The 2% That Changes Everything

In 2019, a tier-one bank in East Africa completed a comprehensive digital transformation programme. The total investment over three years exceeded $45 million, encompassing a new core banking system, a mobile banking platform, an enterprise data warehouse, and an integrated CRM. By every conventional measure, the programme was a success. The systems went live on schedule. The technology performed as specified. The vendor celebrated. The board was briefed on successful delivery.

Eighteen months after go-live, an internal audit revealed something that the delivery metrics had entirely obscured. System utilisation across the bank's 340 branches was at 38 percent. Branch staff were using the new core banking system for basic transactions — deposits, withdrawals, balance inquiries — exactly as they had used the previous system. The CRM was effectively empty, populated only by the data migrated from the old system and updated by no one. The enterprise data warehouse produced the same reports that the old system had produced, in the same formats, for the same recipients. The mobile banking platform had attracted customers, but the cross-selling capabilities it was designed to enable had never been activated because the teams responsible for cross-selling had never been trained to use them.

The bank had spent $45 million to achieve, functionally, a cosmetic upgrade. The technology was new. The organisational capability was unchanged. And the 2 percent of the total programme budget that had been allocated to training — approximately $900,000 spread across 4,200 staff over three years — had determined the outcome of the entire investment more decisively than the $44 million spent on technology.

The 2 Percent Pattern

This is not an isolated case. It is the dominant pattern across enterprise technology investments in African institutions. The numbers vary — sometimes the training allocation is 3 percent, sometimes 5, occasionally as low as 1.5 — but the structural dynamic is invariant. Organisations invest heavily in technology infrastructure and minimally in the human capability to use it. They then measure success by whether the technology was delivered, not by whether the organisation's capability was transformed.

The consequences are predictable and well-documented. A 2023 study by the Technology Services Industry Association examined 1,200 enterprise software deployments across emerging markets and found that training investment was the single strongest predictor of system utilisation — stronger than technology quality, vendor reputation, or executive sponsorship. Organisations investing less than 5 percent of total project cost in training achieved average utilisation rates of 27 percent. Those investing 15 to 20 percent achieved utilisation rates of 78 percent. The difference between these two groups was not a marginal improvement. It was the difference between a successful transformation and an expensive decoration.

For a $45 million programme, the gap between 27 percent and 78 percent utilisation represents approximately $23 million in captured versus wasted capability. The additional training investment required to move from the lower utilisation tier to the higher one: approximately $6.75 million, or 15 percent of total programme cost. The bank saved $5.85 million by underinvesting in training and wasted $23 million in system capability as a consequence. This is not a mistake of execution. It is a mistake of philosophy — a fundamental misunderstanding of what enterprise technology investment actually produces.

The Philosophical Error

The error is this: organisations treat enterprise technology as a product that delivers value upon installation. It is not. Enterprise technology is an instrument that delivers value through use — and the quality of use is determined entirely by the capability of the people using it. A $45 million core banking system in the hands of untrained staff produces no more value than a $5 million system in the hands of trained staff, and considerably less value than a $15 million system deployed with proportional investment in human capability.

This should be obvious, but it runs contrary to every incentive in the enterprise technology ecosystem. Vendors sell software, not capability. Consultants implement systems, not organisational learning. Boards approve capital expenditures for tangible assets, not operational expenditures for intangible competencies. The entire procurement and governance apparatus of enterprise technology is designed to optimise for the wrong variable — the quality and cost of the technology — while treating the variable that actually determines value — organisational capability — as a residual expense to be minimised.

The result is a continent-wide pattern of underperformance that is consistently misdiagnosed. When a $45 million system delivers 38 percent utilisation, the diagnosis is change resistance, or data quality, or integration complexity, or cultural barriers to adoption. These diagnoses are not wrong — all of these factors are present. But they are symptoms of a single root cause: the organisation did not invest in developing the human capability required to use the technology it purchased. The change resistance exists because staff were not prepared for the change. The data quality is poor because staff were not trained to enter data correctly. The integration appears complex because staff do not understand how the integrated system works. The cultural barriers exist because no one invested in changing the culture.

The Mathematics of Training Investment

The financial case for adequate training investment is not marginal. It is overwhelming. Consider the mathematics across three scenarios for a $20 million enterprise technology programme.

In the first scenario — the African enterprise average — training receives 3 percent of the budget, or $600,000. Based on the TSIA data, expected utilisation is approximately 30 percent. The functional value of the system — the capability actually captured by the organisation — is approximately $6 million. The remaining $14 million in system capability sits unused. The effective cost of the capability delivered: $20 million for $6 million in value, or $3.33 per dollar of capability.

In the second scenario, training receives 15 percent of the budget, or $3 million. Expected utilisation rises to approximately 65 percent. The functional value captured: $13 million. The effective cost: $20 million for $13 million in value, or $1.54 per dollar of capability. This is less than half the cost-per-capability of the first scenario, achieved by redirecting $2.4 million from technology to training.

In the third scenario — the global best practice — training receives 20 percent of the budget, or $4 million. Expected utilisation: approximately 78 percent. Functional value captured: $15.6 million. Effective cost: $1.28 per dollar of capability. This represents a 62 percent improvement in cost-efficiency compared to the first scenario, achieved through a reallocation of $3.4 million — a sum that represents less than 17 percent of the total budget.

The maths is clear: every additional dollar moved from technology to training produces a positive return until training investment reaches approximately 20 percent of total budget. Beyond that threshold, returns continue but at diminishing rates. Below that threshold, every dollar withheld from training costs the organisation multiple dollars in wasted technology capability.

Why African Enterprises Underinvest

If the case for adequate training investment is so financially compelling, why do African enterprises consistently underinvest? Five structural factors explain the pattern.

First, budget categorisation. Technology investments are classified as capital expenditure and governed by procurement frameworks designed for tangible assets. Training is classified as operational expenditure and governed by HR or learning and development budgets that are structurally smaller and more vulnerable to cost-cutting. When technology and training compete for the same budget line — which they rarely do — training loses. When they are governed by different budget lines — which they usually are — training is constrained by a budget ceiling that bears no relationship to the technology investment it is meant to support.

Second, vendor influence. Technology vendors design project plans with training as a minor deliverable — typically 5 to 10 days of system navigation instruction delivered in the final weeks of implementation. This training is designed to satisfy the contractual obligation to train users, not to develop organisational capability. Because vendors define the project structure and the organisation follows it, the vendor's implicit training philosophy — minimal, procedural, event-based — becomes the organisation's training approach by default.

Third, evaluation timing. Technology delivery is evaluated at go-live. Training adequacy can only be evaluated months later, when utilisation data reveals whether the organisation has absorbed the new capability. By that point, project budgets are closed, project teams have disbanded, and the connection between training investment and system performance has been obscured by dozens of intervening variables. The feedback loop between training underinvestment and system underperformance is too slow and too noisy to drive real-time corrective action.

Fourth, cognitive bias. The sunk cost of the technology creates a psychological framing where additional investment in training feels like throwing good money after bad. The system is built. The licence is paid. Surely people will figure out how to use it. This reasoning is exactly backwards — the sunk cost of the technology is precisely what makes training investment essential, because without training, the sunk cost produces no return — but it is psychologically compelling and organisationally pervasive.

Fifth, accountability gaps. No one in most African enterprises is accountable for the gap between system capability and organisational capability. The CIO is accountable for delivering the system. The HR director is accountable for general staff development. Neither is accountable for ensuring that the staff can actually use the system that was delivered. This accountability gap creates a no-man's-land where the most consequential investment decision of the entire programme — how much to spend on training — receives the least executive attention.

The 2 Percent Correction

Correcting the 2 percent problem does not require a revolution in how African enterprises approach technology. It requires three specific, actionable changes that any institution can implement starting with its next technology investment.

First, mandate a minimum training allocation of 15 percent of total project cost, ring-fenced and protected from reallocation when other project costs overrun. This should be a board-level governance requirement, not a project-level decision. The ring-fence is essential because training budgets are invariably the first casualty of cost overruns — and cost overruns occur in the majority of enterprise technology projects. Without explicit protection, the training budget will be cannibalised to fund technology delays, and the pattern will repeat.

Second, redefine project success metrics to include capability outcomes, not just delivery outcomes. A project that goes live on time and on budget but achieves 30 percent utilisation after 12 months is not a success. A project that goes live three months late but achieves 75 percent utilisation after 12 months is a success. Until boards and executive committees evaluate technology investments on the basis of organisational capability rather than technical delivery, the incentive to underinvest in training will persist.

Third, establish accountability for the training-capability gap. Assign a named executive — ideally the business sponsor of the technology programme, not the CIO — responsibility for ensuring that utilisation targets are met within 12 months of deployment. This executive should have authority over the training budget, the change management approach, and the post-deployment support model. They should report to the board on utilisation metrics with the same frequency and the same rigour that the CIO reports on technical delivery metrics.

The Strategic Consequence

The 2 percent problem is not a training problem. It is a strategy problem. An organisation that systematically underinvests in the capability to use its technology is an organisation that systematically wastes its technology investment. Over a decade, this waste accumulates to hundreds of millions of dollars across the African enterprise sector — capital that was deployed with strategic intent and delivered with technical competence but never converted into organisational capability.

The organisations that correct this — that shift even 10 percentage points of budget from technology to training — will achieve dramatically higher returns from every technology investment. Not marginally higher. Dramatically higher. The data is unambiguous. The mathematics is straightforward. The only barrier is the institutional willingness to accept that the 2 percent that changes everything is not the technology. It is the investment in the people who use it.

The next time your organisation approves a major technology investment, ask a single question: what percentage of this budget is allocated to ensuring our people can use this system at full capability? If the answer is less than 15 percent, you are not making a technology investment. You are making a technology purchase. And the difference between a purchase and an investment is precisely the difference between a system that sits at 38 percent utilisation and one that transforms how your organisation operates. That difference is the 2 percent. And it changes everything.