Mixed Methods Case Studies: 5 Real-World Examples
Theory is useful, but seeing mixed methods research applied to actual business problems is what makes the methodology stick. These five case studies show how teams used different mixed methods designs to answer questions that single-method approaches couldn't handle.
Each case follows the same structure: the problem, the design chosen, how data was collected and integrated, and what the team learned that they wouldn't have found otherwise.
Case Study 1: SaaS Feature Prioritization
Company type: B2B project management software Design used: Sequential explanatory Timeline: 6 weeks
The Problem
The product team had a backlog of 15 feature requests and needed to prioritize the next quarter's roadmap. They'd previously relied on customer vote counts from their feedback portal, but those votes skewed heavily toward power users and didn't represent the broader customer base.
The Approach
Phase 1 (Weeks 1-3): A MaxDiff study with 400 customers across all plan tiers. Participants evaluated the 15 features in randomized sets, producing a statistically valid preference ranking.
Phase 2 (Weeks 4-6): 12 interviews with customers who represented three distinct patterns in the MaxDiff data: those who strongly preferred collaboration features, those who prioritized automation features, and those whose preferences didn't cluster clearly.
The Integration
The MaxDiff data showed that two features, automated status updates and shared workspaces, ranked nearly identically at the top. A vote count would have called it a tie.
The interviews broke the tie. Customers who wanted automated status updates described using the product for cross-team coordination. They needed the feature because their current workaround (manual Slack updates) was consuming hours each week. Customers who wanted shared workspaces described wanting the feature but couldn't articulate a specific workflow it would enable.
The finding single methods missed: Automated status updates solved a concrete, daily pain. Shared workspaces reflected an aspirational desire. The team prioritized status updates and saw a measurable retention lift within two months.
Case Study 2: Healthcare Patient Experience Redesign
Company type: Regional hospital network Design used: Concurrent triangulation Timeline: 8 weeks
The Problem
Patient satisfaction scores had declined for three consecutive quarters, but the scores didn't indicate where in the patient journey the problems were occurring. The hospital needed both breadth (which touchpoints had the biggest impact) and depth (what the experience actually felt like).
The Approach
Quantitative strand: A survey sent to 1,200 patients who had visited in the past 90 days, measuring satisfaction across 12 touchpoints on a 7-point scale, plus overall NPS.
Qualitative strand: 20 patient journey interviews conducted simultaneously, where patients walked through their most recent visit from start to finish.
Both strands were analyzed independently before comparison.
The Integration
The survey identified three touchpoints with the lowest scores: billing communication (3.2/7), wait time before initial consultation (3.8/7), and discharge instructions (4.1/7).
The interviews revealed something the survey couldn't: these three touchpoints weren't independent problems. Patients described a pattern of information uncertainty, not knowing what to expect at each stage. The wait felt longer because nobody told them how long it would be. Billing was confusing because the cost wasn't discussed during the visit. Discharge instructions were unclear because they contradicted what the doctor said verbally.
The finding single methods missed: The survey would have sent the team chasing three separate problems. The interviews revealed a single root cause, communication gaps, that could be addressed with one systemic intervention instead of three isolated fixes.
Case Study 3: Consumer Brand Repositioning
Company type: Direct-to-consumer skincare brand Design used: Convergent design Timeline: 10 weeks
The Problem
The brand was considering repositioning from "affordable skincare" to "ingredient-transparent skincare." They needed to understand whether the repositioning would resonate with their existing customers and attract new ones, and what "ingredient transparency" actually meant to consumers.
The Approach
Quantitative strand: A conjoint analysis with 500 respondents testing brand positioning attributes (price emphasis, ingredient focus, dermatologist endorsement, sustainability claims) across different combinations.
Qualitative strand: 4 focus groups (8 participants each) discussing skincare purchasing decisions, what "transparency" means in the beauty context, and reactions to competitor positioning.
Both datasets were analyzed independently, then merged using a joint display organized by positioning attribute.
The Integration
The conjoint showed that "ingredient transparency" as a positioning attribute increased purchase intent by 18% among health-conscious segments but had no effect on price-sensitive segments.
The focus groups added critical context. Participants defined "transparency" differently by age group. Younger participants (25-34) meant detailed ingredient sourcing and environmental impact. Older participants (35-50) meant clear labeling and clinical evidence. Both groups responded to the word "transparency," but they wanted fundamentally different things.
The finding single methods missed: The conjoint said "transparency works." The focus groups said "but only if you define it correctly for each segment." The brand launched with segment-specific messaging and saw a 22% increase in new customer acquisition within the first quarter.
Case Study 4: Employee Engagement After Organizational Change
Company type: Mid-size technology company (800 employees) Design used: Embedded design Timeline: 4 weeks
The Problem
Six months after a reorganization that merged two departments, leadership wanted to measure how the transition was going. They had budget for a comprehensive survey but not for a separate qualitative study.
The Approach
Primary strand: An 80-item employee engagement survey covering satisfaction, clarity of role, team dynamics, and confidence in leadership.
Embedded strand: Five open-ended questions placed at strategic points in the survey: after the team dynamics section, after the leadership confidence section, and at the end. Questions like "Describe one thing that's working well since the reorg" and "What's the biggest unresolved issue from the transition?"
The Integration
The quantitative data showed overall engagement at 3.6/5, acceptable but below the pre-reorg baseline of 4.0. The biggest drop was in "clarity of role" (from 4.2 to 3.1) and "team dynamics" (from 4.0 to 3.3).
The embedded open-ends explained both drops. Employees from the acquired department described feeling like "guests in someone else's house", their processes had been replaced wholesale with the other team's processes, and nobody had asked for their input. Clarity of role was low not because job descriptions were unclear, but because employees didn't know whose norms and methods took priority.
The finding single methods missed: The survey alone would have triggered a "clarify roles and responsibilities" initiative. The qualitative data revealed the real issue was cultural integration and employee voice. Leadership shifted from a top-down role clarification exercise to a collaborative process redesign.
Case Study 5: Pricing Model Validation for a Subscription Product
Company type: B2C fitness app Design used: Sequential explanatory Timeline: 5 weeks
The Problem
The app was testing a switch from a single subscription tier ($14.99/month) to a freemium model with three tiers. They needed to validate the pricing structure and understand how existing subscribers would react.
The Approach
Phase 1 (Weeks 1-3): Conjoint analysis with 600 respondents (300 current subscribers, 300 non-subscribers) testing combinations of features, price points, and tier structures.
Phase 2 (Weeks 4-5): 10 interviews with current subscribers whose conjoint responses suggested they'd downgrade to a lower tier, and 5 interviews with non-subscribers whose responses showed interest in the free tier.
The Integration
The conjoint modeled an optimal three-tier structure and predicted that 35% of current subscribers would downgrade. That was a concerning revenue projection.
The interviews with potential downgraders told a more nuanced story. Most weren't motivated by price savings. They were frustrated that they were paying for features they didn't use and felt the current single tier was "wasteful." When shown the mid-tier option, several said they'd prefer it and feel better about their subscription.
The non-subscriber interviews revealed that the free tier needed just two specific features to drive signups. More than that created choice paralysis.
The finding single methods missed: The conjoint predicted revenue loss from downgrades. The interviews revealed that downgrades would actually increase satisfaction and reduce churn, net positive for lifetime value. The team launched the freemium model with confidence and saw churn drop by 15% in the first six months.
Patterns Across These Cases
A few themes emerge:
Quantitative data identifies the "what." Qualitative data reframes the problem. In every case, the qualitative strand didn't just add detail, it changed how the team understood the quantitative findings.
Integration is where the value lives. The individual datasets in each case would have been useful on their own. But the integrated findings led to better decisions than either strand alone would have produced.
The right design matches the question. Sequential explanatory worked when the team needed to explain surprising quant results. Concurrent triangulation worked when they needed validation from two angles. Embedded worked when budget was tight but context was essential.
Platform integration matters operationally. Several of these studies involved quant and qual data in the same system. Teams that had to manage separate tools spent more time on logistics and less on analysis.
Running Your Own Mixed Methods Study
If these cases resonate with the questions your team is facing, Quali-Fi gives you the infrastructure to run studies like these. Surveys, MaxDiff, conjoint, focus groups, and IDIs all live on one platform, so you can move from quant to qual (or run them in parallel) without switching tools or exporting data.
For help choosing which design fits your situation, use our decision framework. For the technical details of connecting your datasets, see our guide on combining qual and quant data.
Start your mixed methods study on Quali-Fi
FAQs
Are these mixed methods case studies based on real companies?
These cases are composites based on common patterns across real mixed methods research projects. The specific companies are fictionalized, but the research designs, findings, and integration approaches reflect actual practice in product, marketing, and organizational research.
Which mixed methods design is most commonly used in business research?
Sequential explanatory design is the most common in business contexts because it follows a natural logic: run the numbers first, then talk to people to understand what the numbers mean. Embedded design is also very common, especially when teams add open-ended questions to existing surveys.
How long do these studies typically take?
The cases in this article ranged from 4 to 10 weeks. Most business mixed methods studies fall in the 4-12 week range, depending on the design. Embedded designs are fastest (2-4 weeks), sequential designs take 5-10 weeks, and concurrent or convergent designs take 6-12 weeks.
Do I need a large sample size for mixed methods research?
Not necessarily. The quantitative strand needs enough participants for statistical confidence (typically 150-500 for surveys, depending on the analysis). The qualitative strand needs enough for thematic saturation (typically 10-30 for interviews, 3-5 for focus groups). See our sampling guide for detailed recommendations.
What's the biggest mistake teams make with mixed methods?
Skipping the integration step. Teams collect both types of data but present them separately, a survey summary and an interview summary in the same deck. The value of mixed methods comes from systematically comparing, connecting, or merging the two datasets. Without integration, you've done two studies, not one mixed methods study.