Agent banking is the most visible layer of Nigeria’s financial inclusion push. Yet behind every umbrella kiosk and POS terminal is a quieter question banks have struggled with for years: where should agents actually go, and what makes them stay active?
We sat down with Olaitan Moses Ojo, lead digital channels & analytics architect at Micro-Save Consulting, who led the analytics redesign that repositioned Nigeria’s FirstMonie agent network. Ojo has also recognized by the Nigeria Technology Awards (NiTA) as “Most Outstanding Digital Channels & Analytics Architecture Professional of the Year.”
TE: What was the FirstMonie problem you were asked to solve?
Ojo: FirstMonie already had scale. The bank had recruited a large number of agents and the programme was visible across Nigeria. But when we analyzed the network in 2016, performance wasn’t matching the ambition. We saw heavy clustering in urban and already-banked locations, weak agent density in many rural and peri-urban LGAs, and a dormancy pattern where too many agents were registered but not meaningfully active.
FirstBank’s question was not “How do we add more agents?” It was “How do we make this network truly inclusive and sustainably active?” That is what MSC was engaged to support, and I led the analytics work that guided the repositioning.
TE: What did you build first?
Ojo: We built a geo-analytics engine that could show financial access in a way the bank could act on. We didn’t start with recruitment targets; we started with mapping need and access.
We integrated several data layers: the locations of branches, ATMs and existing FirstMonie agents, population and economic-activity indicators by LGA and community, and the actual transaction patterns coming through the network. The idea was to stop relying on gut feel and start using evidence to see where the real gaps were.
TE: So what exactly is the “Financial Access Gap” model?
Ojo: It’s a scoring framework that estimates where financial access is weakest relative to demand. In plain terms, the model asks: if you deploy one more well-supported agent here, will it materially increase access, or will it just add another outlet to an already-served area?
We combine four dimensions: need in that community, the current density of formal access points, usage patterns for financial services, and the depth of services available at existing outlets. When need is high, but access and usage are low, the gap score rises. Those are the communities where an agent can change the inclusion map, not just inflate a headcount dashboard.
TE: How did that change FirstMonie’s expansion strategy?
Ojo: It converted expansion into a ranked, phased rollout plan. For each region, FirstMonie could see a prioritized list of LGAs and micro-markets based on their access-gap scores. At the same time, overserved pockets were clearly visible, so the bank avoided adding agents where new sign-ups would only cannibalize existing outlets.
In practice, it shifted the network from opportunistic growth to deliberate coverage. FirstMonie began to recruit and activate agents in rural, peri-urban, and low-income clusters that were previously under-penetrated, while slowing recruitment in already saturated urban micro-markets.
TE: What about dormancy? Placement doesn’t automatically keep agents active.
Ojo: Exactly. Coverage and activity have to be solved together. So, alongside the gap model, we built agent-level scorecards for FirstMonie. The scorecards looked beyond simple transaction counts. They tracked stability of activity over time, service mix, liquidity patterns, and early warning signs that an agent was drifting toward dormancy.
This gave the programme a way to identify anchor agents worth deeper support, spot at-risk agents early, and tailor training or operational help before dormancy became irreversible. It made network management proactive instead of reactive.
TE: Data quality in Nigeria can be chaotic. How did you deal with that?
Ojo: We assumed imperfect data from day one. Step one was cleaning and geocoding, standardizing location records, validating coordinates, and reconciling inconsistencies across systems. Step two was iterative improvement. We launched a workable model and then refined it as new data came in and as field teams corrected anomalies.
Importantly, we built a feedback loop with regional and branch teams. They flagged local realities, security factors, seasonal market behaviour, or obvious mapping errors, and those insights fed into model updates. Over time, the process improved FirstMonie’s own data discipline as well.
TE: What outcomes did FirstMonie see after adopting the model?
Ojo: Internally, three shifts were clear. First, FirstMonie’s footprint expanded to cover nearly all LGAs in Nigeria with more balanced distribution. Second, the share of agents located in high-gap (previously underserved) communities rose significantly. Third, the active-agent ratio improved in multiple regions because the programme was no longer placing agents in weak locations or leaving viability to chance.
The point is not that analytics magically solved everything; it created a repeatable, evidence-based way to expand inclusion and sustain activity.
TE: NiTA recognised you in 2022. How does that tie into this work?
Ojo: For me, it was meaningful because agent banking is often discussed only in terms of physical footprint. The award recognized the hidden layer, the models and decision rules that determine where agents go, how they’re supported, and whether inclusion goals are actually achieved.
TE: What’s the one lesson other banks or fintechs should take from this?
Ojo: Don’t treat agent expansion as a volume race. Treat it as an access-gap and viability problem. If you can measure where the gaps are, prioritize them honestly, and manage agents with the same rigour you use for any other channel, you get a network that grows for impact, not for optics.
This interview has been edited for clarity and length. Published by Techeconomy as part of our coverage on digital finance infrastructure and financial inclusion in Nigeria.