Dispatch · Seeker Intelligence · Vol. I

The Problem With Research That Arrives After the Decision

Traditional research takes weeks. Your investment committee meets tomorrow. Most organisations have learned to make decisions without the intelligence they needed — and call it good judgement.

A
Aminu RabiuFounder, Kofa Insights Limited
October 2026
6 min read

It arrives on a Tuesday. The research you commissioned three weeks ago — the market assessment, the stakeholder analysis, the field findings that were supposed to inform the decision about whether to proceed with the programme in that region.

The decision was made on Thursday of last week.

The investment committee had a slot. The board paper had a deadline. The partner organisation needed an answer before the end of their fiscal quarter. The process moved forward because processes move forward — because waiting for research that is still in fieldwork is not an option when the window to act is open and the people who need to decide are in the room.

So the research arrives after the decision. Which means the research does not inform the decision. It documents it — or, more commonly, it is quietly filed in a shared drive where it confirms what the decision assumed, and nobody spends much time on the parts that don't.

This isn't an unusual situation. It's one of the most common experiences in development finance, impact investing, and applied research. And it's almost never discussed, because discussing it means admitting that most of the intelligence commissioned for important decisions never reaches those decisions in time to change them.

The same engagement — where the research arrived and where the decision happened
Day 1Research commissioned
Day 7Brief finalised
Day 14Decision made
Research still in fieldwork
Day 21Research delivered
Day 1–7
Research scoped and commissioned. Brief finalised. Fieldwork mobilised. The work begins.
Day 8–13
Fieldwork underway. Interviews being conducted. Data being collected. No findings available yet.
Day 14
Investment committee meets. Decision required. Research is not ready. Decision is made on existing analysis, sector knowledge, and the judgement of the people in the room.
Day 15–20
Analysis and report writing. The research continues — for a decision that has already been made.
Day 21
Research delivered. Filed. The parts that align with the decision are noted. The parts that don't are set aside for "future consideration."
The research itself was fine: the fieldwork was genuine, the findings accurate. None of that mattered, because it arrived after the decision it was supposed to inform. The problem was never the quality — it was the timing.

The timeline that nobody talks about

The gap between when a decision needs intelligence and when traditional research delivers it is not a failure of execution. It is built into the design of research itself.

Traditional research is optimised for accuracy and defensibility. It is designed to produce findings that can withstand scrutiny: representative samples, rigorous methodology, careful analysis, thorough report writing. All of those things take time. And the time they take, taken honestly, almost never fits the window that decisions actually have.

A realistic traditional research timeline — what each stage takes, and what the decision is doing at the same time
01

Procurement and scoping

Writing the terms of reference, going through procurement, selecting the research firm or team, finalising the methodology, and agreeing on deliverables. For institutional clients, this involves multiple internal approvals.

5–10 days
Decision process: active
02

Fieldwork preparation

Developing data collection instruments, training field teams, arranging access and logistics, translating materials where needed, piloting the approach. The quality of this stage determines the quality of what comes next.

5–7 days
Decision process: active
03

Data collection

The actual fieldwork — interviews, surveys, observations, community engagement. For qualitative research across multiple geographies, this is where most of the time is spent. Rushing it produces findings that aren't worth having.

7–14 days
Decision process: moving
04

Analysis

Transcription, coding, synthesis, identifying patterns across the data, checking findings against alternative interpretations. This is where the fieldwork becomes intelligence. It cannot be compressed without losing the quality that justifies doing the research at all.

5–10 days
Decision often made
05

Report writing and review

Drafting, internal review, revision, client review, final revision, formatting, delivery. For institutional clients, this stage involves multiple rounds of comment that add time without adding intelligence.

5–10 days
Decision already made
Realistic total timeline4–7 weeksFor a well-conducted qualitative research engagement

Four to seven weeks isn't carelessness. For research designed to produce findings that are accurate, methodologically sound, and defensible to a funder — four to seven weeks is doing it properly. The research isn't the problem; four to seven weeks is simply longer than most important decisions can wait.

What happens in the gap

The space between "research commissioned" and "decision made" is not empty. Organisations fill it with the information they have, the judgment of the people in the room, and the institutional momentum that accumulates the longer a decision has been in process. Three specific patterns repeat across almost every organisation that faces this gap regularly.

What organisations do when the intelligence has not arrived — three patterns and what each costs
Pattern
What happens
What it costs
Use what's available
The organisation makes the decision on the intelligence it has — desk research, sector analysis, the judgment of team members who have worked in similar contexts. It is not that different from what the research would have added. Probably.
The assumption that the research wouldn't have changed it. Sometimes that assumption is correct. Often it isn't. The decisions most worth commissioning research for are the ones where the operating reality was genuinely uncertain — which means the research was most likely to change the outcome. That is also when it is most likely to arrive too late.
Reverse justification
The research arrives after the decision. The team reviews it looking for the parts that align with what was decided. The parts that challenge the decision are noted as "nuances" or "context-specific findings" that don't change the overall direction. The research becomes documentation rather than intelligence.
The research was never going to be used. The decision was made before it arrived. But the process looks like evidence-based decision-making — which means the next decision will be made the same way. The pattern becomes invisible precisely because it always produces something that looks like rigour.
Institutionalise the workaround
After enough decisions made without the intelligence they needed, organisations stop expecting research to inform decisions and start expecting it to document them. The research process and the decision process become parallel tracks that rarely meet. Both continue. Neither serves the other.
The expectation that intelligence cannot reach decisions is wrong — but it becomes self-fulfilling. Organisations that stop trying to get intelligence before decisions make decisions without it indefinitely. The cost is not visible in any single decision. It accumulates across everything the organisation does without the information it needed.
Intelligence that arrives after the decision isn't intelligence — it's documentation. The difference between the two isn't about quality; it's about timing.

Why traditional research is structurally late

The lateness isn't a product of bad execution. It comes from different design objectives. Traditional research and practitioner intelligence are optimised for different things — and when you use the wrong instrument for a decision's actual timeline, the result is predictable.

Two instruments — two design objectives — two different relationships with the decision timeline
Traditional research — designed for
Statistical defensibility — findings that hold up to scrutiny by a funder, a board, or a peer reviewer
Methodological rigour — a process that can be described, replicated, and audited
Broad coverage — enough data points to make claims that go beyond one person's account
Comprehensive documentation — a report that covers the full scope of what was investigated
Sequential process — each stage building on the last, in an order that cannot be compressed without losing quality
Practitioner intelligence — designed for
Decision relevance — intelligence keyed to the specific question the brief was designed to answer, not comprehensive coverage
Speed — 72 hours from brief submission to a decision asset, because the decision cannot wait for a sequential process
Operational depth — the specific, conditional knowledge of one practitioner closest to the context, not a sample of many people's general views
Immediate use — notes formatted for an investment memo, board paper, or design decision — not a research report that requires further synthesis
Honest scope — precisely one practitioner's account of one context, clearly attributed. Not broader than what it is.
Neither is better — they're built for different jobs. The problem is using the instrument designed for comprehensive accuracy when the decision needs rapid, specific intelligence, and having nothing arrive when it would have mattered.

This is the structural mismatch. Research designed for defensibility requires a process that takes weeks. Decisions often have windows measured in days. Using the former to serve the latter produces late intelligence — not because the research was wrong, but because the instrument was designed for a different job.

What 72 hours actually changes

The 72-hour timeline is not a feature but a design requirement: the maximum time between a brief and a decision asset that still allows intelligence to reach a decision before the decision is made. A single practitioner session on the Kofa platform — brief, prep room, structured agenda, transcribed and organised into a decision asset — can happen within that window. Nothing about conventional research can.

The same decision — what changes when intelligence arrives before it rather than after
Moment
Research arrives after the decision
Practitioner session arrives before
Assumption testing
The assumption at the centre of the decision was never tested against operational reality. The research, when it arrived, either confirmed it or challenged it too late.
The assumption is tested before the decision. The practitioner says something that changes the assumption — specifically, conditionally, from direct experience. The decision is made differently as a result.
Board paper
"Based on available analysis and sector knowledge." The intelligence section is thin because the research hasn't arrived. The board approves on the basis of what's there.
The board paper cites a verified practitioner session. Specific, attributed, keyed to the question the board needed answered. The board has evidence, not just analysis.
Investment committee
The committee makes the decision without the ground-level intelligence that was commissioned. The research arrives two weeks later, confirms most of what was assumed, and notes one exception that would have changed the structure of the deal.
The session happens before the committee meets. The practitioner's account of the market identified the same exception — in time for the deal structure to change before the committee voted, not after.
Programme design
The design is finalised before the field research arrives. When the research identifies a community dynamic the design didn't account for, the programme is already contracted and the design cannot change.
The session informs the design. The community dynamic is identified before the design is finalised — when it can still change what the programme looks like, not after the contracts are signed.
Market entry
The organisation enters the market on the basis of desk research and prior experience. The research commissioned to validate the decision arrives after the entry. It identifies two barriers that a practitioner who knew the market would have named on day one.
The session identifies those barriers before entry. Not because the practitioner knew more than the research team — because they knew it faster, specifically, and in direct response to the decision being made.

The intelligence that arrived in 72 hours wasn't more comprehensive than the research that arrived in three weeks. It was more specific, more conditional, and — critically — present when the decision was being made. Specificity that arrives in time to change a decision is worth more than comprehensiveness that arrives too late to change anything. And because the session is preserved on the platform, that intelligence becomes a decision asset — searchable, retrievable through Ask Kofa — so the next decision begins from what this one learned.

The decisions that cannot wait — and the ones that can

Not all decisions have the same relationship with time. Some have windows measured in months — long enough for comprehensive research to be designed, conducted, and delivered. Others have windows measured in days. Knowing which kind of decision you are facing changes what kind of intelligence you should be trying to get.

Six decision types — their real intelligence windows and what that means for how to serve them
Decision type
Window
What traditional research misses
What practitioner intelligence adds
Investment committee
Days
The committee meets on the next available slot. Research commissioned to inform it rarely arrives beforehand. The decision happens on analysis, not ground truth.
One session, 72 hours. The practitioner's account of the market reaches the committee before it votes. The ground-level intelligence is in the board paper, attributed and citeable.
Deal structuring
Days–weeks
Deal structures are negotiated in real time. The information that shapes the terms — market dynamics, operational risks, informal dependencies — needs to be present during the negotiation, not in a report that arrives after the term sheet is signed.
Session timed to the negotiation. Intelligence delivered before the term sheet is drafted. The informal risk that the research would have identified in week three is named in the negotiation instead.
Board paper
1–2 weeks
Board papers have submission deadlines. Research commissioned to inform them is rarely complete by submission. The intelligence section is written from what's available, which is rarely what the board actually needs to evaluate the decision.
Session completed before the paper is drafted. The practitioner's account is cited in the body of the paper — specific, verified, directly responsive to the questions the board will ask.
Programme design
Weeks
Design workshops have dates. The design is expected to be complete before fieldwork can be conducted. Community-level intelligence arrives as the design is being finalised — too late to change it without extending the timeline and budget.
Session in the design window. One practitioner who knows the community provides the ground-level input that shapes the design before it is locked — not in a post-hoc evaluation that identifies what should have been different.
Market entry
Weeks–months
Market entry decisions are made when the opportunity exists. Waiting for comprehensive research to conclude means waiting for the window to close. Organisations enter on the basis of what they know — which is rarely enough.
Session before entry. The practitioner who knows the market identifies the two or three things that desk research cannot. The entry is structured around operating reality, not the description of it.
Annual strategy
Months
Strategy processes have long enough horizons for comprehensive research to be useful — if it is commissioned early enough and structured around the strategic questions being addressed.
Both instruments useful here. Practitioner sessions to define what to measure. Field intelligence to measure it across the relevant geographies. Traditional research to build the longitudinal baseline. The horizon makes sequencing possible.

The pattern across the first five decision types is consistent: the intelligence window is shorter than the research timeline. That's no coincidence, and it's exactly what makes rapid practitioner intelligence valuable — not as a replacement for rigorous research, but as the instrument that actually reaches decisions in the moments they're being made.

The research that arrives on Tuesday for the decision made last Thursday isn't wrong. It's simply answering the right question at the wrong time. Intelligence that changes decisions has to be in the room when the decision is made — not in the shared drive where the decision is filed.

One question before your next decision
On your last important decision — when did the research arrive, when was the decision made, and what would have changed if the intelligence had been present in the room instead of delivered to the shared drive afterwards?

Aminu Rabiu

Founder, Kofa Insights Limited

Your next decision has a window. Intelligence that arrives after it doesn't count.

Post a brief describing the decision and its timeline. The platform handles the practitioner verification, the prep room, and a structured decision asset within 72 hours — before the committee meets, before the paper is submitted, before the window closes — and the session becomes a decision asset your organisation keeps.

Start a Brief→

Every brief that runs on the Kofa platform is structured around the 72-hour design requirement, not a marketing claim. It is the maximum time between brief submission and decision asset delivery that still allows practitioner intelligence to reach a decision before the decision is made. Every engagement — from brief to prep room to session to decision asset — is structured around that constraint.