Introduction

Oscar Wilde built an entire farce on a simple social flaw: the world will sometimes reward the appearance of virtue more than virtue itself. In The Importance of Being Earnest, a name can outperform a character. “Ernest” wins hearts, while earnestness, the real thing, is treated as optional.

It is funny. It is sharp. It is also uncomfortably familiar.

Because in business, and especially in insights, we face the same temptation. To prize the label over the labor. To reward the output as a proxy for the understanding. To confuse performance for progress.

In our world, “Ernest” is the topline that looks decisive. Earnest is the work that makes decisions durable.

And that is the tension I submit to you. Stakeholders want time back so they can make faster, more impactful decisions. Insight people want time to do the first part, to curate the right dataset, validate the signal, and bring the story into focus. We are methodologists and meaning makers, and the only thing we ultimately sell is confidence grounded in truth.

Earnest leadership is the discipline of choosing truth over performance, especially when speed is rewarded.

The puzzle we are always rebuilding

We, as methodologists and ultimately people who decision makers and investors rely on to provide both tactical and strategic direction on a vast array of organizational challenges, are nothing if not purposeful and persistent in the quest for truth and meaning.

We will wrestle for hours, if not days, over the curation of the perfect dataset. We think about business problems as if we were reverse engineering a completed puzzle of the current landscape, requiring more information to better navigate the terrain and add pieces to bring the story into focus.

And yet the pressure is constant. Stakeholders want the story now. Insight people want the story right.

Both are fair.

The risk is what Wilde lampooned. When the room rewards appearances over truth, teams start optimizing for what looks decisive instead of what is dependable. A fast answer can become a costume. A clean chart can become a mask. A confident recommendation can become theater.

Theatre scales. Truth compounds.

Earnest leadership is what keeps the work from becoming performative.

Slow and fast thinking in the moments that matter

In stable conditions, speed is often a feature of good process. You build repeatable systems. You scale what works. You produce reliably. You move.

In dynamic conditions, the machine can become a trap.

When disruption is peaking and information is coming from all directions, the instinct is to accelerate. More dashboards. More trackers. More summaries. More meetings. More inputs. More output.

That is when the best insights leaders do something counterintuitive.

They slow down.

Not to delay. To discern.

Fast thinking under pressure can turn into reflexive thinking. It can mistake noise for signal. It can make yesterday’s mental model feel like today’s map. It can turn an insights team into a factory that produces outputs at scale while quietly losing the plot.

So sometimes the most responsible move is to stop, look, and listen.

Stop the frantic collection of everything. Look for patterns that repeat across sources. Listen for what is changing underneath the words.

Here is a simple decision rule you can use when the environment is volatile.

• Go fast when the question is “what happened” and the data is stable. Retrieve, quantify, trend.
• Go slow when the question is “why it is changing” and signals conflict. Interpret, triangulate, pressure test.
• Go hybrid when speed is required but the cost of being wrong is high. Move quickly on execution, slow down for judgment.

Speed is for retrieval. Slowness is for judgment.

Earnest leadership is the discipline of choosing truth over performance, especially when speed is rewarded.

A client story that puts earnestness to the test

A while back, I worked with a client in the home and garden care sector. They were the market leader in one specific area, confident and established. Then they acquired an adjacent-market competitor with a much more diverse product line, covering a wide range of conditions and problems that property owners and landscape professionals face.

On paper, it looked like a growth story. In practice, it came with a hidden research trap.

Each of these acquired products was low incidence on its own. Not many people used any single one frequently enough to dominate the data. But collectively, across the full set of conditions and use cases, the portfolio represented a significant share.

The business challenge sounded straightforward. Increase awareness and usage. Apply established segments and personas with better, more targeted messaging. Potentially streamline the product line to reduce confusion.

But here was the stake. If they got this wrong, they would optimize for the visible slice of the market and miss the real portfolio opportunity. High incidence categories would hijack the story. The long tail would disappear. Messaging would improve for a few cases, while adoption across the broader portfolio stayed flat.

It would look like progress. It would not be progress.

We convinced them to do something more earnest. Before you fix messaging, understand actual usage and real use cases. That is how you uncover say do gaps. That is how you surface unmet customer needs that have become normalized. That is how you expose friction points that suppress adoption. That is how you map competitive strengths and weaknesses, not in theory, but in lived experience.

And here was the pivot. This would not only improve their messaging. It could improve their offerings.

That single shift changed the nature of the work. We were no longer polishing a story. We were building one that would hold up under scrutiny and guide decisions with confidence.

The method move that made it fast and rigorous

Designing the study was the hard part.

We needed enough records in each product category to speak with confidence. But the categories were low incidence. A broad sample would waste time and money collecting respondents who could not speak to the conditions and use cases we needed.

So we designed a screener that filled a quota schema prioritizing category incidence.

In practical terms, we were deliberate about who entered the study, based on the conditions they faced and the products they used, so we could quickly and efficiently build robust sample sizes across the portfolio. That approach allowed us to validate incidence by category rather than assume it. It also allowed us to map findings back to their existing segments and customer personas.

This is the hybrid model in action.

Fast in execution, because the design prevented waste. Slow in judgment, because we refused to make decisions on incomplete truth.

The payoff was tangible. They stopped debating opinions and started acting on evidence. They discovered which SKUs were confusing, not weak. They learned where friction lived in the journey and where competitive claims were under-supported or over-trusted. They clarified which messages belonged to which use cases and saw where streamlining would remove confusion rather than revenue.

Earnestness did not slow them down. It made them right.

Technology, theatre, and the temptation to discount judgment

Every wave of insights disruption arrives with a promise. Speed. Scale. Automation. Democratization.

And it delivers. The machinery gets faster.

But there is a familiar Wilde-like twist. When a new capability enters the room, organizations can confuse novelty with authority. The loudest output wins. The cleanest summary wins. The fastest answer wins.

Meanwhile, the assets that actually make insights valuable can get pushed to the side or discounted.

Human judgment. Domain authority. Reputation. The hard-earned ability to separate signal from noise. The scar tissue that comes from being wrong, learning why, and building better methods.

These are not soft assets. They are the assets.

AI accelerates early-stage work dramatically. It can draft, classify, summarize, and synthesize in seconds. That is useful. It is also dangerous if it makes weak inputs look authoritative.

The key distinction is this: automation is not authority.

AI is not earnest. It is powerful patterning. It does not carry responsibility. We do.

So treat AI like an associate, not an oracle. Use it to move faster toward the moment where being earnest matters most, the moment you decide what is true enough to act on.

Three simple controls keep AI a multiplier of judgment instead of a substitute for it.

• Provenance: document sources and what the model saw.
• Validation: triangulate with at least one non-AI method, dataset, or expert review.
• Accountability: name an owner who signs off on the claim and owns the consequences.

The Wilde Test for performative insights

If you want a quick way to spot when your function is drifting toward theater, here is a simple check. Call it the Wilde Test.

1. If the chart disappeared, would the conclusion still be true?
2. Are we optimizing for clarity or for applause?
3. What would we advise if we were accountable for the outcome, not the slide?

If the answers feel uncomfortable, that is not a problem. That is a signal.

Earnest leadership is the discipline of choosing truth over performance, especially when speed is rewarded.

Recommendations for leading in earnest through change

Earnestness is not solemn. It is clarity with integrity. It is seriousness with purpose. It is the discipline to resist performance and deliver truth that holds up under scrutiny.

Here are five leader-ready actions you can put into practice.

Make “fit for purpose” your operating system. Define the decision first. Match method and speed to the risk of being wrong. Not every decision requires a gold standard study, but every decision requires honest standards.

Institutionalize pauses. Build slowness into the moments where mistakes are expensive. Pattern review. Bias scan. Second source check. Pre-mortem thinking. Make these normal so speed does not erase credibility.

Publish the chain of evidence. Tell stories with receipts. Make it easy for stakeholders to see how you got there. What you used, what you excluded, what you assumed, and what you need to confirm next.

Protect reputation like capital. Credibility compounds. Treat it as an asset you build deliberately. It is the reason your recommendations travel farther than your slides.

Use AI to expand judgment, not replace it. Let AI accelerate exploration, but keep human accountability explicit. Separate model output from business truth. Require provenance, validation, and ownership.