The phrase "evidence-based" has become so widely used that it has almost stopped meaning anything. Virtually every significant decision in every well-run organisation is described as evidence-based. The analysis was thorough. The data was reviewed. The recommendation was well-supported. The process was rigorous.
And yet, decisions made with abundant evidence still go wrong, repeatedly, and in ways that feel surprising each time. If the evidence was good, why was the decision not?
The answer, more often than not, is that evidence-based decision-making is being confused with something simpler and considerably less useful: decision-based evidence-making.
Decision-based evidence-making is the practice of arriving at a conclusion first, through instinct, preference, political pressure, or the simple momentum of a process that has been moving in one direction for long enough that changing it feels more disruptive than continuing, and then assembling the evidence to support it.
It is extraordinarily common. It is almost never acknowledged. And it produces the appearance of rigour without any of the substance, which makes it considerably more dangerous than simply making a decision without evidence, because at least the latter is honest about what it is.
The tell is in the process. In genuine evidence-based decision-making, the evidence shapes the conclusion. The question is set before the analysis begins. The findings are followed where they lead, including when they lead somewhere inconvenient. In its inverted form, the conclusion shapes which evidence gets presented, how it gets framed, which inconvenient findings get quietly set aside, and which analytical approaches get selected because they are more likely to produce a supportive result.
Most leadership teams have sat through both kinds of process. From the outside, they can look identical. The reports are equally thick. The presentations are equally polished. The language of rigour is deployed with equal confidence. The difference lies not in the appearance of the process but in the integrity of the sequence that produced it.
They require the question to be set before the analysis begins. Not a preferred answer dressed up as a question, but an actual question to which multiple answers are genuinely possible, and to which the organisation is genuinely prepared to respond regardless of which answer the evidence produces. This sounds obvious. It is rarely done well.
They require the evidence against the preferred option to receive the same quality of attention as the evidence for it. This means actively seeking disconfirming information, not as a token exercise designed to demonstrate open-mindedness, but as a serious and sustained part of the analytical process. What would have to be true for this to be the wrong decision? Is there evidence that those conditions already exist? These questions need to be asked by people with the standing to insist on a real answer, not offered as rhetorical gestures at the end of a process that has already reached its conclusion.
They require someone in the room with both the standing and the safety to say, plainly and without apology, that the analysis was designed to reach a conclusion rather than test one. That observation is one of the most valuable things that can happen in a decision-making process. It almost never happens spontaneously, because the social cost of making it is high and the personal benefit is uncertain. Creating the conditions under which it can happen is one of the most important things a senior leader can do for the quality of their organisation's decisions.
One of the clearest markers of a healthy evidence-based culture is what happens to dissent. Not aggressive or unconstructive dissent, but the considered, well-reasoned view that the evidence points in a different direction from the emerging consensus.
In organisations where evidence genuinely drives decisions, dissent is treated as information. It slows the process down just enough to test whether the consensus is real or manufactured. It is examined rather than managed. It changes the outcome of conversations, at least sometimes, because the people in the room are genuinely interested in whether it is right.
In organisations where evidence is used to support decisions rather than inform them, dissent is acknowledged politely, incorporated into the record in a way that demonstrates it was heard, and then set aside as the process moves to its predetermined conclusion. The question worth asking after any significant decision process is a simple one: did the dissenting view receive a genuine hearing, or a performed one?
Evidence-based decision-making, done properly, is not about having more data or better analysis. It is about intellectual honesty at the moment when the evidence does not fully support what you want to do. That moment arrives in almost every significant decision. It is rarely comfortable. And it is the moment that separates organisations that use evidence as a genuine input from those that use it as a sophisticated form of decoration.
The leaders who handle it well do not pretend it is not happening. They name it, examine it, and make a conscious and transparent choice about how much weight to give the evidence relative to other legitimate factors: judgement, experience, values, risk appetite, and the practical constraints of the environment in which they operate.
That is a harder and more honest process than assembling a presentation that points to the answer already chosen. It is also the only version that deserves to be called evidence-based.
Our consultants take the time to understand your situation before offering any perspective on scope or method. There is no obligation attached to an initial conversation, and no expectation that you arrive with a fully formed brief. The clearer your thinking, the more quickly we can advise, but we are equally comfortable helping you develop that clarity as the first step. You might find our Getting Started Guide helpful in this process.
View: Our Contact Page
View: Our Getting Started Guide
Next Article: Alingning Strategy and Execution