One job title ranked first in two opposing Japanese hiring questions

On October 15, 2026, Japanese recruiting-services company NeoCareer released a survey of graduate recruitment in the generative-AI era. The online survey, conducted September 4–8, covered 328 recruiters whose employers permit generative AI for work. This is a selected group, not a representative random sample of all Japanese companies or students.

Some 70.7% said their employers might reduce graduate hiring in the future. Of those 232 respondents, 82.3% reported that moves to reduce hiring had already begun. Among the 191 reporting such moves, 49.2% selected marketing as an affected occupation—the highest share. Yet when the full 328 were asked which occupations they wanted to recruit more heavily in the future, marketing again came first at 33.5%.

The striking parallel is not proof that the same employers are shrinking and expanding identical marketing jobs. The two questions and denominators differ, and neither percentage measures changes in total job vacancies across Japan. What makes the result worth examining is why one broad occupation could appear at the top of two different sets of hiring intentions.

Marketing, customer support and sales appeared among the reductions

Among the 191 respondents whose firms already showed moves to reduce graduate hiring, the top areas were marketing (49.2%), customer success and support (38.7%), and sales (38.2%). These percentages are response shares across multiple job categories, not the proportion of employees or positions eliminated in those occupations.

In a different question posed to all 328 respondents, the roles employers wanted to strengthen were marketing (33.5%), sales (33.2%), and IT or systems-development engineering (31.4%). Sales, like marketing, appeared on both lists. The survey does not reveal whether these were the same companies, or whether stated plans became completed recruitments.

A single label such as 'marketing' can cover content production, campaign execution, customer segmentation, data analysis, brand strategy and performance evaluation. It is plausible that technology affects those activities differently. But NeoCareer's survey did not track hiring changes by specific subtask, so an assertion that basic creative jobs were cut while strategic jobs increased would go beyond the evidence.

Rakuten's 2027 graduate marketing course shows why the job title alone is insufficient

To understand what an actual employer asks of graduates, consider Rakuten Group's official 2027 recruitment materials. Its graduate marketing course describes an initial assignment to the group marketing division and work on strategy across marketing, branding and membership. This is one company's recruitment description, not proof of a nationwide hiring pattern.

Listed responsibilities include planning and driving brand-value initiatives, working with individual businesses to improve marketing efficiency, using analytical platforms to support data-based decisions, and creating marketing guidelines and training or knowledge-sharing programs. The work spans functions well beyond writing promotional copy or posting to social media.

Rakuten's separate graduate design course describes prototyping new user experiences with generative AI and developing AI-enabled creative and design workflows. These are distinct hiring tracks, but they illustrate neighboring responsibilities across data, customer experience and technology integration. Neither job description establishes that Rakuten reduced or expanded marketing headcount because of AI.

What employers mean by marketing may depend on the decisions the role owns

Two advertised 'marketing' jobs may involve different responsibilities. A content-operations role might focus on rapidly producing and testing materials. A marketing analytics role might identify customer segments and interpret campaigns. A cross-business brand role might coordinate multiple departments and justify trade-offs against shared commercial objectives.

Consider two students' projects. One generated fifty ad headlines with AI. Another tested several options, checked how real users responded, and explained why a particular message or budget allocation worked. Both can claim AI use, but those experiences provide different evidence of readiness. The survey does not show which one would have a higher chance of receiving a job offer.

A more practical way to read a vacancy is to separate what the hire will do, which business outcomes the person is expected to influence, and where AI fits into the workflow. These questions convert the vague prediction that 'marketing jobs are increasing or disappearing' into concrete preparation for the particular employer and role.

The 84.8% AI-skills finding does not mean an AI test guarantees an offer

In NeoCareer's survey, 48.2% of the recruiters said they placed very high importance on students' AI-related skills or experiences, while 36.6% placed some importance on them, totaling 84.8%. These are recruiters' stated priorities; the number is not a measure of mandatory AI certifications, formal AI testing or actual success rates.

Simply saying that a student has used a chatbot or image generator does not establish readiness for every role. An employer could be interested in how the student defined a problem, checked a generated result, corrected errors and applied the outcome to a real project. The appropriate evidence depends on the work described in the vacancy.

The same survey found communication skills were the most frequently chosen current priority at 31.1%, followed by teamwork at 30.2% and resilience at 28.0%. Communication was already high in respondents' recollections of the pre-AI period, at 29.6%. It would therefore be unsound to claim this survey proves communication only became important after generative AI arrived.

Why keep hiring graduates? The leading answer was future leadership development

When asked why graduate recruitment should continue in the AI era, 25.0% chose the long-term development of executives and management candidates—the largest response share. Maintaining a balanced workforce by age and numbers followed at 20.4%, while bringing in new perspectives was selected by 12.8%. These are stated motivations, not observed management appointments.

A graduate can be hired not only to complete entry-level tasks today but also to gain knowledge of the firm's products, customers and operating responsibilities over several years. Even if some routine tasks become cheaper to automate, an employer could continue investing in the development of future decision-makers. This does not guarantee that its graduate hiring numbers will remain unchanged.

Candidates also have a reason to ask how that development is expected to work. If entry-level routine work changes, what gives new staff opportunities to understand customers, test decisions and take accountability? 'Developing future leaders' becomes useful career information only when a company can explain the experiences and support behind the phrase.

Four questions to ask when evaluating a Japanese marketing graduate role

First, read the tasks rather than the heading: is it campaign operations, customer research, analytics, or cross-business coordination? Second, verify the real eligibility requirements and selection process instead of assuming that a favorable labor-market trend means you can apply. Rakuten's 2027 marketing track, for example, specifies its April intake, business-level Japanese, and extra written questions in the application.

The Rakuten marketing listing also states that successful applicants must obtain a TOEIC score of 800 or an equivalent qualification before joining, with candidates potentially asked to sit the exam during selection if their current score is lower. These are concrete admissions details distinct from the abstract discussion of AI and marketing. Candidates should confirm that the live posting still applies when they submit.

Third, ask what the organization already uses AI for and where new staff must make their own judgments. Fourth, ask how success and learning are assessed: what responsibilities a graduate receives in the first year, which projects they can contribute to, and how they might gain broader accountability. Those answers make similarly named marketing tracks easier to compare.

BANSEOG VIEW | Assess the work within the occupation, not the occupation's name

The October NeoCareer research put marketing first in both a reduction-moves question (49.2% of 191) and a future-hiring-strengthening question (33.5% of 328). Because the respondent pools and questions differ, those findings do not show a net change in marketing employment or contradictory decisions by the same employers. The survey represents recruiters at GenAI-permitting workplaces, not Japan's whole labor market.

Yet Rakuten's actual marketing-track posting demonstrates a meaningful distinction: a marketing job can combine brand strategy, coordination between operating businesses, analytics, evaluation and staff training. Different employers may decide to automate some tasks, retain other decisions with people, or structure early-career roles differently. None of these choices can be inferred from a job title alone.

Companies need to define the problems their hires will own, while students need to test the tasks, desired business outcomes, AI workflow and development path they are actually entering. The reader-relevant question is not whether the word 'marketing' will survive AI. It is which marketing responsibilities a particular employer needs a person to perform.

BANSEOG VIEW

BANSEOG VIEW | An occupational title does not define the underlying work

Marketing's 49.2% reduction-moves share (n=191) and 33.5% strengthening share (n=328) come from different respondent sets and questions. They do not establish national headcount changes or the same employers simultaneously cutting and expanding.

Rakuten's 2027 marketing track specifies strategy, data analysis, brand work and effectiveness verification. It is an example of actual duties, not evidence of how Rakuten's AI adoption changed the number of hires.

Applicants should separate job duties, accountability for outcomes, AI usage, training paths and exact live eligibility requirements rather than deciding from the occupation label alone.

SOURCES

Primary sources and references

Prepared October 16, 2026. Primary NeoCareer/CareerTrus 2026-10-15 release, online September 4–8 survey of n=328 Japanese new-graduate recruiters at employers that PERMIT workplace GenAI. 70.7% reported possible future cuts (not completed layoffs); among n=232 who saw a possibility, 82.3% (n=191) reported moves to reduce graduate hiring already under way. Marketing was cited by 49.2% of that n=191 in the reduction-moves occupation question, versus 33.5% of ALL n=328 in a separate future strengthening question. Neither identifies same-firm contraction and expansion or national net changes, nor demonstrates which marketing subtasks were affected. AI use importance 84.8% (48.2% very / 36.6% somewhat) is stated priority, not AI certifications, tests or employment outcomes. Present communication 31.1% vs recalled prior 29.6% does not establish causal impact of AI. 25% citing future leaders is purpose, not management transition outcomes. Rakuten 2027 graduate marketing official course includes group brand, membership strategy, analytics, program evaluation, April intake, business-level Japanese, application written answers, and TOEIC 800 or equivalent required of successful candidates before joining. Separate design course states GenAI-driven UX prototyping and AI workflow development; not evidence of AI-caused marketing layoffs or hiring growth. Interpretations are independent Banseog analysis.