Argument · September 2026 · 3 min read

Why Some Business Development Services Units Survive Their Projects

I have designed and established business development services units inside banks, microfinance institutions, NGOs and government bodies across Egypt. Sixteen of them now run independently inside their hosts. This is what was different about their design.

The usual story of a donor-funded business development services (BDS) center goes like this. A project identifies a gap in enterprise support. It rents an office, hires advisors, subsidizes the services to zero, trains several hundred entrepreneurs, reports the number, and closes. Six months after closure the office is empty and the advisors have joined the next project. The gap is still there. It has been filled, briefly, by a subsidy, and a subsidy is not a service.

The question I put to every team before we placed the first unit was not where the entrepreneurs were. It was: who will own this unit on the day the project ends, and what will need to be true for them to keep it open?

The Host Is the Intervention

That question changed the design. Instead of standalone centers, every unit was embedded in an existing institution with its own reason to want business development services: a bank that needed bankable clients, an MFI whose borrowers were failing for reasons a loan could not fix, an NGO with a mandate it could not deliver, a government body with an enterprise target. The host was not a landlord. The host was the client of the unit, and the unit had to earn its place.

Embedding solves ownership and creates a second problem. A unit inside a bank that has to ask the branch manager before it prices a service, hires an advisor or signs a partnership will do none of those things, because the branch manager has other priorities. The units that fail inside hosts fail for lack of decision rights, not for lack of demand.

Delegation Before Delivery

So the governance model came before the service menu. For each unit we negotiated a delegation framework with the host: which decisions the unit could take alone, which needed the host’s sign-off, how the unit reported, how its performance was measured, and how its costs and revenues were accounted for. We wrote operating manuals and performance standards, and put a quality assurance routine in place that the host, not the project, would run.

The pricing conversation was the hardest. Every host wanted the services free, because free services are easy to fill. We held the line that a service priced at zero is valued at zero, by the client and by the host’s own finance department. Units introduced fees in steps, with the host’s consent, and the fee became the evidence that the unit was worth keeping.

What the Numbers Hide

More than thirty-five organizations is a figure I use, and I use it carefully, because the number is not the point. The point is that the model outlived the programs that funded it, which is the result a market systems intervention should be judged on.

A subsidy is not a service. A unit that cannot decide is not a unit. A host that does not pay is not an owner.

The three rules are simple to state and hard to hold under the pressure of a project timeline that rewards activity. If you are designing enterprise support now, I would ask the same question: who owns this on the day you leave, and have you given them the authority to keep it?

The same question belongs before design, not only at exit. I set out what a diagnostic should tell a donor at that stage in What a Market Systems Diagnostic Should Tell a Donor Before Design Starts.

Work With Me on This

I work with institutions and companies on designing and implementing enterprise-support units. See the Service →

← All Perspectives

Organizations I Have Worked With
USAIDGlobal Affairs CanadaWorld BankCentral Bank of EgyptCommercial International BankIFCKfWEuropean UnionHabitat for Humanity InternationalAga Khan Agency for MicrofinanceDrosos FoundationUN WomenUN-HabitatLFS AdvisoryGoss GilroyEnvironment and Development GroupAECOMFINBI