Deep Dive: Coaching, Insight, and AI: Building Change That Lasts
A leader leaves a coaching session holding something true: "When the pressure's on, I soften the message to keep the peace." It lands cleanly - honest, useful feedback. Three days later a hard 1:1 lands on an overloaded calendar, and they catch themselves cushioning the same difficult message all over again.
So where did the insight go?
This is the problem at the centre of coaching: the distance between what we understand in the room and what we can actually reach for when the day gets messy.
It helps to know where the idea of forgetting comes from. More than a century ago, Hermann Ebbinghaus measured how quickly freshly learned material faded from his own memory, and gave us the now-famous "forgetting curve." A modern replication found broadly similar patterns of decay across intervals from twenty minutes to thirty-one days - while also showing that real forgetting is far less tidy than the popular summaries pretend (Murre & Dros, 2015). The workplace lesson survives the mess intact: learning left alone fades, and learning that gets retrieved, applied, and reinforced has a far better chance of sticking.
Picture a coaching insight as a compass in a desk drawer. It can point you toward a better route - but only if you pick it up before you set off. In leadership, the route is a feedback conversation, a hiring call, a tense team meeting, the exact moment an old habit reaches for the wheel.
Coaching already manufactures the hard part: insight, reflection, motivation, a sense of accountability. What learning science adds is the scaffolding those ingredients need to last - spacing, retrieval, deliberate practice, feedback, and cues that arrive on time.
And that is precisely where AI, used with care, stops being a convenience and becomes something more useful: the reinforcement layer that lives between coaching sessions.
TALY is more than just personality. It’s an always-on talent intelligence platform designed to create lasting and effective behaviour change, powered by AskTALY.
Want to see how leaders are using it as a coaching alternative, saving time, and getting to real changes quicker?
Get in touch or book a demo today.
Why good insights fade
Most coaching insight doesn't fade because the session was weak. It fades because memory, attention, stress, and habit all tug us back toward the familiar.
Ebbinghaus tends to get flattened into slogans - "people forget most of what they learn within a day." Tidy, but it claims more than the evidence can carry about real behaviour at work. The sturdier version is the one worth keeping: leave learning alone and recall drops; revisit and use it and retention climbs (Murre & Dros, 2015).
That matters here because coaching insight is rarely a dry fact. It arrives wrapped in emotion and context. In the session the client has room to think - to slow down, name a pattern, hear themselves clearly, trace a behaviour to its impact. Back at the desk the conditions invert. They're interrupted, rushed, juggling risk and relationships and deadlines all at once.
So what slips away isn't only the content. It's the emotional charge of the realisation, the alternative they meant to try, the cue that was supposed to flag the moment, the line in their profile describing how they default under pressure. That isn't a character flaw; it's how the equipment is built.
Spacing: one big realisation is rarely enough
Learning science, put simply, is the scientific study of the factors that make learning stick.
One of the most durable findings in learning science is the spacing effect - we retain more when practice is spread across time rather than crammed into a single burst. The Cepeda team's meta-analysis pooled 839 assessments from 317 experiments across 184 articles and confirmed that the gaps between study and the moment you need to recall genuinely shape long-term retention (Cepeda et al., 2006).
The coaching parallel is hard to miss. A great debrief can produce a powerful "aha," but a single moment - however vivid - is still concentrated learning. Helpful, but rarely enough on its own.
Spaced coaching asks a different set of questions. When should this client revisit the insight? When will the behaviour actually be tested? What prompt could help them retrieve it before the moment, and what reflection could help them learn after it? The growth edge for the profession isn't more sessions, longer reports, or fatter content libraries. It's better reinforcement around the moments that matter.
Retrieval: putting insight into action beats rereading it
A second finding lands just as squarely: retrieval practice. Actively pulling information out of memory does more for long-term retention than studying it again. Roediger and Karpicke showed that testing improved delayed recall even when restudying produced more confidence in the short run (Roediger & Karpicke, 2006), and follow-up work established repeated retrieval as central to consolidating what we learn (Karpicke & Roediger, 2008).
For coaching, the implication is concrete: a client needs more than access to their notes, report, or profile. They need to surface the right insight at the right time.
Before giving feedback: "What does my profile say I tend to avoid?" Before a team meeting: "Whose working style do I need to flex toward today?" After a clash: "What pattern showed up in me, and what would I try differently?"
Each of those is active recall, and each turns a stored insight into a leadership move. It shifts the client from "I remember what we discussed" to "I can use it when it counts." When Dunlosky and colleagues rated ten common study techniques, practice testing and distributed practice came out as high-utility because they hold up across learners, tasks, and settings; rereading and highlighting scored near the bottom (Dunlosky et al., 2013). Looking back over notes feels productive. Being asked to use the insight usually does more.
Behaviour change: awareness is only the opening move
Coaching is generously stocked with self-awareness, and that's valuable. But changing behaviour asks for more than knowing.
The behaviour-change literature offers a clean frame. Michie and colleagues' COM-B model holds that behaviour rests on three things at once: capability, opportunity, and motivation (Michie et al., 2011). A leader needs to be able to act differently, to have real situations in which to practise, and to keep choosing the new move when the old one feels easier. That trio explains why insights stall. A client can know they should be more direct yet never find a safe moment to rehearse it. They can want to delegate while working inside a culture that quietly rewards control. They can understand their derailers perfectly and still miss the cue that says this is the moment to pause.
Two more ideas help close the gap.
Goal-setting research is very consistent here: Locke and Latham's thirty-five-year programme found that specific, suitably stretching goals reliably outperform vague encouragement to "do your best" (Locke & Latham, 2002) - which is why "communicate better" rarely changes anything, while "name the expectation in one clear sentence" might.
Gollwitzer's implementation intentions then attach that goal to a trigger: a simple if-then plan links a situational cue to a chosen response and measurably improves follow-through (Gollwitzer, 1999). In coaching language: "If I notice myself softening the message, I'll state the expectation plainly, in one sentence."
And habit explains the relapse. Lally and colleagues found that repeating a behaviour in a consistent context makes it more automatic over time - though how fast that happens varies enormously between people and behaviours (Lally et al., 2010). When a leader snaps back under pressure, it's usually because the old behaviour simply has more reps on the clock. The remedy isn't a deeper insight. It's more practice.
Coaching works, and it can work even better
The evidence for coaching itself is genuinely encouraging. Theeboom and colleagues' meta-analysis found positive effects across performance, wellbeing, coping, work attitudes, and goal-directed self-regulation (Theeboom et al., 2014); Jones, Woods, and Guillaume found the same across learning and performance outcomes (Jones et al., 2016). A more recent meta-analysis from Cannon-Bowers and colleagues again confirmed coaching is effective overall, while calling for the science behind it to keep maturing (Cannon-Bowers et al., 2023).
That last point is the one TALY cares about: coaching already works, and the next move is to design it more intelligently around how people actually learn, forget, practise, and change.
The training-transfer literature sharpens the target. Baldwin and Ford framed transfer as the generalisation and maintenance of new skills on the job - shaped by the learner, the design of the learning, and the work environment (Baldwin & Ford, 1988). Blume and colleagues' meta-analysis later showed transfer hinges on trainee characteristics, motivation, environment, and support, especially for "open" skills (Blume et al., 2010).
Leadership is an open skill - and so are feedback, collaboration, self-regulation, trust, and conflict. They're messy, social, and shaped by context, which is exactly where coaching earns its keep. The work was never just grasping the concept. It's carrying the concept into the next conversation.
Where AskTALY fits: the retention layer coaching has been missing
Emerging research on AI coaching is still young, so we should be clear and responsible. A systematic literature review by Passmore, Olafsson, and Tee found that AI coaching research is growing, with early evidence suggesting AI coaches can be useful, accepted, and effective for specific tasks, and can aid the overall coaching experience (Passmore et al., 2025). Terblanche and colleagues also found that an AI chatbot coach supported goal attainment, while noting that human empathy and emotional intelligence remain hard to replicate and best left for live coaching sessions (Terblanche et al., 2022).
AskTALY improves the conditions that help coaching outcomes happen.
It does that along the grain of the research:
Spacing - instead of waiting three weeks to revisit an insight, a client can return to it before a meeting, after a hard moment, or in a weekly reflection.
Retrieval - rather than passively rereading a report, they're prompted to pull the relevant insight into a live situation.
Transfer - it connects personality, emotional intelligence, risk, and motivational data to real workplace moments: feedback, conflict, onboarding, delegation, team alignment.
Action - it helps turn a broad intention into a small, testable if-then experiment.
Personalisation - different people forget different things, dodge different conversations, overplay different strengths, and need different cues. A static report treats the profile as finished; AskTALY treats it as a living playbook.
There's a useful academic anchor for the whole idea. Work on just-in-time adaptive interventions in digital health describes delivering the right type and amount of support at the moment a person is both receptive and in need, adjusting to their changing context (Nahum-Shani et al., 2018). Strip away the clinical setting, and that is the design brief for a coaching companion: the right nudge, in the right form, at the moment it can actually be used.
The future of coaching is with AI, not instead of
The science of learning doesn't say coaching is failing. It says coaching has more room to grow because we routinely ask a single session to deliver insight, motivation, recall, transfer, and durable behaviour change across weeks of pressure. That is an enormous ask for one hour.
The next era of coaching will trade bigger reports for better continuity, one-off "aha" moments for the ability to retrieve the right insight at the right time, and static assessment outputs for living, personalised support that helps leaders rehearse the person they're trying to become. AskTALY is a practical way to make that shift. It keeps the profile alive, turns insight into prompts, and helps people move from understanding their patterns to testing new ones in the work itself.
Open your TALY profile, find one place this shows up, and pick a single micro-shift to test this week.
Get in touch to find out more… we really do love talking about this stuff.
Or Book a Demo today to see how easy it is to start using TALY in your business.
References
Baldwin, T. T., & Ford, J. K. (1988). Transfer of training: A review and directions for future research. Personnel Psychology, 41(1), 63–105. https://doi.org/10.1111/j.1744-6570.1988.tb00632.xBlume, B. D., Ford, J. K., Baldwin, T. T., & Huang, J. L. (2010). Transfer of training: A meta-analytic review. Journal of Management, 36(4), 1065–1105. https://doi.org/10.1177/0149206309352880Cannon-Bowers, J. A., Bowers, C. A., Carlson, C. E., Doherty, S. L., Evans, J., & Hall, J. (2023). Workplace coaching: A meta-analysis and recommendations for advancing the science of coaching. Frontiers in Psychology, 14, Article 1204166. https://doi.org/10.3389/fpsyg.2023.1204166Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2006). Distributed practice in verbal recall tasks: A review and quantitative synthesis. Psychological Bulletin, 132(3), 354–380. https://doi.org/10.1037/0033-2909.132.3.354Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students' learning with effective learning techniques: Promising directions from cognitive and educational psychology. Psychological Science in the Public Interest, 14(1), 4–58. https://doi.org/10.1177/1529100612453266Gollwitzer, P. M. (1999). Implementation intentions: Strong effects of simple plans. American Psychologist, 54(7), 493–503. https://doi.org/10.1037/0003-066X.54.7.493Jones, R. J., Woods, S. A., & Guillaume, Y. R. F. (2016). The effectiveness of workplace coaching: A meta-analysis of learning and performance outcomes from coaching. Journal of Occupational and Organizational Psychology, 89(2), 249–277. https://doi.org/10.1111/joop.12119Karpicke, J. D., & Roediger, H. L., III. (2008). The critical importance of retrieval for learning. Science, 319(5865), 966–968. https://doi.org/10.1126/science.1152408Lally, P., van Jaarsveld, C. H. M., Potts, H. W. W., & Wardle, J. (2010). How are habits formed: Modelling habit formation in the real world. European Journal of Social Psychology, 40(6), 998–1009. https://doi.org/10.1002/ejsp.674Locke, E. A., & Latham, G. P. (2002). Building a practically useful theory of goal setting and task motivation: A 35-year odyssey. American Psychologist, 57(9), 705–717. https://doi.org/10.1037/0003-066X.57.9.705Michie, S., van Stralen, M. M., & West, R. (2011). The behaviour change wheel: A new method for characterising and designing behaviour change interventions. Implementation Science, 6, Article 42. https://doi.org/10.1186/1748-5908-6-42Murre, J. M. J., & Dros, J. (2015). Replication and analysis of Ebbinghaus' forgetting curve. PLOS ONE, 10(7), Article e0120644. https://doi.org/10.1371/journal.pone.0120644Nahum-Shani, I., Smith, S. N., Spring, B. J., Collins, L. M., Witkiewitz, K., Tewari, A., & Murphy, S. A. (2018). Just-in-time adaptive interventions in mobile health: Key components and design principles for ongoing health behavior support. Annals of Behavioral Medicine, 52(6), 446–462. https://doi.org/10.1007/s12160-016-9830-8Passmore, J., Olafsson, B., & Tee, D. (2025). A systematic literature review of artificial intelligence in coaching: Insights for future research and product development. Journal of Work-Applied Management. https://doi.org/10.1108/JWAM-11-2024-0164Roediger, H. L., III, & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249–255. https://doi.org/10.1111/j.1467-9280.2006.01693.xTerblanche, N., Molyn, J., de Haan, E., & Nilsson, V. O. (2022). Comparing artificial intelligence and human coaching goal attainment efficacy. PLOS ONE, 17(6), Article e0270255. https://doi.org/10.1371/journal.pone.0270255Theeboom, T., Beersma, B., & van Vianen, A. E. M. (2014). Does coaching work? A meta-analysis on the effects of coaching on individual-level outcomes in an organizational context. The Journal of Positive Psychology, 9(1), 1–18. https://doi.org/10.1080/17439760.2013.837499