Organizations that build leadership capability internally outperform those that rely on external hiring. The performance advantage is real. What remains poorly understood is what internal development actually requires, beyond budget allocation and program calendars.
Internal leadership development produces measurable organizational advantages in capability continuity, cultural alignment, and long-term performance that external hiring cannot replicate. The challenge is that most internal development programs are insufficiently differentiated by development need, insufficiently integrated across modalities, and insufficiently measured for developmental impact to produce returns proportionate to their investment. This article examines the evidence on which development modalities produce returns under what conditions, addresses the 70-20-10 framework and the systematic misapplication it has generated in organizational practice, considers the identification of development-ready individuals, and addresses the measurement of development effectiveness as an organizational learning capability.
Day (2001) established the foundational theoretical distinction between leader development, improving the capability of individual people, and leadership development, building the organizational capacity to produce leadership capability systematically and continuously. Organizations that invest in leadership development as an organizational system, rather than as an aggregation of individual development activities, consistently outperform those that treat development as an HR function rather than a strategic capability investment. Collins (2001) found in his longitudinal study of sustained high-performing companies that all had in common a strong internal leadership pipeline, built through deliberate investment in identifying and developing capability from within, suggesting that the capacity to develop and promote internally is itself a distinguishing organizational competency.
The performance advantage of internal over external leadership supply operates through several mechanisms with distinct practical implications. Internal leaders carry more accurate organizational knowledge, including the informal structures, political dynamics, relationship histories, and cultural assumptions that are essential context for effective leadership but essentially invisible to external hires for the first twelve months or more. They carry relational capital with colleagues, stakeholders, and direct reports whose cooperation their leadership requires and that external hires must build from zero, often during precisely the period when they are simultaneously learning the organizational context that would tell them how to build it most efficiently. And they carry the organizational identity commitment that motivates sustained performance investment beyond immediate role requirements and that external hires may develop over time but rarely arrive with at the level their organizations immediately require.
External senior leadership hires fail at rates substantially higher than internal promotions, with estimates ranging from 40 to 60 percent within the first 18 months across research populations and leadership levels. The direct costs of a failed executive hire, including search fees, relocation, transition costs, and severance, are substantial. The indirect costs are typically larger: the organizational disruption of the onboarding period, the talent system credibility damage that failed senior hires create, the performance impact of extended leadership vacancies, and the retention effects on high performers who observe leadership transition instability as a signal about organizational management quality. Organizations that maintain robust internal pipelines reduce both the frequency and the per-incident cost of executive hiring across the full leadership spectrum, producing a compounding investment advantage that accumulates across the years during which the pipeline is actively maintained.
The return-on-investment argument for internal development must account for the investment required to build a genuine pipeline rather than simply to create development program activity. Most organizations that characterize themselves as committed to internal development are investing in development activity, measured by program participation rates, training hours, and coaching engagements, without investing in the organizational systems that make development activity translate into genuine capability advance. Day and Dragoni (2015) argued that the return to individual development activity is substantially moderated by the organizational context in which it occurs, and that organizations with more sophisticated leadership development systems produce higher returns from equivalent individual development investments than those with less sophisticated systems. The investment case for internal development therefore includes both the individual program investment and the system investment that determines its productivity.
McCall, Lombardo, and Morrison (1988) established through retrospective research with senior executives that the majority of what executives attributed to their leadership development occurred through three modalities: challenging job assignments contributing approximately 70% of attributed development, developmental relationships including coaching and mentoring contributing approximately 20%, and formal education and training contributing approximately 10%. This finding has been extraordinarily influential in organizational development practice and has been misapplied with near-equal consistency in ways that reduce rather than improve development return on investment.
The 70-20-10 finding was a descriptive retrospective account of how development has historically occurred in populations of senior executives. It was not a prescriptive recommendation for investment allocation. The frequent inference that formal development deserves only 10% of the development budget because it contributes only 10% of learning outcomes conflates the historical contribution of a modality to aggregate learning with the investment required to make that modality productive under deliberate design. Formal development produces the conceptual frameworks, analytic tools, and structured self-awareness that allow leaders to extract developmental value from experience and from developmental relationships. Without formal development, experience tends to produce confident routines rather than genuine capability development, and developmental relationships tend to provide encouragement rather than the transformative challenge that accelerates development most effectively.
Day and Dragoni (2015) proposed that the most effective approach integrates all three modalities in a deliberate developmental sequence. In their integrated framework, formal development provides the conceptual vocabulary that allows leaders to extract insights from experience rather than simply accumulating experience; developmental relationships provide the ongoing reflection, challenge, and feedback that converts experience from raw material into transferable learning; and challenging assignments provide the authentic organizational complexity in which newly developed capabilities are tested and established as behavioral habits. The modalities are interdependent: formal development without application contexts produces abstract knowledge without behavioral change; challenging experiences without conceptual frameworks and reflective relationships produce behavioral repertoires without understanding of when and how to deploy them across the varied contexts leadership presents.
The organizational design implication of the interdependence argument is that the productivity of the experiential 70% depends substantially on the quality of the formal 10% that precedes it and the relational 20% that processes it. Organizations that cut formal development investment below the level required to provide conceptual frameworks for the experiences they create are degrading the return from their experiential investment by removing the interpretive structure that makes experience educationally productive. Organizations that create challenging experiences without the developmental relationship infrastructure that provides reflection, challenge, and feedback are generating raw experience without the processing capacity that converts it into transferable learning. The 70-20-10 framework, correctly applied, is an argument for the integration of all three modalities in a deliberate sequence, not an argument for concentrating investment in experiential development at the expense of the formal and relational modalities that determine its yield.
The identification of individuals most likely to produce high returns from development investment is a high-stakes assessment challenge that most organizations approach with inadequate rigor. Silzer and Church (2009) reviewed the assessment literature on high-potential identification and found that the characteristics most consistently associated with high development return were learning agility, the ability to extract learning from varied and challenging experiences faster than peers; cognitive complexity, the capacity to hold multiple competing considerations simultaneously; and leadership motivation, genuine interest in influencing others and shaping organizational outcomes rather than primarily in personal advancement.
Learning agility is particularly important as a targeting criterion because it predicts how much capability an individual will develop from a given developmental experience independently of their current capability level. A high-agility individual at moderate current performance produces higher development return from the same experience than a low-agility individual at the same performance level, because the high-agility individual extracts more transferable learning and applies it more effectively in subsequent contexts. Organizations that target development investment primarily on current performance, the most common high-potential identification approach, are investing in the criterion that best predicts current performance rather than the one that best predicts development trajectory. Learning agility is a better predictor of development trajectory than current performance level, which makes it the more appropriate targeting criterion when the goal is maximizing the development return on investment in people.
The assessment approaches with the strongest evidence base for high-potential identification include structured behavioral interviews probing specific examples of learning-from-challenge behavior, multisource assessment capturing peer and supervisor observations across multiple contexts, structured simulation exercises under conditions approximating the specific leadership demands of target roles, and longitudinal tracking of behavioral development trajectory rather than cross-sectional performance snapshots. Organizations relying primarily on supervisor nomination, the most widespread identification method in practice, systematically underidentify individuals not socially proximate to the nominating supervisor. This introduces biases that perpetuate existing organizational demographic patterns and produce high-potential pools less diverse than the underlying talent base would support.
The equity implications of identification quality extend beyond investment efficiency. High-potential identification processes that systematically underidentify certain demographic groups through structural biases of supervisor nomination and social proximity produce leadership pipelines that fail to represent the full capability distribution of the workforce. Improving identification rigor therefore simultaneously improves development return and organizational equity, making the investment in better identification methodology a return-positive and equity-positive intervention rather than a trade-off between the two.
The organizational investment in leadership development is among the largest single discretionary expenditures in most organizations of significant size, and simultaneously among the least rigorously evaluated for the behavioral outcomes it is designed to produce. Kirkpatrick and Kirkpatrick (2006) proposed the four-level evaluation framework distinguishing participant reaction, learning of content, behavioral change following the program, and organizational results attributable to those changes. Research on evaluation practices consistently finds that the overwhelming majority of development programs are evaluated primarily at the reaction level, with rigor declining sharply at each subsequent level. This gap means organizations are unable to distinguish programs producing genuine behavioral change from those producing high participant satisfaction without development impact.
The behavior change criterion is the most practically significant evaluation level and the most rarely implemented with rigor. Behavior change measurement requires pre- and post-development assessment of the specific behaviors targeted, conducted in conditions approximating where those behaviors are most consequential, by observers with adequate behavioral baseline knowledge and observational access to detect meaningful change. Most programs omit this step because it is organizationally costly, because the results might unfavorably evaluate programs that are institutionally well-regarded, or because program designs did not specify behavioral targets with sufficient precision to make measurement criteria derivable from them.
Organizations that implement rigorous behavior-change evaluation consistently improve program quality faster than those that do not, because the measurement creates accountability for programs to produce behavioral outcomes rather than the participant satisfaction that justifies continuation. The transition from popularity-based to effectiveness-based evaluation requires organizational leadership commitment to the principle that development investment is justified by behavioral outcomes rather than by the experience of engaging in development activity. This commitment is widely endorsed in the abstract and inconsistently applied in practice, because behavior-change measurement is organizationally uncomfortable in ways that reaction measurement is not: it reveals which programs are not producing the behavioral impact their investment requires, and that revelation creates pressure to discontinue programs with institutional momentum and constituency.
The organizational learning capability that emerges from sustained, rigorous development effectiveness measurement is itself a competitive advantage. Organizations that have tracked the connection between specific development program features, specific development experiences, and specific behavioral outcomes over multiple cohorts possess empirical knowledge about what works for their specific organizational context that external vendors and benchmarks cannot provide. This accumulated knowledge, treated as a proprietary capability rather than routine administrative data, allows the organization to improve its development investments with each cohort and to build a development system increasingly calibrated to produce the specific behavioral capabilities that its competitive requirements demand.