One Monday morning, my team and I stared at a funnel that should have been improving. The product was decent. People who got through onboarding tended to stick.
The problem was getting them there. Every extra thousand sign-ups came with a bigger marketing bill, more abandoned verification steps, and more tickets from people who simply did not trust a brand new app with their money or data. If you have built in Africa, you know the feeling: distribution can be brutally expensive.
This is why a different playbook is winning right now: bundle into someone else’s distribution rails first, then earn the right to own more of the journey later.
In this piece, I will break down build versus bundle, why bundling is changing the growth math across African tech, and a quick decision model you can run before you spend.
Build vs bundle, in plain English
Build means you own the full customer journey. You acquire users, convert them, onboard them, take payments, and keep them coming back. Your app is the destination.
Bundle means your product rides on top of an existing channel: USSD, WhatsApp, a bank channel, an agent network, a telco menu, or a super-app container. Your product is the value, but the container is the distribution.
Bundling is not a shortcut. It is a trade: you give up some margin and control in exchange for cheaper acquisition, faster activation, and smoother cash collection.
Why the pivot is happening now
Three forces are pushing teams away from a build-first mindset.
- Trust and friction are not evenly priced. A new standalone app often pays a trust tax that shows up as KYC drop-off, higher support cost, and slower conversion.
- Distribution has consolidated. The channels with daily user attention are limited, and the biggest rails already belong to telcos, banks, agent networks, and chat platforms.
- Finally, unit economics have tightened. More founders are being pushed to prove payback, not just growth.
The model: unit economics is a distribution problem
Here is the blunt version: your unit economics are only as good as your distribution.
Build and bundle move different levers. Building gives you more control and potentially higher long-term margin, but it usually raises CAC and time-to-first-value early on.
Bundling tends to lower CAC and lift activation because the user is already in a familiar environment. The trade-off is partner fees, experience constraints, and sometimes slower iteration.
How bundling changes the metrics that matter
CAC usually drops because you are borrowing trust and attention.
Activation improves when the channel already has identities, saved payment methods, or predictable device patterns. Every removed step is a conversion lift.
Retention can improve when the product becomes part of a routine inside a channel users already visit. It can also suffer if the host platform owns the relationship, so you need to design for repeat value.
Payback often improves because embedded payments and collections reduce failed transactions and reconciliation headaches.
Four bundling playbooks that work in practice
- Start with the lowest-friction front door (USSD or WhatsApp): If your first job is to test demand, do not force a full app download. Make the first journey simple: check eligibility, create intent, book, order, or get a quote.
- Embed payments early: Many products fail not because demand is weak, but because paying is painful. Payment rails reduce drop-off and improve cash collection.
- Use agent-led onboarding for high-trust moments: For SMB and offline-first products, agents can turn a confusing process into a guided one. Treat the agent flow like product design: clear scripts, incentives, and feedback loops.
- Partner-led KYC and verification: If compliance is slowing you down, work with institutions that already verify identities. It can compress time-to-activation and reduce risk.
When you should still build a standalone experience
Bundling is powerful, but it is not always right. I would lean toward building when:
- The workflow is deeply differentiated and cannot survive inside a constrained container.
- Your product needs a direct relationship (community, network effects, or high-touch support).
- The customer is high value and high frequency, and owning the channel pays off.
- You need rapid iteration that a partner approval process would slow down.
The one-page scorecard: should you build or bundle first
Give yourself 1 point for each statement that is true today. If you score 6 or more, bundling first is usually the safer bet for the next 6 to 12 months.
Scorecard statement | True today? (1/0) |
Your CAC is rising and paid acquisition is not getting cheaper. | |
Users drop off during KYC, verification, or first payment. | |
Your users already live inside a few dominant channels (telco, bank, chat, agent). | |
Trust is a top objection for your category. | |
Your product value is clear in one or two actions. | |
You can deliver value with a limited UI (menu, chat, lightweight web). | |
Payments or collections are a major source of churn or support cost. | |
You have a realistic partner option with aligned incentives and governance. |
Distribution is product strategy
In African tech, distribution is not a growth tactic. It is strategy.
The teams that win the next cycle will design products that plug into rails, learn fast, and then gradually own more of the journey. If your unit economics are struggling, do not just add features. Re-check how your product gets adopted, trusted, and paid for.
Suggested reading:
- GSMA: The Mobile Economy Sub-Saharan Africa (annual reports).
- World Bank: Global Findex Database (financial inclusion and payments adoption).
- IFC and AFI: publications on digital financial services, KYC, and agent networks.
Tobi Ogunmoyero is a product leader and mentor turning customer insight into scalable products. He leads teams, writes about product strategy, and drives measurable impact.
