The imported model fails politely
A consulting model is a set of assumptions about the environment it will operate in. Most of the models that arrive on the continent were built somewhere else, for somewhere else, and they carry assumptions that do not hold here: reliable connectivity, a particular regulatory backdrop, deep benches of pre-trained specialists, clients who buy slideware. When those assumptions break, the model does not announce it. It simply produces advice that cannot be executed.
The imported model delivers a strategy that is internally coherent and locally inert. The transformation roadmap assumes infrastructure that is not there. The data programme assumes a regulatory posture that does not apply. The target operating model assumes a labour market that does not exist in that shape. The deliverable is impressive and the outcome is nothing, because the distance between the recommendation and the conditions of execution was treated as someone else’s problem to close.
We start from the opposite premise. The conditions of execution are the whole problem.
What “African realities” changes in the work
The phrase risks becoming a slogan, so it is worth being concrete about what it actually changes.
Infrastructure becomes a variable rather than a constant. Connectivity, power, and data residency cannot be assumed stable across a deployment. A solution that is correct in a lab and fragile in the field is not correct yet. The failure is seldom dramatic. A field tool that assumes a live connection does not crash in a low-coverage area, it quietly captures nothing, and the gap stays invisible until a decision is made on data that was never collected. Designing offline-first is not gold-plating. It is the difference between a dataset and a dataset with holes you cannot see.
Regulation becomes a design input from the first conversation. Data protection on the continent is not a footnote. South Africa’s POPIA, and the comparable frameworks emerging in other jurisdictions, change what may be collected, where it may live, and how it must be governed. A practice that treats compliance as a closing checkbox builds rework into every engagement, a point that becomes sharp the moment you start a data programme in a regulated sector like healthcare.
And talent is read accurately. The shortage on the continent is rarely raw capability. It is structured opportunity and applied development. A practice designed for local reality builds capacity as it delivers and leaves teams stronger rather than dependent on the next engagement.
Localisation is the harder version, not the cheaper one
Localisation is often heard as “the cheaper version.” It is closer to the opposite. Designing for the actual environment costs more thought up front and far less rework later, because it refuses the convenience of assumptions that do not hold. This is why our model is embedded rather than advisory. We put people inside the client environment, accountable for the outcome on the same ground they are advising about - we call them the Boots - so the advice is tested against reality while it is being given, not after we have left. The same instinct drives how we build software fast without breaking it. The measure is simple. Does the advice survive contact with the conditions it was given in? If it does, it was built for the ground. If it does not, it was borrowed.
Key takeaways
- Consulting models encode environmental assumptions, and imported ones fail quietly when those assumptions do not hold here.
- “African realities” is concrete: variable infrastructure, consequential local regulation, and capable but underserved talent.
- Localisation is a rigour standard, not a discount. It is the harder version done properly.
- Embedded delivery keeps advice honest by testing it against the conditions it is given in.