Where Discovery Interviews and Marketing Research Are the Same Conversation

Last year I watched a product manager and a marketing lead sit in the same customer interview. The customer was a mid-market operations director, describing how she chose their product over a competitor. The PM’s notes: the integration pain points, the specific workflows the product didn’t support, the three features she wished existed. The marketer’s notes: the language she used to describe the problem (“I needed something that just worked with our ERP”), the competitive alternatives she considered, the moment in the buying process where she almost chose the competitor. Neither person was wrong. Both were doing their job. And when the interview was over, they went back to their respective teams, filed their notes in their respective systems, and never spoke about the same conversation again.

Six months later, the product team had validated three new features based on discovery interviews that included this one. The marketing team had updated the positioning based on customer research that included the same interview. The features and the positioning were coherent — they happened to describe the same customer reality from different angles. But nobody in the building knew that, because the PM and the marketer had never compared notes. The product team built what the customer needed. The marketing team described what the customer wanted to hear. The overlap was accidental, and the missed overlaps were too.

This is the pattern I want to name and fix: discovery interviews (product) and marketing research (customer interviews, win/loss, persona work) are asking the same questions from different angles — but the org chart separates them, and the separation costs both functions the signal they can’t get alone. The PM’s discovery produces validated needs. The marketer’s research produces positioning language and segment understanding. When the two don’t talk, the product builds the right thing and markets it wrong, or markets the right story and builds the wrong thing, or — most commonly — both functions spend double the customer contact budget to learn half the total picture.

The same questions, different vocabularies

The overlap between product discovery and marketing research isn’t superficial. It’s structural. Both functions are asking the same four questions, just with different names on them:

What job is the customer hiring this product to do? Product calls this Jobs to Be Done — the functional, emotional, and social progress the customer is trying to make. Marketing calls this “customer motivation” or “buyer persona” or “pain point.” The vocabulary differs; the question is identical. When the PM asks “what were you trying to accomplish?” and the marketer asks “what drove you to look for a solution?”, they’re asking the same question with different emphasis — the PM emphasizing the job, the marketer emphasizing the trigger. Both framings are useful, and the combination is more complete than either alone.

What alternatives did the customer consider, and why? Product calls this competitive analysis or opportunity mapping. Marketing calls this win/loss analysis or competitive intelligence. The PM wants to know what the customer’s workflow looked like before the product — the competitive alternatives that April Dunford’s framework starts with. The marketer wants to know what the customer compared the product to during the buying process. Both are asking “what else could have solved this problem?” — the PM from the demand side, the marketer from the competitive landscape side. The customer’s answer contains both, and extracting both from one conversation is twice as efficient as extracting each from separate ones.

What language does the customer use to describe the problem? This is marketing’s bread and butter — the exact phrases customers use when they describe their pain, which become the copy on the landing page, the subject line of the outreach email, the framing in the sales deck. Product discovery also captures this language, but PMs tend to paraphrase it into product vocabulary — “the customer needs a better export flow” instead of “I need to get this data into the spreadsheet my manager reviews.” The paraphrase is useful for engineering; the original language is useful for positioning. When the PM preserves the verbatim and shares it with marketing, the marketing team gets positioning input that no focus group can produce, because it comes from a customer describing their real problem in their real words, with no awareness that they’re generating copy.

What would make the customer switch? Product calls this churn risk or the forces of progress — the push, pull, anxiety, and habit forces that govern behavior change. Marketing calls this “switching triggers” or “competitive displacement drivers.” Both are asking: what would have to be true for this customer to change their behavior? The PM wants the answer to inform the roadmap — which opportunities to target. The marketer wants the answer to inform the go-to-market — which segments to pursue and which messages to lead with. The answer is the same, and it comes from the same customer story.

The four questions aren’t a coincidence. They’re the fundamental questions of understanding demand — what the customer needs, what they currently use, how they talk about it, and what would change their behavior. Product and marketing arrived at the same four questions from different starting points, and the fact that they’re asking the same questions from different departments is the first sign that the org chart, not the customer, created the separation.

Where they genuinely diverge

I want to be honest about the differences, because the argument isn’t “product and marketing should do the same thing.” They shouldn’t. The divergence is real and useful, and it lives in two places.

Product discovery is solution-agnostic; marketing research is solution-aware. The PM’s discovery interviews, done well, deliberately avoid mentioning the product. The customer describes their struggle in their own context, and the PM decompresses from the stated need to the underlying opportunity. The marketer’s research, by contrast, necessarily involves the product — the customer is asked how they found it, what they thought of the demo, what made them choose it or not, what they tell others about it. The PM is learning what to build. The marketer is learning what to say. Both are learning from the customer, but the PM needs the customer to forget the product exists, while the marketer needs the customer to evaluate it. This divergence is genuine, and it means the interview scripts can’t be identical — but they can share the first half.

Product discovery is continuous; marketing research is periodic. The continuous discovery habit means the PM is talking to customers every week, building an evolving picture of the opportunity space. Marketing research typically happens in bursts — a quarterly persona refresh, a win/loss analysis after a competitive loss, a market-sizing exercise before a campaign. The cadence difference means the PM has a rolling, current picture while the marketer has a periodic, deep picture. When the two share signal, the PM’s frequency compensates for the marketer’s depth, and the marketer’s depth compensates for the PM’s tendency to focus on one opportunity branch at a time.

Product discovery validates; marketing research positions. The PM is asking “is this problem real, and is it worth solving?” — the desirability gate. The marketer is asking “how do we frame this problem so the right customers care?” — the positioning gate. Both need the same customer understanding, but the PM uses it to decide what to build and the marketer uses it to decide what to say. The separation makes sense at the point of action. It doesn’t make sense at the point of learning.

The cost of the separation

The cost is double, and it’s measurable.

Double the customer contact budget. The PM books interviews for discovery. The marketer books interviews for research. Both are scheduling conversations with customers, both are asking overlapping questions, and both are consuming a finite resource — customer attention and willingness to participate. I’ve worked with teams where the PM and the marketer were each running eight customer conversations a month, and at least three of those conversations overlapped in subject matter. Sixteen conversations, eight of which were redundant. At an average cost of one hour of internal time plus customer goodwill per interview, that’s eight hours of waste per month — and the goodwill cost is harder to measure but real, because customers who get interviewed twice by the same company start to question whether anyone is listening.

Half the signal, twice the filing. The PM’s notes go into the discovery system — snapshots, the tree, the signal cards. The marketer’s notes go into the research repository — personas, competitive battlecards, message testing results. Neither system references the other. The result is two incomplete pictures of the same customer, stored in separate systems, accessible to separate teams. The PM knows what the customer needs but not how they describe it in the buying process. The marketer knows how the customer describes it but not what they’d actually switch for. The combined picture — needs plus language plus competitive context plus switching triggers — is the complete demand signal, and it lives in neither system.

Misaligned outputs. The worst cost is the one that takes months to surface. The product team, working from discovery interviews, builds a feature that solves a real customer problem. The marketing team, working from separate research, positions the feature in a way that doesn’t match the customer’s mental model. The feature is right, the positioning is wrong, and the mismatch produces confused adoption — customers who need the feature don’t find it because the messaging doesn’t resonate, and customers who respond to the messaging don’t need the feature. This is the failure I watched happen with the PM and marketer I described at the top: the features and the positioning were accidentally coherent because the customer happened to describe the same reality in both conversations. The failures — the features that didn’t match the positioning, the positioning that didn’t describe the features — are invisible because they never shipped as a mismatch; they shipped as separate products and separate campaigns that happened not to align.

The bridge: shared signal, separate analysis

The fix isn’t to merge product and marketing research. It’s to share the raw signal and let each function analyze it through their own lens. In practice this means three things.

Shared interview calendar. The PM and the marketer maintain a single calendar of customer conversations. When the PM books a discovery interview, the marketer is invited — not to ask questions during the interview, but to listen and take notes on the marketing-relevant aspects (language, competitive context, buying triggers) while the PM takes notes on the product-relevant aspects (pain points, workflow, opportunity). When the marketer books a win/loss interview, the PM is invited for the same reason. The customer gets one conversation, not two. Both functions get the signal they need. The shared calendar also reveals redundancy: if the PM and the marketer were both planning to talk to customers in the mid-market operations segment this quarter, they now know it, and they can coordinate instead of doubling up.

Shared verbatims, separate interpretations. This is the mechanical fix that does the most work. Both the PM and the marketer capture verbatims — the customer’s exact words — and both store them in a shared signal repository. The PM interprets the verbatims as opportunities. The marketer interprets them as positioning language. Both interpretations are valid, and the separation of verbatim from interpretation is what lets each function use the same raw material without contaminating the other’s analysis. The signal card format already has the right structure: verbatim mandatory, interpretation separated, source tagged. The only change is that the PM and the marketer both write cards from the same conversation, and the cards live in the same directory.

Shared language audit, quarterly. Once a quarter, the PM and the marketer sit together for an hour and compare the language they’re hearing. The PM has been hearing customer problems in the context of discovery — what’s broken, what’s missing, what workaround exists. The marketer has been hearing customer language in the context of buying — what they searched for, what they compared, what they told their boss. The overlap is the positioning sweet spot: the language that describes a real problem in terms the customer actually uses. The divergences are equally valuable: the PM’s language reveals needs the marketer hasn’t framed yet, and the marketer’s language reveals buying contexts the PM hasn’t considered. The quarterly audit produces a shared document — half a page, no slides — that both functions reference in their work. The PM uses it to write user stories in the customer’s vocabulary. The marketer uses it to write positioning that matches the customer’s mental model.

The JTBD bridge

The framework that sits naturally at the intersection of product discovery and marketing research is Jobs to Be Done, and it’s worth naming explicitly because it gives both functions a shared vocabulary for the first time.

JTBD’s switch interview — the structured conversation about why someone adopted or abandoned a product — contains everything both functions need. The PM learns the job the customer was hiring the product to do, the struggling moment that triggered the search, and the forces that governed the switch. The marketer learns the same things, plus the language the customer used to describe the job to themselves and to their colleagues, the alternatives they compared, and the moment in the buying process where the decision tipped. One interview, two functions, complete demand signal.

The switch interview is also the format where the PM’s and the marketer’s questions are most naturally interleaved, because the customer is already telling a story — a narrative arc from struggling moment to solution — and both the PM’s questions (“what did you try before this?”) and the marketer’s questions (“what did you tell your boss about why you chose this?”) fit inside the same narrative without disrupting it. A well-run switch interview doesn’t feel like two separate research sessions bolted together. It feels like a conversation about a real experience, and the product and marketing signal comes out of the same storytelling.

I’ve started recommending that PMs and marketers co-run switch interviews — not alternating questions, but one person leading and the other listening, swapping roles between interviews. The leading function drives the decompression and the story excavation. The listening function captures the signal their discipline cares about that the leading function might miss. After three interviews, both functions have a richer picture than either would have alone, and the shared experience produces the informal alignment that a shared-doc process can’t replicate.

What changes when you bridge the gap

I want to be concrete about the outcomes, because “better alignment” is the kind of phrase that sounds good in a slide and means nothing in practice.

The positioning matches the product. When the marketer’s positioning language comes from the same discovery signal the PM’s roadmap comes from, the product and the message describe the same customer reality. The landing page says what the product does in the words the customer uses to describe their need. The demo walks through the workflow the customer actually follows. The onboarding addresses the struggling moment the customer actually has. This sounds obvious, but it’s rare — most positioning is written from the marketer’s research and the PM’s feature list, which are two different documents describing two different versions of the customer. Shared signal collapses them into one.

The roadmap reflects buying context. When the PM hears the customer’s buying context — what alternatives they considered, what triggered the search, what they told their boss — the roadmap includes opportunities that account for competitive reality, not just product capability. A feature that solves a real need but in a category the customer isn’t searching in needs a different go-to-market than the same feature in a category the customer is already shopping. The PM doesn’t need to own the go-to-market, but they need to know it exists, because it affects prioritization — a feature that the marketing team can position effectively is a feature that will get adopted, and a feature that can’t be positioned is a feature that will ship and sit.

The customer contact budget is halved. This is the simplest outcome and often the one that gets executive sponsorship. One shared interview calendar, one set of conversations, two functions extracting signal. The customer’s willingness to participate doesn’t change; the internal duplication does. At eight shared interviews a month instead of sixteen separate ones, the organization saves eight hours of internal time and — more importantly — eight units of customer goodwill. That goodwill compounds: customers who feel heard once, deeply, by a coordinated team are more likely to participate again than customers who feel interrogated twice, shallowly, by two teams who don’t know about each other.

The org chart is the problem, not the people

I want to close by naming the structural issue, because the fix I’ve described is a bridge over an org-chart gap, and bridges are workarounds, not solutions. The PM reports to the CPO or VP of Product. The marketer reports to the CMO or VP of Marketing. Their research budgets are separate, their tools are separate, their review cadences are separate. The separation is structural, and it’s not accidental — it reflects a real difference in what each function does with the signal.

But the customer doesn’t experience the org chart. The customer experiences one company, one product, one message. When the PM and the marketer sit in the same interview and take different notes, the customer has no idea. When the PM builds a feature the marketer can’t position, or the marketer positions a story the PM didn’t build for, the customer notices — not as “the product team and the marketing team are misaligned,” but as “this product doesn’t do what the website says,” or “this product does something I didn’t know I needed and nobody told me about it.” The org chart is invisible to the customer. The misalignment it produces is not.

The shared signal practice — shared calendar, shared verbatims, quarterly language audit — is the lowest-cost intervention I’ve found that bridges the gap without requiring an org restructure. It costs two hours a month of joint time and produces both functions’ customer understanding from the same source. The alternative is continued duplication, continued misalignment, and continued double-spending on a finite customer attention budget.

The PM and the marketer in that interview I described at the top were both good at their jobs. The problem wasn’t their skill. It was that the org chart told them their jobs were different, so they filed their notes in different systems, and six months later neither knew the other had been in the room. The customer’s story was told once and heard twice, separately, incompletely. The fix is the simplest one in this post: hear it together.

Further reading

• Discovery and Customer Understanding — the discovery frameworks (JTBD, opportunity trees, Mom Test) that product uses; the marketing overlap starts here.
• Positioning: The Choice You Make Before the Market Does — where the marketing research converges with the product’s competitive-alternative analysis; Dunford’s five components begin with the same question the PM asks in discovery.
• Finding Opportunities: Interviews, JTBD, and the Opportunity Space — the JTBD switch interview as the shared format where PM and marketing signal naturally interleave.
• The Signal Pipeline — the signal-card format that accommodates both product and marketing interpretations of the same verbatim.
• Making Discovery a Habit — where the shared interview calendar fits in the weekly cadence without disrupting the PM’s existing discovery rhythm.