Building Data-Driven Organisations
"Data-driven" has become one of those phrases that sounds like a strategy but isn't. Most organisations in Nigeria — SME, NGO, public institution — that describe themselves as working toward data-driven decision-making are, in practice, still making most decisions the way they always have: through experience, instinct, and internal politics, with data referenced occasionally as supporting evidence rather than consulted as the primary input.
This is not a criticism. It is a reflection of what typically goes wrong when organisations try to become data-driven, and why the conventional approach to the problem — buy a dashboard, hire an analyst, declare a strategy — tends not to work.
This article argues that the problem is almost never the absence of data, and almost always the absence of a connected data chain. The solutions that work are operational, not strategic.
Why Data-Driven Fails: The Four Most Common Breakdowns
Organisations invest in data collection — forms, surveys, field tools, CRM systems — but the data goes into a repository that nobody is assigned to analyse. It accumulates. Reports are generated once a quarter for a board presentation. The data that could be influencing daily operational decisions sits idle.
Customer data is in WhatsApp. Sales data is in a spreadsheet. Bookings are in a paper diary. Financial data is in a banking app. Each source has partial information; none of them talk to each other. A meaningful picture of organisational performance requires someone to manually pull data from four different places and reconcile it — which means it almost never happens, and when it does, it's already out of date.
NGOs produce donor reports. Government agencies produce ministerial briefs. SMEs produce accounts for tax purposes. All of these are retrospective, formatted for an external audience, and built to demonstrate compliance rather than to inform decisions. The internal team rarely sees the data in a form that would help them do their jobs differently.
The most fundamental failure: data is collected without a clear operational question that it's meant to answer. "We should track more data" is not the same as "we need to know which inquiry source converts to a booking, so we can stop wasting time on the ones that don't." The question determines what to collect, how to process it, and how to present it — in that order.
The Data Chain: What Actually Has to Work
A functioning data chain has four links. Every link has to be operational for the chain to produce value.
Collection is where most attention goes. It is usually not the weakest link. Most organisations are already collecting data, even if informally. The question is whether the data being collected is the data that actually matters for the questions you need to answer.
Storage is where the chain most commonly breaks in the Nigerian context. Data collected in one place doesn't make it to a central, accessible store. WhatsApp messages stay on phones. Form submissions sit in an email inbox. The data exists, but it's not accessible as data — it's accessible only as a sequence of messages that would take hours to synthesise manually.
Analysis is where most organisations think the problem is. They assume they need an analyst, a data scientist, a BI tool. In reality, for most operational decisions at the SME and NGO scale, analysis doesn't require sophisticated statistical methods. It requires a clean feed of data in one place, filtered by the right variables, displayed in a form that makes the answer obvious.
Decision is the link that almost never gets designed. Who sees the data? When? In what format? What operational decision does it inform? If nobody has defined who acts on the data and how, even a perfect dashboard produces no change. The decision link requires an owner — a person or team who is responsible for reviewing the data and adjusting based on what it shows.
The most valuable intervention is almost never a better dashboard. It's connecting the collection and storage links so that analysis is possible in the first place.
What This Looks Like for Different Organisation Types
For SMEs, the data chain is typically short and the storage problem is most acute. The fix is usually a simple integration: customer interactions flow automatically into a spreadsheet; booking confirmations append a row; payment notifications trigger a log entry. The "dashboard" can be a filtered view of that spreadsheet. The decision is made by the owner, weekly, based on a thirty-second review.
For NGOs, the collection problem is more complex because data comes from multiple field locations in multiple formats. The fix requires standardising the collection instrument (consistent form, consistent fields), building a central aggregation point (a database or connected spreadsheet), and designing a reporting view that programme staff can actually navigate. Donor reports become a by-product of the operational data rather than a separate manual exercise.
For government institutions, the decision link is usually the weakest, because data review and decision-making are separated by multiple organisational layers. The design challenge is producing a summary view that is actionable at the level where decisions actually get made — which is usually not the executive suite, but the mid-level operational manager.
The AI Readiness Dimension
TDA's research arm — the AI Readiness Index — provides evidence-based benchmarking for Nigerian public institutions specifically across five dimensions of organisational readiness: strategy, data infrastructure, technical capacity, governance, and leadership. The Index is directly relevant to the data chain question: most institutions assessed at a low readiness level are failing at the storage and decision links, not at strategy or ambition.
View the AI Readiness Index ↗Where to Start
The most effective starting point is a data audit: map what data exists in your organisation, where it lives, and what operational question it's closest to answering. In most cases, the answer will reveal that the storage link is the primary constraint, and that fixing it doesn't require new collection infrastructure — it requires connecting what already exists.
TDA works with organisations at all stages of this process — from initial audit through to building the automation that closes the storage gap and makes operational data genuinely accessible.
Be the first to leave a comment on this article.