Global – October 28, 2026 -- Path2Response is offering custom digital audiences built on purchase records tied to 273 million individuals, with no model development or onboarding fees and payment only through clients' regular media buys.
Transaction data replaces interest-based inference in the targeting model
The company argues that the real divide in audience targeting is between fact and inference, not demographics. It says Facebook lookalike modeling builds largely on content engagement, while Google's targeting leans heavily on site visits and browsing context. Both read interest, which the company calls an inference by design.
Path2Response models start from the transaction itself. Records of past purchases, tied to core demographics, are submitted monthly by a growing network of retail and nonprofit contributors. Daily retail site-visitor signals and nonprofit gift-donation data are layered on top to capture real-time intent.
Dataset spans 40 retail categories and 19 nonprofit causes
The data is mapped to households and includes exact purchase dates and dollar amounts. Because it draws on recent transactions and daily visitor signals rather than a static file, the company says the resulting audiences reflect current buying behavior.
Audience fees fold into media spend instead of upfront charges
The company says marketers have traditionally absorbed development fees, onboarding charges and heavy minimum commitments before knowing whether a data source would perform. Its alternative is pay-for-performance licensing. Its team builds the model and delivers it to a client's seat ID in the preferred DSP, social or CTV platform. The starting point is either a seed audience Path2Response already has or one the client provides.
On social platforms and The Trade Desk, the targeting fee is a percentage of media buy. On DSPs, it is CPM-based and included in media cost. If an audience misses ROAS, reach and scale targets, Path2Response says it is not paid.