In , a Hungarian physician named Ignaz Semmelweis observed a chilling discrepancy in the mortality rates of two maternity clinics at the Vienna General Hospital. In the first clinic, where medical students were taught, the death rate from puerperal fever was nearly 10%; in the second, where midwives trained, it was less than 4%.
10%
4%
The statistical reality Semmelweis faced in the clinics of Vienna.
Semmelweis eventually realized that the students were coming directly from autopsies to the delivery room, carrying “cadaverous particles” on their hands. He instituted a mandatory washing of hands in a chlorine solution, and the mortality rate in the first clinic plummeted to match the second.
Despite the data being undeniable and the results immediate, the medical establishment of the time did not merely ignore him; they were offended by the suggestion that a gentleman’s hands could be unclean. They chose the comfortable anecdote of their own status over the uncomfortable evidence of the sink, and Semmelweis eventually died in an asylum, his data discarded because it lacked the social permission to be true.
The Ritual of the Data-Driven Roadmap
We often imagine that we have moved past this, that the presence of a heatmap or a statistical significance calculator prevents us from being the doctor who refuses to wash his hands. We have replaced the “gentleman’s status” with the “founder’s intuition,” but the mechanism remains the same.
In the modern boardroom, we perform the ritual of the data-driven roadmap with the solemnity of a high mass. There are slides. There are percentages carried out to the second decimal point. There are interviews with users who were paid $75 in Amazon gift cards to tell us their hopes and dreams regarding a navigation menu.
We spend , or perhaps , assembling a mountain of evidence that suggests Version A is the clear winner for the hero section of the new homepage. The bounce rates are lower, the clarity scores are higher, and the eye-tracking software suggests that people are actually looking at the value proposition rather than the stock photo of a smiling woman in a hard hat.
Then the meeting happens.
The founder, or the VP, or the person whose name is on the building, leans back and stares at the screen. They look at Version A-the winner-and then they look at Version B, the one that the data suggests is a failure. They mention that they showed both to their spouse over a glass of wine the previous evening, and the spouse found Version A a bit “cold.”
Or perhaps they mention a competitor they admire who uses a style more like Version B. Within eleven seconds, the mountain of evidence is reduced to rubble.
The room, sensing the shift in the gravitational field, begins to adjust. The researcher, who spent forty hours on the analysis, finds themselves nodding and saying that Version B does, in fact, have a certain “emotional resonance” that the data might not have fully captured.
Because the leader’s intuition precedes the collective agreement, the meeting is not a forum for discovery but a theater for alignment.
The waste here is not the choice of Version B. It is entirely possible that Version B is actually better. Intuition is not a mystical force; it is often just pattern recognition that happens too quickly for the conscious mind to articulate. The founder might be right about the “coldness” of the design in a way that a heatmap cannot quantify.
The catastrophe is not the decision; it is the sequence. It is the three weeks of performative research that was never intended to be an input, only a costume for a choice that had already been made in the private theater of a single mind.
Therefore, the cost of a “data-driven” culture that ignores its data is not just the budget of the research, but the institutionalized cynicism of the people who produce it.
The Courtiers of Data
When you ask a team to find the truth and then you show them that the truth is a variable dependent on your mood at dinner, you are teaching them that evidence is decorative. You are training your best people to stop looking for what is real and start looking for what is defensible to you.
They will begin to filter their findings through the lens of your known biases. They will stop bringing you the uncomfortable data that Semmelweis brought to the doctors of Vienna, and they will instead bring you a mirror. They will stop being researchers and start being courtiers.
This is the central friction of the modern agency-client relationship. The agency wants to prove its value through “rigor,” which usually means more slides and more data points. The client wants to feel confident, which usually comes from a sense of personal resonance.
The Monarchy of Intuition
When these two forces collide, we get the “retrofitting” phase. This is the period after the decision has been made but before the project moves forward, where the agency must now edit the deck to make it look like the data actually pointed to Version B all along. It is a lie that everyone in the room participates in so that they can maintain the collective fiction that they are a rational organization.
What if we stopped pretending that every decision was a mathematical proof? What if we acknowledged that some things are decided by taste, some by hierarchy, and some by evidence?
The most efficient organizations are those that make the basis of their decisions explicit. If a project is being driven by the founder’s vision, the team should know that from day one. There is no need to spend $14,000 on user testing if the founder is going to override it anyway. That money could be spent on making the founder’s vision as beautiful and functional as possible.
The friction arises when we pretend we are in a democracy of data while living in a monarchy of intuition.
Defined scope, direct path to launch.
In our work at Coherent Agency, we have found that the antidote to this performative waste is a rigid adherence to a defined sequence.
If you publish your prices and your process, you are essentially making a promise to the market about how reality will be handled. When a client knows exactly what happens in the discovery phase, the wireframing phase, and the design phase, there is less room for the “eleven-second pivot.”
By defining the tiers of engagement-whether it is a $3,500 Launch site or a $12,000 Enterprise build-you are creating a framework where the scope and the decision-making criteria are settled before the first pixel is moved.
When the price is fixed and the timeline is set, the “performance of research” becomes a liability rather than a billable asset. If we only have four weeks to go from discovery to launch, we cannot afford to spend three of them building a deck that no one is going to follow.
We have to find the most direct path to a functional, beautiful result. This forces a level of honesty that is often missing in more amorphous, hourly-billed relationships. We have to ask: “Does this data actually matter to the final product, or are we just using it to feel safe?”
If we look back at the doctors in Vienna, their mistake was not just a lack of hygiene. It was a lack of a feedback loop that they were willing to honor. They had a “process” for understanding disease, but it was a process designed to uphold their status, not to save their patients.
Modern digital projects are often the same. We have a “process” for design, but it is often a process designed to make the stakeholders feel important, not to make the website convert.
The researcher who spent nine days on the left-hand column in the opening scene is not just tired; they are being fundamentally devalued. They are learning that their expertise is a secondary characteristic of the project, a “nice to have” that can be discarded at the first sign of an anecdote.
The Slide-Maker’s Paradox
If this happens enough times, that researcher will stop being a researcher. They will become a slide-maker. They will stop looking for the “cadaverous particles” on the hands of the organization and start looking for the chlorine that smells the least like a challenge to the status quo.
The paradox of the modern workplace is that we have more data than ever before, yet we seem to have less clarity on how to use it. We treat data like a security blanket-something to hold onto when we are scared of making a mistake-but we drop it the moment we see something we like better.
The real innovation is not in the collection of more data, but in the courage to decide, beforehand, which data we are actually willing to be wrong for.
If you are going to pick the homepage your spouse likes, do it on Monday. Do not wait until of research have been completed on Friday. Your team will respect the honesty of your ego more than the dishonesty of your “data-driven” process.
They can work with an ego; they can even sharpen it. But they cannot work with a ghost, and they cannot build a future on evidence that they know is only there to serve as a costume.
Ultimately, the goal of any professional project should be the removal of the unnecessary. This includes unnecessary meetings, unnecessary revisions, and most importantly, the unnecessary performance of rationality.
When we move toward transparent pricing, fixed timelines, and explicit decision-making, we are not just being “efficient.” We are being honest. We are acknowledging that while data is a powerful tool, it is not a substitute for leadership. And leadership, at its best, is the ability to make a choice and own the consequences, rather than hiding behind a 40-page deck and a wife’s dinner-table feedback.