Building Research Infrastructure to Measure and Disrupt the Illicit Massage Industry Organizational Context

The Network is a data-driven nonprofit focused on measuring and disrupting the illicit massage industry (IMI) in the United States. Rather than centering primarily on advocacy, our core function is the systematic collection, structuring, and operationalization of open-source digital signals that indicate commercial sexual exploitation occurring under the cover of massage businesses.

The IMI operates through highly visible online marketing ecosystems, including sex-buyer review forums and classified advertising platforms. These digital traces create measurable signals that, when captured longitudinally, allow for structured analysis of business operations, network relationships, and market dynamics.

Data Assets and Infrastructure
Over the past several years, we have built a scalable data pipeline to capture and normalize publicly available information associated with suspected IMI storefronts. Our dataset currently includes:

  • 1M+ customer reviews scraped from sex-buyer forums
  • 40,000+ business advertisements
  • Structured business identifiers (name, phone, address, aliases)
  • Longitudinal changes in marketing language and platform usage
  • Multi-source signal integration across five independent data streams

These data are inherently noisy, unstructured, and adversarial in nature. Businesses frequently change names, phone numbers, URLs, and ad copy to evade detection. As a result, we have invested heavily in:

  • Entity resolution and de-duplication methods
  • Cross-source signal triangulation
  • Classification frameworks to distinguish likely legitimate from suspected illicit storefronts
  • Secure infrastructure with controlled authentication and role-based access

In 2025, we launched a centralized Data Platform to serve as a secure “single source of truth,” enabling internal analysts and vetted partners to access standardized, queryable data products. This platform automates substantial portions of ingestion, cleaning, enrichment, and labeling workflows that would otherwise require intensive manual effort.

Research Collaboration Model
Our institutional position is deliberate: we do not aim to lead novel theoretical scholarship on trafficking markets. Instead, we focus on building and maintaining high-quality, structured datasets that can support rigorous research.

We have formal research collaborations underway with:

  • University of Alabama
  • George Mason University
  • A forthcoming partnership with Colorado State University

Our objective is to provide a national-scale tranche of structured IMI data to academic partners to support:

  • Signal detection and classification modeling
  • Network analysis of business clustering and movement patterns
  • Linguistic analysis of advertising and review content
  • Market adaptation and displacement studies
  • Policy and enforcement impact evaluation
  • Research Opportunity

The illicit massage industry presents a rare paradox: it is both exploitative and digitally transparent. While individual victims remain hidden, the storefronts themselves rely on discoverability to function economically. This creates an opportunity for computational and data-driven inquiry.

We believe that pairing operational data engineering capacity with academic methodological rigor can produce:

  • Stronger predictive signals of exploitation risk
  • Better differentiation between legitimate and illicit businesses
  • More precise, evidence-based interventions
  • Reduced unintended harm from overbroad enforcement

While our long-term goal is to defeat the illicit massage industry, we remain interested in enabling a research ecosystem capable of achieving greater analytic sophistication than the adaptive strategies used by traffickers themselves.

Call to Collaboration
We are seeking research partners interested in working with structured, longitudinal, multi-source data on the illicit massage industry at national scale. We welcome collaboration across criminology, public policy, computational social science, machine learning, natural language processing, and network science.

By combining infrastructure, domain-specific data engineering, and academic rigor, we aim to accelerate the pipeline from intelligence to insight to decision to measurable defeat.

Presenter: Ian Hassell