Enterprise Intent Data Vendor Checklist

Use this checklist before the demo controls the agenda

Intent-data vendors often package signal collection, identity, contact validation, segmentation, delivery, and services under one label. A procurement team should score those layers separately and require evidence for material claims.

1. Define the job

• Is the goal net-new acquisition, CRM prioritization, retention, media activation, property research, or predictive scoring?
• Which audience, geography, behavior, recency, and business outcome matter?
• Which system owns the record after delivery?
• Who is accountable for compliance, campaign execution, and measurement?

2. Inspect provenance and rights

• What source categories contribute to the signal and identity record?
• What is observed, licensed, modeled, inferred, or appended?
• Which timestamps are delivered?
• What contractual rights permit the stated use?
• Which purposes, channels, industries, or geographies are prohibited?
• How are consumer opt-outs, disputes, and deletion requests handled?

3. Define the signal

• What behavior qualifies for each category?
• Is the event person, household, address, or device level?
• How is repeated activity handled?
• What is the source-to-delivery latency?
• What lookback windows and refresh schedules apply?
• Can the vendor provide evidence without exposing proprietary publisher identities?

4. Evaluate identity and fields

• What matching inputs and confidence methods are used?
• What does the published match rate actually measure?
• Which fields are validated, by what method, and how often?
• How are conflicts, shared households, moves, and stale contacts represented?
• Can provenance and timestamps survive export into downstream systems?

5. Review security and operations

• Request an architecture summary, control evidence, data flow, subprocessor list, incident process, retention schedule, deletion process, and access-control model.
• Confirm encryption, transfer methods, audit logs, backup approach, environment separation, and customer-side responsibilities.
• Review whether exports can be limited by role, geography, purpose, and volume.

6. Design the proof of concept

Pre-register the market, category, time window, required fields, exclusions, channel, offer, owner, control, measurement period, and stop criteria. Use a representative cohort. Do not let the vendor choose only the easiest geography after seeing the data.

Score usable-record rate, duplicates, contactability, delivery, engagement, qualified progression, conversion, total cost, operational burden, complaints, and opt-outs. Preserve cohort identifiers so the team can explain the result.

7. Normalize commercial terms

Compare subscription minimums, seats, export limits, overages, contract term, renewal, exclusivity, replacement rules, permitted use, data retention, services, integrations, and termination rights. Divide the total program cost by usable outcomes—not by rows in a file.

8. Establish the review cycle

Assign an owner for source documentation, product claims, compliance changes, security evidence, campaign outcomes, and consumer feedback. Time-sensitive claims should include a verified date and a review-by date. A good vendor relationship becomes more transparent as the system evolves.