Why agencies are the wrong choice for reward infrastructure
The agency model for reward programmes made sense when there was no alternative. The merchant relationships, the physical logistics, the market knowledge — agencies provided genuine value that didn't exist elsewhere. In 2026, that alternative exists. Here's the full operational, commercial, and strategic case for treating reward issuance as infrastructure — not a service to outsource.
This is not an argument that agencies are bad. It's an argument that for reward programme delivery specifically, the agency model has been superseded by a better option — and continuing to use it is a choice with real costs that many organisations haven't fully examined.
The analogy is payment processing. Twenty years ago, if a business in Lagos wanted to accept card payments, they hired a team to build the infrastructure, or outsourced the complexity to a managed service that took a significant margin. Today, they integrate Paystack or Flutterwave in a week and the payment processing works programmatically at a fraction of the managed service cost. The complexity didn't disappear — it moved from visible effort to invisible infrastructure.
Reward delivery is at the same inflection point. The complexity of issuing local-currency reward cards, delivering them to Nigerian phone numbers via WhatsApp or USSD, managing multi-market catalogues — that complexity hasn't disappeared. It's been absorbed into platform infrastructure. The question is whether you're paying infrastructure prices or agency prices for the same outcome.
What agencies were solving and why it mattered
The agency model for African reward programmes existed because several genuinely difficult problems needed solving:
- →Merchant relationships — which brands would participate, at what redemption values, in which markets
- →Physical logistics — printing, distributing, and reconciling physical vouchers and scratch cards
- →Market knowledge — understanding what reward categories and values worked in Lagos versus Nairobi
- →Compliance — navigating regulatory requirements across multiple markets
- →Customer service — handling recipient queries about reward delivery and redemption
These were real problems. Agencies that could solve them were genuinely valuable. The problem is that most of these problems have now been solved at the infrastructure level — baked into platforms that handle them as default behaviour, not as bespoke service delivery. The value proposition of the agency has been eroded by the very infrastructure it helped create demand for.
The five costs of the agency model
Margin on every transaction
Agencies make money on the spread between what they charge you for a reward and what the reward actually costs. A ₦5,000 reward card issued through an agency typically costs the buyer ₦6,000–₦7,000 when agency margin and operational handling fees are included. At volume — tens of thousands of reward issuances per month — this margin compounds into significant budget inefficiency. Infrastructure platforms charge a flat per-issuance fee with no spread on the reward value.
Lead time on every campaign
An agency-managed programme requires a brief before anything happens. Brief review, proposal, approval, campaign setup, asset production, testing. A standard timeline from brief to live is two to four weeks. For time-sensitive programmes — a product launch promotion, a sports event reward campaign, a regulatory-driven retention initiative — this lead time is operationally crippling. A configured API integration issues the first reward seconds after the trigger fires.
Opacity on performance data
Agency-managed programmes typically produce end-of-campaign reports. You see aggregate data: total rewards issued, total redemptions, overall redemption rate. What you don't see in real-time — because the agency is the system of record, not you — is the granular performance data that would let you optimise the programme while it's running. Which recipient segments are redeeming? Which categories are chosen most often? Which delivery channels are performing? With infrastructure that you control, this data is yours in real-time.
Scaling friction
When a programme performs well and you want to double the volume, an agency arrangement requires renegotiation of scope, revised pricing conversations, additional operational capacity. Scaling a programmatic API integration is a configuration change — you've already built the integration, you just issue more rewards. The scaling economics are categorically different.
Market expansion friction
Adding a new market to an agency-managed programme means finding an agency with market presence, negotiating a new contract, building new merchant relationships, running a parallel setup. Adding a new market to an API infrastructure programme is a configuration change — add the country, specify the catalogue, set the currency.
Every time you brief an agency, you're paying two to four weeks of lead time and a 20–40% margin for something that should take minutes and cost nothing extra.
When agencies still make sense
This argument has limits. There are contexts where agencies remain the right choice:
One-off bespoke events
A single, unique reward activation — a corporate anniversary event, a once-per-year sales conference recognition — where the setup cost of an API integration isn't justified by a single use.
High-touch VIP programmes
Executive recognition programmes, board-level client gifting, or situations where the human element of the reward delivery is itself part of the value.
Market entry before infrastructure exists
If you're entering a market where a platform doesn't yet have local brand partnerships or delivery infrastructure, an agency with existing local relationships may bridge the gap while infrastructure is built.
Genuinely complex fulfilment
Programmes where the reward is a physical experience — a trip, a dinner, an event — that requires human coordination and cannot be reduced to a digital delivery.
Outside these specific contexts — for structured, repeating, volume reward programmes — the agency model is the legacy option. Not the wrong option in absolute terms, but the option that costs more, moves slower, and gives you less control than the alternative.
The transition: moving from agency to infrastructure
For organisations currently running agency-managed reward programmes, the transition to programmatic infrastructure has a practical sequence:
The migration doesn't have to be disruptive. The most common approach is to run the next campaign after a natural cycle end-point through the API platform while the agency relationship transitions to strategic advisory rather than delivery management. The agency's market knowledge is still valuable. Their role in the delivery chain isn't.
Infrastructure approach
How QIFTS works — from API call to recipient redemption
The full flow: trigger → issuance → delivery → redemption. No agency in the middle.
Related reading
What reward programmes actually cost — and what it costs to skip them
The full cost structure of reward programmes, including the hidden costs of agency management that rarely appear on budget lines.