The survey results were unambiguous. Seventy-eight percent of customers rated ice dispensers as “very important” when purchasing refrigeration units. The product team celebrated. Their emphasis on dispenser technology was validated. The engineering investment was justified.
I asked a different question: how many customers actually use the dispensers regularly?
The answer, buried in usage data no one had examined: thirty-eight percent.
That forty-point gap between stated importance and actual behavior represented one of the largest opportunities I’ve ever uncovered in a transformation. Customers said they wanted dispensers because they’d always said they wanted dispensers. It was the expected answer. But their behavior revealed a different truth—a truth that opened a $47 million opportunity in non-dispenser units that competitors had ignored.
This is the difference between customer research that asks and customer research that understands. Most companies do the former. The ones that win do the latter.
The Fatal B2B Mistake
Business-to-business companies make a predictable error: they study their direct customers while ignoring the end users whose needs actually drive demand.
The logic seems reasonable. Your customer is the business that buys from you. You should understand their needs, their buying criteria, their decision-making process. Customer research means researching the customer.
But in B2B contexts, your direct customer is often an intermediary. They’re a retailer, a distributor, a contractor, an OEM. Their needs are shaped by their customers’ needs. And their customers’ needs are shaped by end users who may be several steps removed from your transaction.
The refrigeration manufacturer selling to appliance retailers isn’t really serving those retailers. They’re serving the homeowners who will ultimately use the product. The retailers are a channel. Understanding the channel without understanding the end user produces strategies optimized for the wrong audience.
Harvard Business Review’s work on jobs-to-be-done demonstrates that customer needs exist independent of the products that serve them. End users have jobs they’re trying to accomplish. Your product either helps accomplish those jobs or it doesn’t. The intermediaries in your distribution chain are largely irrelevant to that fundamental question.
The dispenser insight came from going around our direct customers—the retailers—to study actual homeowners. Two hundred interviews, conducted in homes, observing real kitchens, watching actual behavior. What we learned contradicted everything our retail partners told us, because our retail partners were passing along assumptions they’d never tested either.
The TCO Intelligence Framework
Customers don’t buy products. They buy solutions to problems. The solution’s value isn’t the purchase price—it’s the total cost of ownership across the entire lifecycle.
Most customer research focuses on purchase criteria: features, specifications, price point, brand preference. This research is nearly useless for understanding true customer economics.
The TCO Intelligence Framework maps three cost layers:
Direct Costs These are the visible costs customers consciously evaluate: purchase price, installation, training, warranty. Direct costs are the easiest to research and the least differentiating to understand. Every competitor gathers this information.
Indirect Costs These costs are real but often unattributed: energy consumption over product lifetime, maintenance requirements, supplies and consumables, space and infrastructure needs. Indirect costs frequently exceed direct costs but receive fraction of the attention in purchase decisions.
In one analysis, we discovered that a competing product with a 15% lower purchase price had 40% higher energy costs over its operational life. Customers choosing on price were making economically irrational decisions—but no one had presented the comparison in terms they could evaluate.
Hidden Costs These costs are invisible even to customers who experience them: productivity lost during downtime, opportunity costs of limited capability, management attention required for problem products, reputation risk from unreliable performance.
Hidden costs are the hardest to quantify and the most powerful to expose. The customer who understands that a “cheaper” product actually costs more—when hidden costs are included—becomes a customer who buys on value rather than price.
MIT Sloan research on customer analytics demonstrates that companies understanding total cost of ownership outperform competitors by significant margins. They can price on value rather than competing on cost. They can justify premium positions with economic proof. They can help customers make better decisions that strengthen relationships.
Building TCO Intelligence requires research methods that go beyond surveys. You need to observe customer operations, access their financial data where possible, and map the complete economic picture they experience.
Pain Point Archaeology
Customers can tell you what they want. They often can’t tell you what they need.
This isn’t because customers are stupid. It’s because they’ve adapted to current limitations. The pain they experience daily has become background noise. The workarounds they’ve developed feel like normal operations. The problems they’ve given up trying to solve have been mentally reclassified as unchangeable constraints.
Pain Point Archaeology digs beneath surface-level feedback to uncover needs customers themselves can’t articulate.
The Five-Why Method
When a customer expresses a preference or complaint, ask why. Then ask why again. And again. Five levels of “why” typically reach root causes invisible at surface level.
Customer: “We need faster delivery.” Why? “Because we run out of inventory.” Why? “Because we can’t predict demand accurately.” Why? “Because our forecasting system doesn’t have good data.” Why? “Because we don’t know what’s selling until it’s already sold.” Why? “Because our point-of-sale system doesn’t integrate with ordering.”
The surface request was faster delivery. The root need is demand visibility. These require completely different solutions—and the customer asking for faster delivery has no idea their real problem is data integration.
Compensating Behavior Analysis
Watch what customers do, not just what they say. Compensating behaviors—workarounds, adaptations, manual interventions—reveal unmet needs customers have stopped recognizing.
The operator who always checks a setting twice before starting a machine is compensating for an unreliable default. The manager who maintains a shadow spreadsheet is compensating for inadequate reporting. The customer who orders extra inventory “just in case” is compensating for unpredictable quality.
Each compensating behavior represents a pain point so normalized that customers don’t think to mention it. Observational research—watching work happen in customer environments—reveals what interviews never will.
The Frustration Diary
Ask customers to document frustrations in real-time over a week or month. What made them curse under their breath? What required extra effort? What went wrong and required recovery?
The frustration diary captures incidents that surveys miss because customers forget them. The daily irritation that’s not severe enough to remember when asked about “problems” accumulates into significant opportunity for whoever solves it.
Nielsen and Qualtrics have published extensively on research methodologies that capture customer needs beyond surface preferences. The consistent finding: observational and behavioral methods reveal opportunities that direct questioning misses.
Conversion Death Point Analysis
Somewhere in your customer journey, prospects fail to convert. They enter your pipeline and then disappear. They engage with your product and then stop. They consider purchase and then choose alternatives.
Conversion Death Points are where opportunities go to die. Finding them requires tracking the complete journey and identifying exactly where failure occurs.
Journey Mapping
Map every step from initial awareness through purchase decision and ongoing usage. Not the theoretical journey—the actual journey customers experience. Where do they enter? What do they encounter? Where do they drop off?
Digital journeys are easier to track—web analytics reveal where visitors abandon. Physical and relationship-based journeys require customer interviews and observation.
Failure Mode Analysis
At each death point, ask: why do prospects fail here? What would have to be different for them to continue?
Common failure modes include:
Information gaps: prospects can’t find answers to questions that block progress. Effort barriers: the next step requires more work than prospects are willing to invest. Trust deficits: prospects don’t have enough confidence to proceed. Value uncertainty: prospects can’t determine whether the offering is worth the cost. Competitive alternatives: prospects find easier or better options elsewhere.
Each failure mode suggests different interventions. Information gaps need content. Effort barriers need simplification. Trust deficits need proof. Value uncertainty needs demonstration. Competitive losses need differentiation.
Recovery Testing
Once you identify death points, test interventions. If prospects abandon at pricing pages, test different price presentations. If they disengage after initial meetings, test different follow-up approaches. If they choose competitors, test different positioning.
Conversion improvement often generates more value than customer acquisition. The company converting 10% of qualified leads has dramatically different economics than the company converting 5%—and the improvement requires understanding where and why failure occurs.
Customer Job Analysis
Clayton Christensen’s jobs-to-be-done framework, developed at Harvard Business School and advanced by the Christensen Institute, provides the theoretical foundation for deep customer understanding.
Customers don’t buy products. They hire products to do jobs. The job exists independent of any product—and understanding the job reveals opportunities invisible when you focus on the product.
Functional Jobs
These are the practical tasks customers need to accomplish. The homeowner buying a refrigerator needs to preserve food. The manufacturer buying equipment needs to produce output. The business buyer purchasing software needs to complete processes.
Functional jobs are usually obvious but often too broadly defined. “Preserve food” doesn’t distinguish between different preservation needs—fresh ingredients for daily cooking versus bulk storage for large families versus specialized requirements for particular items.
Decomposing functional jobs into specific sub-jobs reveals differentiation opportunities. The refrigerator optimized for fresh produce storage serves a different job than the refrigerator optimized for frozen bulk storage, even though both “preserve food.”
Emotional Jobs
Beyond functional outcomes, customers have emotional needs products can serve. They want to feel confident in their choices. They want to avoid anxiety about reliability. They want to impress colleagues or satisfy family members.
Emotional jobs often drive decisions more than functional jobs, especially when functional alternatives are comparable. The business buyer choosing between equivalent solutions often chooses the one that makes them feel safest—the vendor unlikely to create problems that reflect poorly on their judgment.
Understanding emotional jobs explains preferences that functional analysis cannot. Why do customers choose more expensive options when cheaper alternatives perform identically? Often because the expensive choice makes them feel more secure.
Social Jobs
Products also serve social functions. They signal status, demonstrate values, and shape how others perceive the customer.
The executive who buys premium equipment is partly buying functionality and partly buying the signal that premium choice sends about their organization. The consumer who chooses sustainable products is partly buying performance and partly buying the identity that sustainability represents.
Social jobs explain brand premiums, feature preferences, and choices that seem irrational from pure functionality perspective.
Stated Preferences vs. Revealed Behavior
The dispenser insight—78% stated importance, 38% actual usage—illustrates a fundamental research challenge: what people say and what they do frequently diverge.
Survey research captures stated preferences. These are what customers believe they want, what they think they should want, or what they’ve been conditioned to expect. Stated preferences are influenced by survey design, social desirability, and respondent fatigue.
Behavioral research captures revealed preferences. These are what customers actually choose when facing real decisions with real consequences. Revealed preferences reflect true priorities, not hypothetical ones.
The divergence between stated and revealed preferences creates both risk and opportunity. The risk: building strategy on stated preferences that don’t predict behavior. The opportunity: finding gaps between what customers say and what they do that reveal unmet needs.
Usage Data Analysis
For products with usage tracking, compare stated preferences against actual usage patterns. Features customers say they value but rarely use are candidates for elimination. Features customers use heavily but don’t mention in surveys are candidates for emphasis.
Choice Analysis
When customers make actual decisions, which factors predict their choices? Regression analysis on historical decisions often reveals that stated priorities (quality, features, brand) matter less than unstated factors (convenience, relationship, habit).
A/B Testing
When possible, test alternatives rather than asking preferences. The email subject line customers say they prefer may generate lower open rates than the subject line they didn’t choose. The product configuration customers rank highest may convert worse than lower-ranked options.
Behavioral evidence trumps survey data. When they diverge, trust behavior.
The 200-Interview Revelation
In the refrigeration transformation, we conducted two hundred in-depth interviews with homeowners—not our retail customers, but the end users whose needs ultimately determined success.
We didn’t ask what features they wanted. We watched how they used their kitchens. We observed where they stored things and why. We documented the problems they’d adapted around so thoroughly they’d forgotten they were problems.
The interviews revealed:
Ice dispensers were used regularly by only 38% of households despite nearly universal stated importance. The assumption that dispensers were essential had never been tested against behavior.
Water filtration mattered more than any executive had realized. Households with filtered dispenser water used dispensers regularly. Those without filtration avoided them due to taste concerns.
Door configuration—how refrigerator doors opened relative to kitchen layouts—created daily friction in many households that no one mentioned in surveys because they’d accepted it as normal.
Counter depth—whether refrigerators protruded beyond counter lines—was an aesthetic priority driving purchase decisions for a significant segment that our feature-focused research had missed entirely.
Each insight suggested product opportunities competitors were missing because they relied on the same survey data everyone else used. The two hundred interviews—qualitative, observational, behavioral—generated more actionable insight than years of quantitative surveys.
Building Customer Intelligence Capability
Deep customer research isn’t a project. It’s a capability that requires ongoing investment.
Regular Customer Immersion
Executives and product developers should spend time with customers regularly—not in conference rooms, but in customer environments. Watch customers work. Observe usage in context. Experience the problems customers experience.
Some companies mandate customer visits quarterly. Others rotate staff through customer-facing roles. The mechanism matters less than the outcome: continuous exposure to customer reality that prevents drift into internal assumptions.
Behavioral Data Infrastructure
Where possible, instrument products to capture usage data. Connected products enable continuous behavioral research at scale. Usage patterns reveal priorities surveys miss.
Even without connected products, behavioral data exists. Customer service records, return patterns, warranty claims, support ticket content—all provide behavioral signal about customer experience.
Research Methodology Diversity
Different questions require different methods. Surveys for broad patterns. Interviews for depth. Observation for behavior. Experiments for causation.
Companies that rely exclusively on any single method develop blind spots. The survey-only company misses behavioral truth. The observation-only company lacks scale. Methodology diversity creates comprehensive understanding.
Cross-Functional Integration
Customer insight should inform every function—product development, marketing, sales, operations, service. Siloed research that lives in marketing and never reaches engineering is waste.
Build processes that distribute customer understanding across the organization. Share research broadly. Require customer evidence in strategic discussions. Create cultures where customer reality checks internal assumptions.
The Magnificent Obsession Mindset
The phrase “Magnificent Obsessions” reflects what deep customer understanding requires: obsessive curiosity about customer reality that refuses to accept surface-level answers.
Most companies research customers. Few obsess over understanding them. The difference shows up in products that solve problems customers didn’t know they had, in positioning that resonates at emotional levels, in relationships that competitors can’t replicate.
The dispenser insight wasn’t found because someone ran a better survey. It was found because someone asked a different question and refused to accept the easy answer.
Customer research that goes beyond surveys and demographics requires that obsessive orientation. It requires watching behavior, not just asking questions. It requires digging beneath stated preferences to reveal underlying needs. It requires tracking complete journeys, analyzing total costs, and understanding jobs at functional, emotional, and social levels.
The companies that develop this capability see opportunities others miss. The companies that don’t wonder why their research never produces differentiation.
The customers are telling you what they need. Just not in surveys. Not in words. In behavior, adaptation, workaround, and choice.
The question is whether you’re obsessed enough to listen.
Todd Hagopian is the founder of https://stagnationassassins.com, author of The Unfair Advantage: Weaponizing the Hypomanic Toolbox, and founder of the Stagnation Intelligence Agency. He has transformed businesses at Berkshire Hathaway, Illinois Tool Works, and Whirlpool Corporation, generating over $2 billion in shareholder value. His methodologies have been published on SSRN and featured in Forbes, Fox Business, The Washington Post, and NPR. Connect with Todd on LinkedIn or Twitter.