Predictive Modeling scores every individual on the probability of one specific outcome — the output is a ranked list with a 0–1 score. Signature Segmentation groups a first-party file into named archetypes based on shared values, behaviors, and motivations — the output is labeled audience groups with profiles and an activation-ready file.

Signature Segmentation can be run on a Predictive Modeling output (e.g., a specific decile band) to reveal the distinct motivational groups within that scored population.

Signature Segmentation delivers a portable file — a labeled customer file at the individual level that can be used in any environment (CRM, Snowflake, email platform, CDP, CMS, paid media). Dynamic Segmentation in Ignite and Cortex enables in-platform building and digital activation.

Install and Enrichment deliver raw attribute access. Enrichment appends Resonate attributes to your records; Install delivers a population-level file you join yourself. Signature Segmentation is a finished archetype framework built from that same underlying data — Resonate does the segmentation work and hands back a named-archetype report plus a labeled file.

Two input types at launch: a first-party file (CRM), or a specific decile band from a prior Resonate Predictive Modeling output. Tag-based audiences (pixel-tagged site visitor pools) are on the Phase 2 roadmap.

HEM (SHA256 default, SHA1, or MD5), MAID, and IP address (IPv4).

20,000 matched records. Because match rates vary by ID composition, plan raw submissions accordingly — 40,000+ on a HEM-only file, less on multi-ID files.

40–60% on HEM-only files. 60–80% on files with multiple ID types.

If the matched universe falls below the minimum threshold for reliable clustering, the engagement pauses at intake. The Data Integrations team will advise on how to supplement the file before proceeding.

Within two weeks from receipt and validation of the file to delivery.

An average of six times per year, approximately every 8 weeks is the standard cadence. Quarterly, semi-annual, or annual cadences are available.

The locked scheme is applied to the updated file. Delivered outputs include the updated segmented file and a Drift Analysis report.

Drift Analysis monitors whether the composition of the audience has shifted meaningfully since the scheme was locked. It reports one of three states: stable, drifting, or rebuild recommended.

Re-runs classify new records against the existing locked scheme. Rebuilds regenerate the scheme and produce new archetypes. Rebuilds are triggered by one of three signals: Drift Analysis recommends it, the 24-month backstop is reached, or a major business event has changed the file's composition.

Each segment arrives with a name and a description that summarize what defines the group — the values, behaviors, and motivations that distinguish that segment from the others. The Named Segment Report goes deeper, showing the differentiating attributes that shaped each segment, its size relative to your file, and how it compares to the other segments in the same scheme. The intended use case and industry context you provide at intake inform how the segments are described, so the framing matches how your team will apply them.

If a segment isn't landing clearly for your team, your Resonate contact can walk you through the interpretation.

Between four and eight, in most engagements. The count is determined by the data to produce the highest-quality, usable scheme — not fixed in advance. Segments are always distinct and mutually exclusive.

Equal-sized groups aren't what naturally exists in an audience. Signature Segmentation sizes each segment by what the data supports, so sizes vary.

Questions? Contact your Customer Success Manager or resonatesupport@resonate.com.