Japan’s Product Problem Is No Longer Making Things—It Is Selling Them

The product is finished; the growth has not begun

Imagine the launch meeting. Engineering demonstrates a stable product. The roadmap has been delivered. Marketing has a website and a release. Sales has a deck. Customer success has an onboarding checklist. Every department completed its assignment, yet no one can answer the first commercial question in one sentence: for which customer, facing which urgent alternative, is this product the best choice now?

That unanswered sentence changes everything downstream. If the segment is vague, research mixes unlike customers. If the problem is vague, positioning becomes a list of features. If the value is vague, price looks arbitrary. Sales improvises. Marketing optimizes clicks that do not become revenue. The roadmap absorbs the loudest request. When growth disappoints, each function blames the handoff.

Prodelia Partners, a Tokyo consultancy founded in January 2026, calls this the gap between making a product and delivering its value to a market. Its Japan PMM Market Report 2026, released July 21, surveys the people who stand in that gap: product marketing managers, product managers, marketers, business developers, executives and others. The findings describe an organizational function that already exists informally but lacks a name, authority, method and career.

The report is important less because it discovers a national defect than because it makes a hidden job visible. Someone in every product company must synthesize customer evidence, choose a market, define competitive meaning, decide how the offer is packaged, equip revenue teams, orchestrate launch and return what the market teaches to the product team. When that work belongs to everyone, it often belongs to no one.

What the survey actually found

The online survey ran in June 2026, used 22 questions and retained 166 responses after removing six duplicates or out-of-scope submissions from 172 collected. The report says every displayed question uses the full n=166. Respondents were asked about roles, organization, GTM bottlenecks, methods, careers, AI use and future staffing.

On the central question, 45.2% said they strongly felt a GTM problem and 41.6% somewhat felt one. Asked to select the single phase where new businesses or products most often stall, 45.2% chose market entry, 22.9% post-launch scaling, 18.7% planning and 13.3% development. The combined result suggests that respondents perceive nearly seven in ten bottlenecks after the idea has moved toward the market, but the survey did not measure actual project cycle times or failure records.

Within GTM, 27.1% selected positioning and value definition as the single hardest phase. Sales-marketing coordination followed at 17.5%, market and customer understanding at 16.9%, target-segment definition at 13.9%, messaging at 13.3%, channel and pricing at 7.8%, and measurement at 3.6%. In a separate multiple-choice question, the leading organizational problems were insufficient people and skills, selected by 59.6%; unclear ownership, 44.6%; fragmented customer insight, 42.2%; and lack of common understanding across planning, sales and marketing, 30.7%.

The apparent paradox is that PMM work is already widespread. Dedicated PMM titles existed at 27.7% of respondents' companies. Another 34.3% said the function existed without that title and 30.1% said another role performed some of it. Adding those categories produces the report's 92% figure. Yet respondents rated clarity between “making” and “delivering” roles at 2.4 out of five, availability of methods and education at 2.3, and career-path clarity at 2.2. Only 7.2% gave career clarity a four or five.

Before calling it “Japan,” read the sample

The report itself contains the most important warning: this is effectively a survey of Japan's B2B SaaS and product market, not a representative value for all industries. SaaS and IT account for 84.9% of respondents. Manufacturing is 4.8%, retail and distribution 3.0%, healthcare 1.8%, finance 1.8% and the public sector 0.6%. The headline about Japan's ability to make things should therefore be read as metaphor, not a conclusion about factories.

Company size is concentrated too. Firms with 51–300 employees account for 44.6% and those with 301–1,000 another 28.3%. Only 16.9% are at companies with 50 or fewer staff, while 10.2% work at companies above 1,000. The sample describes growth-stage organizations where specialization is becoming necessary but boundaries remain fluid.

Respondents are unusually close to the subject: 27.7% identify as PMMs, 18.7% marketers and 18.7% product managers. Business development contributes 10.2% and executives 9.0%. Fifty-seven percent have less than three years in their stated domain. This makes the responses valuable as frontline testimony and unsuitable as an estimate of how the average Japanese worker or chief executive views PMM.

The report does not publish its recruitment frame, response rate, sampling weights, confidence intervals, raw data or subgroup cross-tabs. All questions display n=166 even where “not applicable” might affect interpretation. Several central questions ask what respondents think is the bottleneck, not what independently recorded performance data shows. The publisher's claim that this is Japan's first quantitative PMM reality survey is based on its own search as of July 20.

There is also a commercial interest to disclose. Prodelia sells PMM training, frameworks and consulting. The last pages of the report introduce its academy and prices, after a question finds that 78.3% of respondents are interested in external training, methods or consulting. That alignment does not make the findings false. It makes independent replication, transparent recruitment and raw aggregate tables especially valuable.

PMM is not a new name for advertising

Product marketing management sits between product truth and market choice. It starts with evidence: interviews, usage, win-loss analysis, competitors, sales calls, support tickets, renewal and churn. It converts that evidence into decisions about ideal customer, urgent problem, category, differentiation, proof, packaging, price, route to market, launch sequence, adoption and enablement. Then it measures whether the market behaved as expected and changes the product story—or the product itself.

Marketing communications asks how to reach and persuade an audience. Product management decides what problem the product should solve and what should be built. Sales works individual opportunities. Customer success makes customers realize value after purchase. PMM is the connective operating function: it defines the commercial truth those teams use together.

The boundary varies by company. In an early startup, a founder may perform it. In product-led software, product managers and growth teams may own much of it. In enterprise SaaS, PMM may lead positioning, launches, competitive intelligence and sales enablement while revenue operations manages the funnel. In a manufacturer, the closest role may be product planning, technical marketing, application engineering or overseas business development.

A new title is not the objective. Clear decisions are. Who names the primary customer? Who owns positioning? Who can reject a feature request that does not fit the segment? Who decides packaging and price? Who writes the launch brief? Who measures adoption by promised use case? Who returns lost-deal evidence to the roadmap? If those rights are explicit and exercised well, an organization has a PMM system even without the acronym.

The job began before software—and before Japan's economic miracle

The conventional origin story starts at Procter & Gamble in 1931. Neil McElroy, then working on Camay soap, wrote an internal memorandum proposing “brand men” responsible for studying shipments, identifying weak markets, developing plans, coordinating execution and testing results. The memo was not merely about better advertising. It put one person close to the full commercial performance of a product.

Consumer-goods companies refined brand management across the mid-century. Technology companies later adapted the logic to products whose markets and capabilities changed faster. Product management came to represent the internal problem-and-roadmap side; product marketing increasingly represented the market, positioning and adoption side. The separation was never clean, because market evidence should shape what is built and product reality constrains what can be promised.

Software subscriptions intensified the need. A packaged product could be launched, sold and replaced on a long cycle. SaaS changes weekly and must earn renewal continuously. Distribution is cheap, so competitors appear quickly. Buyers research before speaking to sales. Usage can be measured by account, cohort and feature. Packaging can change without a new assembly line. The product-market conversation becomes permanent.

The deepest idea in McElroy's old memo survives: do not organize only around the factory that produces the thing. Organize around the market performance of the thing. Japan mastered the first half through production, quality and supplier systems. The 2026 survey asks whether its software companies have built an equivalent management system for the second.

Japan learned to sell by learning to make

It would be historically wrong to describe postwar Japan as a nation of engineers indifferent to customers. Its export achievement joined manufacturing and market insight. Companies learned statistical quality control, built supplier capabilities, compressed defects and lead times, adapted products to foreign households and created brands that became shorthand for reliability. Toyota's production system did not merely lower cost; it treated problems and customer demand as signals that should pull work through the system.

Sony did not win global living rooms because transistors were intrinsically persuasive. It made portability understandable and desirable. Toyota and Honda matched fuel efficiency and reliability to changing American needs. Panasonic built distribution and a broad promise around household life. Canon, Nintendo, Shimano, Keyence and many others found distinctive combinations of technology, application knowledge, channels and service. Japan has abundant counterexamples to the claim that it cannot market.

The old system nonetheless favored a particular kind of commercial learning. Long product cycles, domestic distributors, keiretsu ties, large sales forces and stable corporate customers could carry market knowledge through relationships. Engineers visited factories. Salespeople translated customer demands. Product planning reconciled headquarters and channels. Much of the knowledge remained tacit, embedded in people and routines rather than named as a cross-functional discipline.

That system is powerful when customers are known, quality attributes are legible and improvement is cumulative. It is weaker when a company must create a category, serve unknown buyers through digital channels, choose among radically different segments or alter pricing every quarter. SaaS exposes the invisible coordination work because there is no distributor or multi-year model cycle to hide its absence.

Success created its own blind spot

“Good products sell themselves” was never literally true. It became a plausible internal story because Japanese quality was once an enormous and visible differentiator. When imported goods were associated with defects, reliability was positioning. When cars consumed less fuel and lasted longer, engineering created a message the market could understand. A technical advantage and a commercial advantage temporarily overlapped.

As competitors adopted quality management, manufacturing reliability became an entry ticket rather than a unique claim. Value migrated toward software, ecosystems, user experience, data, financing, service, brand, speed and integration. A machine's metal could be excellent while its configuration interface, developer tools, pricing, support model or international channel made it difficult to buy.

The Japanese government has recognized this gap for years. The 2017 manufacturing white paper described the growing importance of customer-centered design thinking in a culture historically inclined toward technology- and market-seed starting points. The Patent Office's 2018 Design Management Declaration argued that design should enter management, branding and innovation from the upstream stage, discover latent user needs and use iterative development. These were not aesthetic campaigns; they were institutional attempts to connect capability to human value.

The 2023 SME White Paper sharpened the point. It found that research-intensive small companies often believe they could commercialize more new products if people helped bridge core technologies and market needs. Its case of Nissin Kogyo shows that precision metal-forming expertise survived the collapse of cathode-ray-tube demand by finding applications in electric-vehicle batteries and medical devices. The technology mattered. So did a coordinator who understood both the capability and the new buyer.

That bridge is the enduring subject. “Monozukuri versus marketing” is a false choice. The better equation is capability multiplied by market connection. If either approaches zero, the commercial result does too.

Software changed the unit of competition

A manufactured product reaches a design freeze because changing tooling, certification and supply chains is expensive. A cloud product can change after breakfast. That flexibility creates a different management problem: the organization can build many things, so it must become better at deciding which value to pursue and how to recognize it.

Japan's digital-transformation record gives the survey a larger context. The government's 2018 DX report warned that legacy systems could impede transformation and create a “2025 digital cliff.” A 2025 Software Modernization Committee report found that Japanese DX remained oriented mainly toward making existing operations efficient, with relatively few cases tied to new business creation. Modernizing software without changing the value model can produce a faster old company.

The OECD's 2026 economic survey presents the macroeconomic version. Japan spends heavily on research and development, yet the composition is concentrated: 79% of R&D is in business, about 86% of that business R&D in manufacturing, and only around 4% in ICT versus 31% in the United States. SMEs account for about 5% of business R&D, the lowest share in the OECD comparison. Japan is not short of technical effort. Its challenge includes diffusing innovation across services, smaller firms and new growth companies.

SaaS makes the gap measurable. A factory can optimize yield while the market remains distant. Software records activation, time to value, feature adoption, expansion and churn. A company can see that customers fail to reach the promised outcome. The difficult part is agreeing what those signals mean and who can change product, offer and motion in response.

Subscription economics also punish the ceremonial launch. Revenue does not arrive once at shipment; it is earned across retention. A message that wins a trial but attracts the wrong customer raises support load and churn. A discount that wins an enterprise contract may hide weak adoption. GTM is therefore not the publicity surrounding release day. It is the complete design of customer acquisition, value realization and renewal.

The hardest work is “value language”

The report's largest single GTM bottleneck—positioning and value definition at 27.1%—can sound soft beside code and revenue. It is not copywriting. It is a chain of falsifiable choices.

A position names the customer, context, alternative, important outcome, distinctive capability and proof. “AI-powered business platform” is not positioning because it excludes no buyer, identifies no urgent change and establishes no advantage. “For 100–500 employee Japanese manufacturers that lose export quotes in manual engineering review, this service converts drawings into a compliant cost estimate within one day” can be tested. Sales can find such accounts. Product can measure the workflow. Finance can model willingness to pay. Competitors can be named.

Japanese B2B teams face a particular temptation to describe specifications because specifications are safe. A feature is demonstrably present. A customer outcome depends on context and makes a stronger promise. Yet buyers do not fund feature existence; they fund reduced risk, revenue, time, labor, compliance or strategic option value. PMM translates without inventing. The promise must remain supported by product evidence.

Localization deepens the problem. A phrase that works in Japan's consensus-driven enterprise sales may fail in a self-service American funnel. The buyer, budget owner, user, security reviewer and reseller may be different people. Overseas GTM is not translating a Japanese brochure. It is rebuilding the buying map: category vocabulary, reference customers, procurement, price anchors, channel incentives and acceptable proof.

One function, five adjacent jobs

Introducing PMM without defining decision rights can make the survey's “everything person” problem worse. The new hire receives launch slides, website copy, competitive sheets, analyst relations, pricing, user research and every task that falls between departments—yet no authority to choose a segment or alter a roadmap.

The operating solution is not a perfect RACI chart. It is a small set of explicit decisions, one accountable owner for each and a recurring forum where customer evidence is adjudicated. Product and PMM should jointly own the product-market thesis. Finance, sales and PMM should shape pricing. Customer success should own realized-value evidence. The chief executive must resolve category, segment and resource choices that cross functions.

PMM also needs a measurable mandate. “Support the launch” is a task. “Increase activation among accounting firms with 20–100 employees from 35% to 55% without worsening 90-day retention” is an outcome. The second gives the role permission to question audience, onboarding, promise and product.

AI makes words abundant and judgment scarce

The survey's respondents are not waiting for AI. Only 2.4% say they do not yet use it in PMM or GTM work. Fifty-nine percent use it to generate messaging ideas, 54.8% for content and materials, 46.4% for data analysis, 41.6% for research summaries, 39.2% for personas or customer understanding and 36.7% for market or competitor research.

Those numbers measure claimed use, not quality. Generative AI can turn one positioning brief into fifty headlines, summarize interviews and cluster sales notes. It can also manufacture consensus from biased inputs, erase exceptions, hallucinate competitors and produce fluent messages that no buyer recognizes. More content does not solve weak positioning; it industrializes it.

Respondents understand the paradox. Four in five expect the value of cross-functional people to rise as AI spreads, and 77.7% say their work already crosses three or more domains. When drafting becomes cheap, the scarce work is selecting evidence, resolving conflict and taking responsibility for a market bet. AI can propose segments. It cannot be accountable for firing the wrong customer, changing the price or diverting engineering effort.

A mature AI workflow preserves provenance. Every synthetic insight links to calls, tickets, usage or research. Different customer segments remain distinguishable. Personally identifiable and confidential data stays governed. Generated claims pass product, legal and evidence review. Teams measure whether AI-assisted work changes conversion, adoption or retention—not how many assets it produces.

The missing system is a learning loop, not a launch checklist

A product organization can perform every standard GTM deliverable and learn nothing. It can produce personas, a positioning document, launch tiers, a sales deck and a dashboard, then store them across folders. The system becomes real only when market evidence changes a decision.

The loop begins with a thesis. A defined segment has a specific costly problem and will choose the product for a defensible reason. Before launch, the company writes disconfirming evidence: which interviews, usage patterns, willingness-to-pay results or competitive losses would prove the thesis wrong. It instruments the promised journey. Then it reviews leading and lagging signals by segment.

Interview evidence must be separated from compliments. Pipeline must be separated from unqualified interest. Activation must reflect the customer's first meaningful outcome, not login. Adoption must connect to use cases, not aggregate clicks. Retention must be cohort-based. Win-loss analysis must include deals the company never saw and customers who chose to do nothing.

The meeting cadence matters. Weekly, teams inspect live friction. Monthly, they review segment health and sales evidence. Quarterly, leaders decide whether to persist, reposition, repackage or stop. Every decision records evidence, owner, expected effect and review date. That creates organizational memory when people transfer—a recurring Japanese enterprise problem highlighted by Prodelia's July collaboration with Findy.

Measure the promise, not the department

PMM fails when evaluated by activity counts: launches supported, pages written, events attended, competitive cards updated. Those are useful inputs and dangerous goals. They encourage a department to ship artifacts while revenue and product teams continue operating from different truths.

No single metric belongs uniquely to PMM because the work is cross-functional. The correct scorecard follows the market thesis. At the top: share of research from the defined ideal customer, problem frequency, message comprehension and willingness to change. Through acquisition: qualified-account rate, segment conversion, sales-cycle length, competitive win rate and discount. After purchase: time to first value, promised-use-case adoption, support burden, retention, expansion and referenceability.

Each release should carry an evidence plan. Did the target account notice? Did the intended role understand? Did sellers use the narrative? Did the feature reach meaningful use? Did behavior change? Did the benefit persist? A launch date without these questions is a delivery milestone, not a market outcome.

The survey's finding that only 3.6% choose measurement and improvement as the hardest GTM phase should not be read as proof that measurement is solved. Respondents may feel the upstream ambiguity more acutely, or may not reach a mature measurement loop at all. The report does not cross-tab measurement difficulty against company maturity. Good positioning and good measurement are complements: a precise promise tells a company what to observe.

Hiring one PMM will not repair the company

Sixty-one percent of respondents say their organizations intend to establish or increase PMM capacity over the next 12–24 months, while 2.4% expect contraction. Eighty-three percent predict that PMM's importance will rise over three years. Demand may be real, but hiring into an unresolved operating model can create a highly articulate bottleneck.

Before opening a requisition, leaders should identify the failure. If no one speaks to customers, build research access. If product and sales use different segments, assign a product-market thesis owner. If launches fail at adoption, connect customer success to roadmap and messaging. If pricing is political, establish a pricing council and willingness-to-pay research. If international sales stalls, hire local market discovery, not merely translation.

A PMM's seniority should match the decision. A junior content-oriented hire cannot resolve segment conflict between executives. A strategic PMM without analytical and operational support becomes a presentation layer. A single PMM supporting ten products cannot stay close to customers. The report's low career-clarity score matters because companies must define progression from research and launch craft to portfolio, pricing, category and executive leadership.

Japan need not import an American job description intact. It can map existing strength: sales engineers who understand deployment, product planners who understand technology, customer-success leaders who know adoption, overseas staff who know buying context, designers who know unmet needs. The aim is to recognize and train a cross-boundary profession while preserving domain depth.

The lesson returns to the factory

Although manufacturers are only eight respondents in the report, its logic matters to physical industry. Modern machinery is sold with software, monitoring, maintenance, financing and outcome guarantees. Automobiles are becoming software-defined. Industrial equipment becomes a data service. Components enter global design ecosystems before they enter factories. The product increasingly includes the buying, integration and operating experience.

METI's manufacturing work frames digitalization as more than QCD—quality, cost and delivery. It also identifies service and platform models, faster quotation, customer-value creation and new business opportunities. The government's updated mobility DX strategy sets an ambition for Japanese companies to capture 30% of global software-defined-vehicle unit sales in 2030 and 2035. Such a goal cannot be won by vehicle assembly alone; it requires platforms, developer relationships, data governance, recurring services and market architecture.

A manufacturing PMM may therefore translate application engineering into a repeatable market offer. It identifies which operating problem a sensor solves, packages hardware with analytics, builds proof from installed performance, equips distributors, determines who owns the recurring software budget and returns field evidence to engineering. This is not “selling harder.” It is designing the whole offer.

Japan's strongest model may be the fusion of genba and GTM. Go to the place where value is used, observe the work, make the problem visible, form a hypothesis, change the offer, measure the result and standardize what works. That resembles kaizen. The difference is that the object of improvement is not only the production process; it is the product-market system.

A 90-day operating plan

The plan deliberately starts with disagreement, not a rebrand. Most organizations already possess scattered market knowledge. Sales knows objections, success knows failed value, product knows usage and finance knows discounting. PMM's first contribution is to make conflicting evidence visible and force a choice.

One narrow segment is enough for the first loop. Companies often resist focus because exclusion feels like lost revenue. But a broad market hides causality. A team cannot learn whether positioning works when every customer buys for a different reason. Focus is a temporary instrument for producing reliable knowledge, not a permanent refusal of opportunity.

What stronger evidence would look like

Prodelia's report has done useful pioneering work: it publishes every question, distribution, full sample size, company-size profile and a prominent warning about industry bias. It distinguishes single-choice, multiple-choice and five-point questions. It does not disguise the report's commercial sequel. Those are meaningful strengths.

A second edition could establish a probability-based or clearly documented recruitment frame, record response rate, publish a codebook and anonymized aggregate data, and separate company-level from individual-level answers. Stratified samples could compare manufacturing, consumer, finance, public sector and SaaS; small, mid-sized and enterprise firms; dedicated PMM organizations and informal ones.

Longitudinal evidence would be more powerful than intention. Follow firms that add PMM capacity and comparable firms that do not. Measure positioning clarity, launch cycle, qualified conversion, win rate, activation, retention and revenue efficiency before and after. Control for company stage and product quality. Ask multiple people within the same company whether role clarity agrees. Compare perception with operational data.

The central claim also needs a symmetrical test. If 45% perceive GTM as the largest stall point, what predicts the 13% who select development? Are low-quality products underrepresented because the sample comes from market-facing practitioners? Does dedicated PMM correlate with better outcomes—or merely with larger companies and more complex portfolios? Does “92% have the function” survive a behavioral definition rather than self-labeling?

Independent universities, industry groups and government agencies should repeat the questions. A shared benchmark could become valuable infrastructure for Japan's product economy. The purpose is not to award PMM a permanent organizational territory. It is to discover which combinations of customer evidence, decision rights and market practice produce better products and stronger businesses.

The bridge is now part of the product

The headline says Japan's problem is no longer making things but selling them. Taken literally, it is wrong. Development remains difficult. Quality remains decisive. Japan's research system, factories, software teams and small suppliers face real technical constraints. Eight manufacturing respondents cannot retire a national industrial tradition.

Taken as a description of a changed bottleneck in growth software, it is illuminating. Development methods, cloud tools, open source and AI have made building faster. Distribution has become global. That raises the relative value of choosing: which customer, which problem, which category, which promise, which proof, which price and which learning loop.

“Selling” is too small a word for this work. The objective is not to decorate a finished invention or pressure a customer. It is to connect the market early enough that the right thing is made, explain it truthfully enough that the right customer can choose it, deliver value quickly enough that the customer keeps it, and return evidence rapidly enough that the company learns.

Japan's manufacturing tradition offers the right moral framework. Quality is not inspected in at the end; it is designed into the process. Market fit should be treated the same way. It cannot be added by a launch campaign after development. It must be built through continuous contact among customer, product and revenue work.

The survey's 92% figure may not show that Japan has already solved PMM. It shows that people are already doing fragments of the job because organizations cannot function without it. The next step is not fashionable titles. It is to give that work evidence, authority, memory, metrics and a career.