Hikari Soma日本語

Data Scientist · Marketing Analytics Consultant

Turning consumer behaviour data into marketing decisions teams can act on.

Many companies now hold more consumer data than they are able to use. I work with purchase histories and product usage logs, which show what customers did rather than what they say they prefer, and identify which of those behaviours are worth building a campaign around.

MSc, Waseda University
Graduate School of Sport Sciences
Adjunct Researcher
RISB, Research Institute for Sport Business
Yokohama, Japan
Available for consulting engagements

Location: Yokohama, Japan

What I do

Most of my work begins from one of these three situations.

Segmentation & CRM strategy

You have purchase data, but your segments are out of date

Cluster and co-purchase analysis on actual purchase histories, to find the groups that really exist and identify which are worth investing in.

Behavioural log analysis

You know what customers say, not what they do

Device usage logs, settings data and purchase records establish real behaviour; causal inference identifies what drives it.

AI-driven marketing

Your team wants to use AI in marketing

Automating analysis workflows, accelerating Python-based data work, and building the material a team needs to use these tools well.

Selected work

Recent engagements. All client work is carried out as an independent consultant.

  1. Sony Group Corporation

    Consumer electronicsIndependent contract · April 2026 – presentSQL · Databricks · Python · Factor analysis

    The challenge

    A business unit held large volumes of device and purchase log data, but had no reliable way to describe how customers actually used the product, and therefore no basis for deciding which behaviours its marketing should target.

    What I did

    • Built metrics for ongoing effectiveness measurement, developed as consumer behavior features. Identified segments via behavioral clustering on customer surveys, improving targeting precision.
    • Supported the design of survey scales for measuring customer motivation, and through analysis of the resulting data, revealed an objective, evidence-based picture of the customer journey.

    Outcome

    The business unit moved from planning on judgement to planning on evidence, with one set of measures both directing budget and testing afterwards whether it worked.
  2. KURAND

    E-commerce · SakeIndependent consultantCluster analysis · Co-purchase analysis · CRM strategy

    The challenge

    A sake e-commerce brand wanted to build marketing capability in-house rather than keep buying analysis. Its customer segments no longer matched how people actually bought, and no one on the team could tell that from the data.

    What I did

    • Ran cluster and co-purchase analysis on customer data to find the groups that genuinely existed.
    • Translated the findings into customer personas the team could work with.
    • Coached the marketing team on reading customer data themselves, so the next analysis would not need me.

    Outcome

    The new personas broadened the sales strategy, and the team came away able to interpret its own customer data rather than waiting for the next report.
  3. Sola of America, Inc.

    Gaming · Mobile · VRIndependent consultantGoogle Ads · Influencer strategy · Short-form video

    The challenge

    A Japanese studio needed to reach a US audience across two products, a mobile app and a VR game launching on Steam, in a market where discovery runs through creators rather than paid media.

    What I did

    • Ran Google Ads and directed social media strategy for the mobile app.
    • Led an organic influencer campaign for the North American Steam launch of the VR title.
    • Secured partnerships with around twenty US-based YouTube and TikTok creators through direct outreach.

    Outcome

    Over 30,000 views in two and a half months, entirely organic and with no advertising spend, together with a feature placement on SideQuest.
  4. SoftBank Group Corp. — Axross Recipe

    Technical educationIndependent contractMachine learning · Data science · Technical writing

    The challenge

    Working professionals learning machine learning and data science can find plenty of material explaining how a method works, and very little explaining what it is worth once applied to a business.

    What I did

    • Wrote educational content and technical articles on machine learning, data science and AI for the Axross Recipe platform.
    • Used a background in consumer psychology to carry each method through to the business insight it produces, rather than stopping at the technique.

    Outcome

    Consistently among the platform’s top contributors, with strong feedback from a broad subscriber base.

About

Hikari Soma
Hikari Soma, Yokohama, 2026

My work sits between academic research and commercial practice. As an Adjunct Researcher at RISB, the Research Institute for Sport Business at Waseda University, I study the psychology and behaviour of consumers in professional sports. As a consultant, I apply the same quantitative methods to product and purchase data for companies in e-commerce, entertainment and gaming.

I also write in English and Japanese on where marketing and data science meet, including the gap between what consumers say they prefer and what their purchase records show.

Experience

  1. Data Scientist (Independent Contract)

    Sony Group Corporation

    April 2026 – Present

  2. Adjunct Researcher, RISB

    Waseda University

    August 2023 – Present

  3. Independent Marketing Consultant

    Self-employed

    April 2023 – Present

  4. Technical Content Author (Independent Contract)

    SoftBank Group Corp.

    February 2023 – February 2024

Education

Master of Sport Sciences

2023

Waseda University, Graduate School of Sport Sciences

Research grounded in consumer behaviour and psychology, including structural equation modelling (SEM), factor analysis, construct validity, and causal inference methods.

Writing

Articles on marketing and data science, grounded in real consumer behaviour data.

  • Decoding Consumer Preferences: How Multiple Factors Shape Buying Decisions

    How businesses can decode complex psychological, economic, and ethical factors to align marketing strategies with real consumer behavior.

    Read on LinkedIn
  • How will AI technologies change marketing?

    How AI technologies revolutionize customer personalization, campaign ROI, and dynamic pricing, alongside the ethical and technical hurdles teams must navigate.

    Read on LinkedIn
  • Predictive Marketing Using Big Data: How to Apply That?

    How marketers can turn big data and predictive statistical models into reliable forecasts for customer retention, product recommendations, and campaign ROI.

    Read on LinkedIn

Contact

I am open to consulting work and collaboration. If any of the below sounds like what you need, I would be glad to hear from you.

Currently interested in

  • Consumer data strategy for e-commerce, entertainment and gaming companies
  • Sponsorship evaluation and analytics for sports organisations
  • Helping marketing teams integrate AI into their workflow