The Virtue of a Question

Quark Insights - Asking Better Questions

Asking questions are an essential and fundamental function of producing meaningful analyses. Taking time to channel curiosity, even with what may seem to be a “stupid question”, may reveal something critical that removes a blind spot. Asking good questions is thoughtful and a bit of an art. This week, I explore how to release your inner interrogator.

How Questions Impact Insights

Good research is not about having answers, it is about asking better questions.

Early in my career, I learned that the questions I hesitated to ask were usually the ones everyone else needed answered. That realization changed how I approached every meeting, dataset, and client conversation that followed, because I stopped treating questions as a sign of weakness and started treating them as a tool for clarity. After all, where do you acquire the knowledge to ask better questions if you haven't studied Sociology and your next gig isn't as a focus group moderator?!

A question, in a research or business context, is a focused prompt that defines what you are trying to understand or decide. It matters because every strategy, analysis, or presentation is built around a question, whether you say it out loud or not, and that hidden question quietly steers what you look at and what you ignore.

In CTV and streaming work, this shows up all the time. A brand might say, “Did our streaming campaign work?” and the team quietly translates that into “Did it hit our planned reach and frequency?” If the brand’s real concern is whether streaming helped them connect with a new audience or support upper-funnel perception, that silent translation can send the analysis down the wrong path. The original question was too fuzzy; the silent question was too narrow. Neither was truly aligned with the business need.

Questions do two things:

• They expose assumptions.

• They define the path to insight.

If you start with weak or incomplete questions, even the best data will lead you in the wrong direction, because you will be busy answering something that doesn’t actually matter to the business. But when you ask the right question, the analysis tends to organize itself around it; the filters you choose, the comparisons you run, and the way you tell the story all line up behind a clear objective.

Imagine a streaming-first brand that aligns everyone on one core question: “Where did this CTV campaign create meaningful differences in how viewers engaged with our brand compared to our other video channels?” Suddenly, all the noise drops away. Every chart, every cut of the data, every takeaway either supports or refines that question. You have a spine to build the narrative around.

The Two Types of Questions in Insights

Not all questions do the same job. In practice, they fall into two distinct categories that you can consciously use as tools instead of letting them blur together.

1. Confirming Questions (What You Think You Know)

A confirming question is a targeted, often closed-ended prompt designed to validate or refute a specific hypothesis or assumption. These questions matter in business because decisions almost always rest on assumptions, about customers, performance, or markets, and confirming questions either support those assumptions or expose where they are off.

In the CTV/streaming world for example, you might encounter:

• “Did our streaming campaign add reach beyond our linear buy?”

• “Did viewers exposed in CTV environments stay with the ad longer than in social video?”

• “Did our premium publisher mix help us avoid overserving the same viewers?”

These questions are efficient. They help you prove or disprove something quickly, which is crucial when executives are weighing options under time pressure and need a clear go/no-go decision. But they are limited, they only operate within the boundaries of what you already suspect, which means they rarely surface new opportunities or risks on their own.

Here’s a hypothetical that keeps it safely conceptual. Suppose a retail brand has a working belief that “CTV is where we reach high-value viewers.” Internally, that belief has been repeated so often it feels like fact. When you build a measurement plan, you intentionally write confirming questions into your framework: “Do the CTV-exposed viewers show the engagement patterns the brand associates with ‘high value’ compared to other channels?” “Are those behaviors present in all streaming environments, or only certain types of inventory?” You are not assuming any outcomes; you are designing the analysis so that if the assumption holds up, you can show how, and if it doesn’t, you can show where the story needs to change.

In that sense, confirming questions act like a pressure test. They clarify whether the stories people tell themselves about CTV and streaming performance have enough support to stay in the playbook, or need to be rewritten before they drive another quarter’s worth of decisions.

2. Exploratory Questions (What You Don’t Know Yet)

An exploratory question is an open-ended prompt designed to uncover patterns, possibilities, or problems that are not yet defined. It matters in business because growth, innovation, and risk mitigation almost always come from things you did not see coming, and exploratory questions are how you deliberately go looking for those.

In CTV and streaming, exploratory questions sound like:

• “What patterns are we seeing across different audience segments and streaming platforms?”

• “Where do we appear to be over-delivering impressions, and where might we be under-investing?”

• “What is surprising or counterintuitive in how viewers are interacting with our creative in streaming environments?”

These questions are where new insight lives. They help you uncover blind spots, unexpected behaviors, and opportunities you were not initially looking for, which is often where the most valuable strategic shifts originate. They tend to generate longer, richer answers, and they require more listening and genuine curiosity to use well.

Take a hypothetical direct-to-consumer brand that has been running CTV and streaming video campaigns for a few quarters. Instead of going straight to, “Did we meet our campaign benchmarks?” you sit down with the data and ask exploratory questions: “Which audience clusters seem to engage differently with our brand in streaming environments versus other video channels?” “What happens when we look at day of week or time of day in relation to how viewers interact with our creative?” You are not claiming any particular outcome, you are deliberately opening the door to see what the data has to say before you decide what matters.

As you explore, you may notice patterns that don’t fit the brand’s existing narrative, certain creative variations behaving differently across platforms, or specific audience slices responding in ways nobody anticipated. Even in a purely hypothetical scenario, this illustrates how exploratory questions serve a different purpose than confirming ones: they help you discover what you didn’t think to ask yet.

How to Use Both, In Sequence

The best researchers do not choose one type over the other, they use both, in order, as part of a deliberate questioning strategy. You can think of it as moving from curiosity to confirmation, using each type of question to play a specific role in the analysis.

Think of it as a two-step process:

Step 1: Start Exploratory

Begin wide. Understand the landscape before narrowing in so you are not locking onto the first obvious pattern and treating it as the whole story. Starting exploratory matters in business because it keeps you from prematurely answering the wrong question and missing subtle but critical signals, like emerging segments or shifts in how viewers use streaming.

In a CTV context, this might look like a brand that just shifted a substantial portion of its budget from linear to streaming for the first time. When the campaign wraps, your first pass at the data is guided by exploratory prompts: “What stands out across audiences, platforms, or dayparts?” “Are there any creative, placement, or environment combinations that behave differently than the rest of the campaign?” You are not starting from performance claims; you are starting from curiosity.

That early exploration gives you a map. You begin to see where the campaign feels consistent and where it feels unusual, where the brand’s existing assumptions seem aligned with the patterns and where something new might be going on. None of this depends on specific numbers, it depends on the way you frame the questions.

Step 2: Move to Confirming

Once you spot a pattern, shift to validation. This matters because businesses do not act on hunches, they act on evidence, and confirming questions stress-test your early observations so they can hold up under scrutiny. It is the difference between “We noticed something interesting” and “We noticed something interesting that is strong enough to influence what we do next.”

Continuing the hypothetical, say your exploratory pass suggests that certain streaming apps appear to behave differently for a particular audience cluster. You then move into confirming mode: “Is this difference consistent across multiple cuts of the data, or does it appear only in a narrow view?” “If we reframe the question slightly, does the pattern still show up, or does it fade?” Again, you are not asserting outcomes; you are showing how the sequence of questions changes the quality of thinking.

This two-step dance, explore, then confirm, creates a loop where curiosity opens doors and discipline decides which doors are worth walking through. Over time, you can make it part of your default workflow: every analysis begins with a handful of exploratory questions, followed by a clear set of confirming questions that support the recommendations you ultimately make.

A Practical Way to Sharpen Your Questions

If you want to improve quickly, use this simple habit that builds questioning discipline into your everyday work.

Before analyzing anything, write down:

• One question you think you are answering.

• One question you are not sure how to answer yet.

A “question you think you are answering” is usually a confirming question, it reflects the stated objective of the project or the metric you have been asked to report on. A “question you are not sure how to answer yet” is almost always exploratory, it points toward the edges of the problem, the places where you feel a little uncertain or curious.

In a streaming advertising setting, the “think you are answering” question might be something like, “Did our CTV campaign support our brand objectives in the way we intended?” That keeps you honest about reporting against the brief. The “not sure how to answer yet” question might be, “What did this campaign suggest about how different types of viewers experience our brand in streaming environments compared to other channels?” That second question pushes you to look beyond the obvious and consider the campaign as a learning opportunity, not just a performance checkpoint.

Over time, you will notice something important: the second question, the uncertain one, is usually where the real value is. That is where you discover nuances about viewer behavior, channel dynamics, or creative impact that were not in the original brief but make your work meaningfully more useful. Even when you stay in hypothetical territory, just naming that second question in your own notes changes how you use the data.

For example, imagine you are preparing a quarterly insights deck for a streaming-heavy brand. Your first question anchors you “How did our CTV and streaming campaign behave relative to what we planned?” Your second question invites exploration,“What did we learn that should change how we think about streaming as part of our broader mix?” You haven’t committed to specific numbers, but you have committed to a richer level of thinking.

The Real Skill: Knowing What the Moment Requires

Strong researchers are not just good at asking questions, they are good at asking the right type of question at the right time. Over time, this becomes a judgment call you make almost instinctively, based on who you are talking to and what decision is on the table.

• If a stakeholder needs proof, lead with confirming questions.

• If the problem is unclear, lead with exploratory ones.

• If the data feels too clean or too obvious, that is usually a signal to dig deeper with open-ended inquiry.

In practice, every conversation has a different job: some are about deciding, some are about discovering, and some are about aligning. Matching your question type to that job keeps meetings focused and makes your analysis feel relevant rather than theoretical.

Picture three hypothetical moments in a single planning cycle for a streaming advertiser.

In a leadership review, the room wants confidence that the streaming strategy is directionally sound, you lean into confirming questions that clarify whether the plan matches the stated goals. In a strategy workshop for a new CTV initiative, nobody is sure yet how success should be defined, you default to exploratory questions that surface possible definitions and help the group see options. In a working session with the creative team, the results look straightforward at first glance, you bring in open-ended questions that ask, “What might we be missing here?” and invite the team to look beneath the surface.

The goal is not to ask more questions. It is to ask questions with intent, questions that clarify objectives, challenge assumptions, and open room for discovery where it matters most. Because in the end, better questions do not just lead to better answers, they lead to better thinking. And better thinking is what turns data into insight, and insight into decisions people feel confident acting on. As the saying goes, you don't know what you don't know. Better to hedge your bets and think about what you're trying to accomplish and work towards more qualified questions, and it may take asking more questions of your client and team to get to those.

Need Some Help?

If your team is juggling messy datasets, tight deadlines, and constant requests, Quark Insights helps you turn everyday analysis into clear stories that drive decisions. Contact me today at LisaC@Quark-Insights.com!

Explore more articles, tools, and templates at the Quark Shop!