The history of customer research and the illusion of customer understanding

For decades, B2B companies have been trying to understand their customers—and they’ve been failing in surprisingly consistent ways.

They’ve done punch-card mailers.
They’ve sent surveys.
They’ve tracked scores.
They’ve debated what a “7 out of 10” really means.

And still, most companies are still guessing.

They don’t guess because they don’t care—they guess because those tools were never designed for how humans actually think.

From punch cards to platforms

Customer research didn’t start with modern SaaS tools. It started with something much simpler: structured systems designed to capture feedback at scale. Punch cards, mail-in surveys, and early quantitative methods all shared the same goal—to turn subjective human experiences into something measurable.

These systems were rigid and slow, but they introduced a powerful idea: If you could standardize responses and collect enough of them, you could begin to see patterns. And if you could see patterns, you could make better decisions. That logic made sense at the time, and it became the foundation for everything that followed.

As technology improved, those same principles were carried forward—with better interfaces and faster collection methods.

The rise of surveys—and the promise of scale

When web-based surveys became mainstream, they felt like a breakthrough moment for B2B research. Companies could suddenly gather feedback from hundreds or thousands of customers in a fraction of the time it used to take. 

Data became easier to collect, store, and analyze.

For many organizations, this was transformative. Surveys made it possible to quantify customer sentiment, track trends over time, and bring a level of rigor to decision-making that hadn’t existed before. They became the default tool for understanding the customer voice.

But that progress came with a tradeoff—one that wasn’t immediately obvious.

In addition to making research more efficient, surveys also fundamentally changed how feedback was captured by forcing customers to translate their experiences into predefined structures.

The hidden (and dangerous) assumption behind every survey

Every survey, no matter how well designed, operates on the same underlying assumption: that customers can accurately compress their thoughts, experiences, and emotions into structured responses.

In practice, that means asking people to:

• Rate something on a scale
• Choose from a list of predefined reasons
• Summarize their thinking in a sentence or two

By design, surveys collect data that’s superficial: Studies have shown that people use just five words and only 15 seconds when answering survey questions.

The problem is that real human experiences don’t naturally fit into those constraints.

In B2B especially, decisions are rarely driven by a single factor. They’re shaped by competing priorities, internal dynamics, perceived risks, and incomplete information. When you ask a buyer to select “the primary reason” for a decision, you’re not uncovering truth—you’re asking them to simplify it.

Why this breaks down in B2B

This limitation exists in all research, but it’s amplified in B2B environments.

Unlike B2C, where large datasets can help smooth out inconsistencies, B2B operates with smaller sample sizes and higher stakes. Each deal carries more weight, and each piece of feedback has the potential to influence major decisions across product, sales, and marketing.

That changes the nature of what “good data” looks like.

In B2B, the real value lies in both identifying patterns and understanding the reasoning behind them. A single detailed explanation of why a deal was lost can be more valuable than dozens of surface-level responses.

That’s why many teams begin to feel a disconnect, even when their survey programs appear healthy on paper.

“The surveys were fine as a starting point, but they weren’t driving the full level of depth that we wanted to see in our program.”
—Krysha Nair, VP of Product Marketing at StackAdapt

This experience isn’t unique. Once teams start relying on survey data to inform real decisions—messaging, positioning, sales strategy—they begin to notice the gap between what the data shows and what they need to know. 

The signals are there, but the story behind them is incomplete.

“We had a decent response rate, so we were getting data—but it also felt like there was some information we were missing out on.”
—Jessica Lovell, Senior Product Marketing Manager at Momentive Software

That missing information isn’t more data points. It’s the context behind them.

The “7 out of 10” problem

Consider one of the most common outputs in customer research: a numerical score.

On the surface, it feels precise. A 7 out of 10 is something you can track, benchmark, and report on. It fits neatly into dashboards and executive updates.

But the moment you try to act on it, the limitations become clear.

A 7 out of 10 could mean the customer liked your product but had concerns about pricing. Or it could mean that they trusted your team but ultimately chose a competitor. Or it could reflect internal constraints that had nothing to do with your solution at all.

The number captures sentiment—but it also strips away meaning.

Without context, teams are left to interpret what that score represents. They debate it in meetings, layer on assumptions, and try to connect it to other signals in the data. Sometimes they’re right. Often they’re not.

At that point, decision-making starts to rely less on insights and more on guesswork.

Or more bluntly: It relies on hope.

And hope is not a strategy.

The cost of shallow understanding

This gap between data and understanding has real consequences.

When companies rely on incomplete signals, they often misdiagnose the root causes of their performance. They may attribute a lost deal to pricing when the real issue was positioning, or they might focus on product gaps when the challenge was in the sales process.

Over time, these small misinterpretations compound. Strategies drift. Messaging becomes less effective. Opportunities are missed—and not because the information wasn’t available, but because it wasn’t fully understood.

StackAdapt saw this clearly when they moved beyond surveys and began interviewing lost deals directly. After partnering with Clozd to capture candid, in-depth buyer feedback, they found that 60% of their loss interviews included clear signals for future re-engagement—transforming closed-lost deals into a new, more predictable source of pipeline.

“What that means is that those losses are coming back around, so we have another—and often better—shot at winning that business.”
—Ryan Nelsen, CMO at StackAdapt

In other words, a majority of their “lost” deals weren’t truly lost—they just lacked the insights they needed to act.

That kind of opportunity is nearly impossible to uncover through structured survey responses alone.

What was optimized for—and what was lost

Over time, the evolution of research tools followed a predictable path. We optimized for speed, scalability, and efficiency—because those were the most visible constraints.

And to be clear, those improvements mattered. They made it possible to bring customer feedback into more decisions and across more teams.

But in solving for those constraints, we introduced a new one: We made it easier to collect data—but harder to capture meaning.

That tradeoff is easy to miss because the outputs still look useful. Dashboards fill up. Reports get generated. Metrics move.

But underneath it all, something critical is missing: a clear understanding of what customers are actually thinking.

The uncomfortable truth

Most B2B organizations don’t understand their customers as well as they think they do.

And it’s not because they lack effort or data. In many cases, they’ve invested heavily in both—building robust survey programs, tracking metrics over time, and creating systems designed to capture the voice of the customer.

But those systems were built to measure responses—not to uncover reasoning.

That distinction matters more than it seems. 

When you’re measuring responses, you’re capturing what customers say in a structured format. When you’re uncovering reasoning, you’re trying to understand how they think, what influenced their decisions, and what context shaped their perspective. Those are fundamentally different goals—and they require different approaches.

Over time, this gap creates a subtle but important disconnect. Teams begin to rely on data that feels directionally helpful, even if it’s not fully complete. They can identify trends, but they still find themselves debating what those trends actually mean. They can see outcomes, but the underlying drivers remain partially obscured.

So they fill in the gaps. They layer on assumptions. They rely on past experience. They interpret signals as best they can.

And sometimes, they get it right.

But most of the time they don’t.

What they’ve been capturing are signals—useful, but incomplete—when they actually need insights.

What comes next

For years, companies have accepted a fundamental constraint in research.

You can have scale, or you can have depth.

Surveys give you reach. Interviews give you understanding. But combining the two has always been difficult, if not impossible, to do efficiently. As a result, organizations built their processes around that limitation—using surveys to gather broad input, and reserving interviews for moments when deeper insights were absolutely necessary.

That tradeoff shaped research programs and influenced how teams made decisions, how confident they felt in their conclusions, and how quickly they were able to act. In many cases, it created an environment where data was abundant, but clarity was still hard to come by.

And because that constraint felt structural—something inherent to the nature of research—it largely went unchallenged.

But that’s starting to change.

The question is no longer whether companies have to choose between scale and depth.

It’s what happens when they don’t.

‍