft's coaching culture program is on record at 670% ROI. Intel's contributes around $1B a year in operating margin. The case for coaching has been made. What never existed was an economic model that let you offer it to the whole org.
Human coaching is excellent and not the layer to disrupt. AI coaching is how you reach the cohort the human-coaching model was never going to reach. The two layers operate in combination: senior leadership keeps its human coaches, and the rest of the org gets the same kind of structured support, available in the moments that matter, at a price that fits a normal L&D budget.
A year in, the org looks different.
The managers you would have worried about are not the managers you worry about. The high-potential ICs you wanted to keep are still here, growing, and saying so. The bench is no longer something to count on. You can see it developing.
You stop being the coaching multiplier of last resort. The development conversations are happening in your manager layer, not just escalating to you. Hard conversations are landing because they were prepped, not because someone got lucky. The org learns at the speed of its work, which is what every other operational system in your company has been doing for years and what people development never quite figured out until now.
The plan is shorter than most enterprise rollouts:
You can book a demo to walk through what activation looks like for an org your size, or start with a free 30-minute session to feel it for yourself before you bring it to your team.
Q: How is this different from the executive coaching I already use?
A: Human executive coaching is excellent for the senior leadership tier where the depth and price point fit. Huckleberry is built for the cohort below that line, the manager and IC layer that human coaching has never reached at scale. Many organizations using AI coaching at scale also retain human coaching for the senior leadership tier. The two layers operate in combination.
Q: What does HR or my admin actually see in the analytics?
A: Engagement metrics and high-level aggregate themes across the org. Session volume and engagement trends, plus broad coaching themes that emerge at the org level. Never individual transcripts, and never anything specific enough to identify what a particular manager or team is working on. The architecture forecloses individual access by design, which is what makes the aggregate signal trustworthy.
Q: Will this satisfy our privacy and compliance review?
A: Huckleberry publishes a Data Protection Addendum covering personal data handling, processing roles, and security commitments. The architectural privacy posture (no compliance API, no admin override to session content) is designed to address the objections that often block other AI tools at legal review. Audio is not stored, and customer conversation data is not used for model training.
Q: How quickly do we see impact?
A: Activation typically happens within days. Behavioral impact compounds over weeks. The retention and engagement numbers most orgs measure are quarterly, so a meaningful read on the metrics typically comes 60 to 90 days in. Anecdotal manager feedback (a manager saying "I handled this conversation completely differently because I prepped with the coach") usually starts in week one.
Q: What about international and multi-jurisdictional considerations?
A: Huckleberry's privacy architecture removes the underlying exposure under the ADA (the US Americans with Disabilities Act) and equivalent disability-disclosure frameworks in other jurisdictions, which is the question employment counsel typically raise on AI coaching tools. Data residency and regional deployment options are available for orgs with specific requirements. Your legal team can confirm specifics for your jurisdiction.
Q: How does this fit alongside our existing manager training program?
A: Training transfers concepts in a classroom setting. AI coaching applies them in the moment of the actual situation. They work in combination, and AI coaching is often what makes existing training finally translate into manager behavior, because the application happens at the point of need rather than three months after the workshop.
You know what coaching did for your career. The org you're building deserves the same kind of support, at the layer where the work happens and the careers that follow yours are being formed.
Book a demo to walk through what scaled AI coaching looks like for your org. Or try a free 30-minute session yourself, the way most exec champions do. $20 per seat per month for teams. Huckleberry launched publicly at Transform 2026 and is live now.
.png)
.png)
