Field Intelligence vs. Expert Calls: When You Need One and When You Need the Other
One question is best answered by one practitioner who knows it from the inside. Some questions only reveal their answer across fifty conversations. Knowing the difference changes how you brief.
A
Aminu RabiuFounder, Kofa Insights Limited
September 2026
8 min read
Two briefs arrive at Kofa in the same week. Both are from development finance institutions. Both concern agricultural markets in East Africa. Both are well-written and specific about what they need.
The first asks: "We are deploying a receivables-based lending product with dairy cooperatives in Western Kenya. We need to understand why qualified cooperative members are not applying — specifically whether it is the product design, the branch dynamics, or something else we are missing."
The second asks: "We are evaluating whether to expand our agri-input lending programme from three districts to twelve. Before we commit, we need to understand how input dealer financing actually works across different geographic contexts in the region — what holds, what varies, and where our assumptions about the model break down."
Both briefs are good, and both would produce useful intelligence. But they are not the same kind of question, and answering them well takes different instruments.
Two briefs — the same surface, two completely different requirements
Expert call brief
"Why are qualified cooperative members in Western Kenya not applying for our receivables-based lending product?"
Question typeWhy is something specific happening in a specific context
What it needsOne practitioner with direct experience in that exact context — loan officer, cooperative leader, or branch manager who has seen this from the inside
What it producesA deep, specific, conditional account of what is actually happening and why
Timeline72 hours from brief to decision asset
Field intelligence brief
"How does input dealer financing actually work across twelve districts in East Africa — and where does our model break down?"
Question typeWhat pattern emerges when many people answer the same question across many contexts
What it needsStructured conversations with dealers, loan officers, and cooperative leaders across multiple districts — coordinated through embedded local coordinators
What it producesA pattern of evidence across geographies — what is consistent, what varies, and where assumptions hold or break
TimelineScoped separately — weeks, not hours
This matters because the most common mistake in practitioner intelligence isn't asking bad questions. It's taking a good question and running it through the wrong instrument: expecting a single conversation to reveal a pattern that only emerges across many, or commissioning a large-scale field exercise for a question one well-chosen session would have answered better, faster, and more precisely.
Both run on the same platform — the same brief, the same prep room, the same decision assets. What changes is the instrument: one session with one practitioner, or structured conversations across many.
What an expert call is built for
An expert call, a single session with a single verified practitioner, is the right instrument when the question has a specific answer that one person close enough to the right context is positioned to give. Many questions that feel like they need scale actually need depth. The seeker thinks they need to speak to many people to understand a market. What they actually need is to speak to the right one person who has been in the market long enough to see how it works.
What an expert call is built for — and where it reaches its design limits
Dimension
Best suited for
Design limits
Question type
Why something is happening. How something actually works. What a formal analysis is missing. Whether a specific assumption holds in a specific context.
Not: how prevalent is this pattern? One practitioner's account is deep, not representative. It describes their vantage point, not the distribution across many contexts.
Geography
A specific geography where the right practitioner has direct experience — the district, the branch, the community where the decision will play out.
Not: what is consistent across twelve districts. A single practitioner knows their geography deeply. They cannot speak for what holds elsewhere with the same authority they bring to their own context.
Decision type
Pressure-testing a specific assumption before it becomes expensive to be wrong about. Understanding a specific barrier before a product is designed around it.
Not: validating a model across a market. One session produces one account. If the decision requires evidence from multiple geographies, one session is insufficient regardless of how good it is.
What it produces
A specific, conditional, citeable account of operating reality from one practitioner's direct experience. Structured notes within 72 hours.
Not: a pattern. A pattern requires variation — the same question answered across enough different contexts for the consistent elements to become visible and the variable ones to be distinguished from noise.
What field intelligence is built for
Field intelligence is an engagement type that deploys the coordinator network across many contexts — structured conversations with multiple communities, geographies, or market participants across African markets, through coordinators already embedded in them, following a shared methodology designed around the objectives of the brief. Where an expert call goes deep into one context, field intelligence goes across many.
What field intelligence is built for — and where its design limits sit
Dimension
Best suited for
Design limits
Question type
What holds consistently across geographies. What varies and why. Whether an assumption that works in one context breaks in another.
Not: why is this specific thing happening in this specific place. Field intelligence reveals pattern. It does not substitute for the depth a single expert session provides when the question is specific and contextual.
Geography
Multiple districts, communities, or market contexts where the same question produces different answers that, taken together, reveal something no single context could show alone.
Not: a substitute for deep local knowledge. Structured interviews across many contexts are not the same as the years of proximity that produce a single practitioner's deep operational understanding.
Decision type
Market expansion decisions where performance across geography matters. Programme design where the model needs to hold across diverse communities. Portfolio decisions where risk concentration needs assessing against a broader pattern.
Not: pressure-testing a single specific assumption. That is faster and more precisely addressed by one expert call. Field intelligence deployed for a question a single session would answer is expensive overkill.
What it produces
A pattern of evidence — structured conversations reviewed through multiple layers of quality assurance, synthesised into findings that reveal what is consistent and what varies.
Not: a representative statistical sample. Field intelligence produces qualitative pattern, not a probability distribution. If the answer requires statistical representativeness, a properly designed survey is the right instrument.
An expert call answers: what is actually happening here? Field intelligence answers: what pattern emerges when that same question is asked everywhere it matters? Different questions, different instruments.
The questions that tell you which one you need
Before deciding which instrument to use, five questions about the brief will usually make the right choice obvious.
Five questions that identify the right instrument before the brief is submitted
01
Is the question asking why something is happening — or how often and where?
→ Expert call
Why questions have specific answers that one person close to the context can give. "Why are qualified borrowers not applying" is a why question. The answer is in one loan officer's account.
→ Field intelligence
How often and where questions require pattern across many contexts. "How prevalent is this barrier across twelve districts" is a distribution question that no single account can answer.
02
Does the decision depend on understanding one specific context deeply — or on understanding how something varies across many contexts?
→ Expert call
A product deployment in a single district. An investment in a specific company. A programme in a specific community. One context, understood deeply.
→ Field intelligence
A market expansion across twelve districts. A programme that must work across diverse communities. A portfolio spanning multiple geographies. Variation is what matters.
03
Is the goal to pressure-test a specific assumption — or to build a map of how something works across a market?
→ Expert call
One assumption, tested by one practitioner with direct experience of whether it holds in the relevant context. Fast, precise, accountable to one vantage point.
→ Field intelligence
A model of how a system works across a geography — built from structured conversations with multiple participants across multiple contexts. The map emerges from pattern, not a single account.
04
How much time does the decision have?
→ Expert call
72 hours from brief to a decision asset. The right choice when the timeline is short and the question is specific enough that one account can change the decision.
→ Field intelligence
Weeks, scoped to the question. The right choice when the decision is significant enough to justify a longer intelligence cycle — and when pattern across many contexts is genuinely required.
05
Does the seeker know what question to ask — or do they first need to understand the landscape before designing the right question?
→ Expert call first
When the question is not yet precise enough for a field intelligence brief, one expert session clarifies what to measure — before commissioning the broader exercise. This is often the right order.
→ Field intelligence second
Once the expert call has clarified what to measure, field intelligence can measure it at the scale the decision requires. The expert call designs the methodology. Field intelligence executes it across geography.
How the coordinator network powers field intelligence
Field intelligence deployments are fundamentally different from conventional qualitative research in African markets — not because the methodology is different in theory, but because the people doing the research are different in kind. Most qualitative research that tries to reach community-level intelligence sends researchers in from outside the geography, with no prior relationship to the communities they study. The research question travels well. The trust needed to get genuine answers often doesn't.
External research team vs Kofa coordinator network — the structural difference
External research team
Researchers arrive without established relationships. Community members calibrate their answers to what a stranger can be trusted with — which is usually less than what they know.
Researchers operate on a timeline set by the research budget, not by the community's rhythms. Trust that takes months to earn is expected to be present within a two-week fieldwork window.
Community leaders perform cooperation for researchers they do not know — public agreement that may not reflect private reality, recorded by someone who cannot tell the difference.
Researchers leave. The community has no ongoing relationship with the research process and no investment in whether the findings are accurate.
The research captures what communities are willing to share with outsiders — different from what they would share with someone they already trust.
Kofa coordinator network
Coordinators have spent years in the communities where they work. The relationships through which they approach participants already exist — with trust established before the first question is asked.
Coordinators understand community rhythms, seasonal patterns, and the moments when genuine conversation is possible. Deployment is timed around community reality, not research budget.
Community leaders know the coordinator. The conversation can be private, frank, and honest in ways that a public consultation with a stranger cannot be. The coordinator can tell performed from genuine agreement.
The coordinator is still there after the research. Their ongoing presence means the relationship through which the research was conducted continues — and that findings can be verified against what actually happens next.
The research captures what communities share with someone they trust — considerably closer to what they know, and considerably more useful for decisions that depend on operating reality.
When to use both — and in what order
The two instruments aren't alternatives — they're complements, and they work best in a specific order. The expert call defines what to measure; field intelligence measures it at scale. An organisation that commissions field intelligence before it knows what it's looking for ends up with a large volume of structured conversations about the wrong question.
Three organisations — how expert calls and field intelligence worked in sequence
DFI · Agri-input lending · East Africa
"We are evaluating expansion from three to twelve districts. We need to understand whether our credit model holds across that broader geography — and where it breaks down."
Expert Call · First
Session with a senior credit analyst across three current districtsRevealed that the model's core assumption — that dealer receivables follow the planting cycle — held in two of three districts but broke in the third due to an informal credit relationship between dealers and a single large cooperative. Defined the variable to test across the expansion geography.
Field Intelligence · Second
Structured conversations with dealers and cooperative leaders across all twelve proposed districtsMapped which districts had similar cooperative concentration risk, which had genuinely distributed dealer networks, and where the model would perform as projected vs where it needed redesign. The expansion proceeded in eight districts, was redesigned for three, and paused for one.
NGO · Maternal health · Tanzania
"Our pilot showed strong uptake in one district. Before we scale to five, we need to understand whether what drove uptake there will work elsewhere — or whether the district was exceptional."
Expert Call · First
Session with the community health extension worker who ran the pilot districtIdentified that uptake was driven primarily by one community midwife's relationship with a specific women's cooperative — and that facility timing during the harvest season was the barrier that had been removed. Both factors were potentially specific to the pilot district.
Field Intelligence · Second
Structured conversations with health workers and community leaders in the five proposed districtsRevealed that two districts had equivalent midwife relationships that could anchor the programme, two had similar harvest-season barriers needing the same timing redesign, and one had a different community structure entirely. The programme scaled differently in each — correctly this time.
Impact Fund · Portfolio construction · West Africa
"We are building a portfolio of last-mile distribution businesses across three countries. Before we finalise the investment thesis, we need to understand what operational risks are structural across the model and which are geography-specific."
Expert Call · First
Session with a logistics manager who had run last-mile distribution across all three countriesIdentified three structural risks present in every context — checkpoint dependency, seasonal route reliability, informal transporter relationships — and two that were country-specific. Defined the risk variables to map across the portfolio geography.
Field Intelligence · Second
Structured conversations with distribution managers and transporters across all three portfolio geographiesMapped concentration of each structural risk across proposed portfolio companies. Revealed that two target investments had concentrated checkpoint dependency requiring active management. The portfolio was constructed with that risk explicitly priced and mitigated.
The pattern holds across all three. The expert call produced the question that field intelligence then answered at scale. Skip the call and the field exercise would have confirmed the assumption rather than tested it; skip the field work and the call would have left the organisation with a deep account of one context it might have wrongly generalised to many. Used together, they produce what neither can achieve alone. And every deployment — one session or across many — leaves decision assets behind, searchable and retrievable through Ask Kofa, so the next expansion starts with what the last one learned.
One question before your next decision
On your next market intelligence brief — are you asking why something is happening in a specific context, or what pattern holds across many? The answer determines which instrument you need, and getting that choice wrong is the most common — and most expensive — mistake in practitioner intelligence.
Aminu Rabiu
Founder, Kofa Insights Limited
Start a brief. The platform runs everything around it.
Submit a brief describing the decision you are making — whether it needs one session or structured conversations across many, the platform runs the engagement. Where our network is active, Kofa finds the right people; anywhere else, bring your own.
Every engagement — expert call or field intelligence — runs on the Kofa platform and produces decision assets your organisation keeps. Expert calls deliver a structured decision asset within 72 hours; field intelligence deployments are scoped through a dedicated conversation before any commitment is made — the question and methodology are agreed before the network is engaged.
Kofa uses Google Analytics to understand how visitors use our site — which pages are most useful and how we can improve. Analytics is configured without advertising features, and your data is never used for advertising. Privacy Policy.