Definition
Email service provider (ESP) selection is the structured evaluation and procurement process for choosing the technology platform that will manage an organisation's email marketing programme. Given the centrality of the ESP to email operations — it handles subscriber data management, template rendering, send execution, deliverability infrastructure, analytics, and often automation and personalisation — the selection process requires rigorous, quantifiable evaluation across multiple weighted criteria. The standard evaluation framework weights criteria by relative importance: deliverability capabilities (25%), reflecting the fundamental requirement that emails reach the inbox; automation and campaign management functionality (20%), representing the operational core of the platform; analytics and reporting depth (15%), enabling performance measurement and optimisation; API quality and integration ecosystem (15%), determining how well the platform fits into the broader marketing technology stack; scalability and reliability (10%), ensuring the platform can handle peak volumes and growth; compliance and security features (10%), increasingly critical in the evolving regulatory landscape; and support and customer service quality (5%), acknowledging the ongoing partnership aspect of ESP relationships.
The selection methodology progresses through four stages: market discovery and RFI (request for information) to identify potential vendors from the hundreds of available ESPs, ranging from enterprise platforms such as Salesforce Marketing Cloud and Oracle Responsys to mid-market solutions such as Klaviyo and Braze to specialised platforms for specific industries; vendor shortlisting based on weighted scoring against the defined criteria, typically reducing the candidate pool to three to five vendors; structured proof-of-concept demonstrations where shortlisted vendors execute standardised test scenarios covering data import, template creation, segmentation, automation workflow building, integration testing, and analytics reporting; and final platform selection with total cost of ownership (TCO) calculation including licensing, implementation, migration, training, and ongoing operational costs. The process typically takes three to six months from initiation to contract signature.
Best Practices
Define and weight evaluation criteria before beginning any vendor research, ensuring the weighting reflects your organisation's specific priorities rather than generic industry defaults. An organisation with sophisticated automation requirements may weight automation capabilities at 30% rather than the standard 20%, while a programme prioritising global compliance might weight compliance and security at 20%. Document the weighting rationale and obtain stakeholder consensus before proceeding to vendor evaluation.
Conduct structured proof-of-concept demonstrations with standardised test scenarios that simulate your actual email programme requirements, not generic use cases provided by the vendor. Test scenarios should include: importing a sample subscriber data set with your data model and segmentation logic; building a complex automation workflow with decision branches, delays, and dynamic content; creating an email template that matches your design system specifications; establishing a third-party integration via the platform's API; and generating a standardised analytics report. Each scenario should have defined success criteria and be scored consistently across all vendors.
Evaluate total cost of ownership over a three-year period, not just first-year licensing costs. Include: platform licensing fees (account-based, contact-based, or send-volume-based pricing models), implementation and migration costs, template development and design system adaptation, integration development for third-party systems, data migration and quality assurance, training costs for platform learning and certification, and ongoing operational costs including platform management headcount, additional API usage fees, and volume-based overage charges.
Assess deliverability capabilities through a demonstration of platform infrastructure, not just feature documentation. Request: the platform's IP management approach (dedicated versus shared IPs, IP warmup support, automated feedback loop registration), authentication management (SPF, DKIM, DMARC configuration support, BIMI capabilities), deliverability monitoring and alerting capabilities, and access to dedicated deliverability expertise through the platform's professional services or customer success team. Verify actual email client support and rendering capabilities through independent testing.
Include contractual protections in the service agreement that address: data ownership and portability (explicit ownership of all subscriber data, templates, and configurations, with guaranteed data export capabilities and format), service level agreements for send latency, uptime, and support response times, data processing agreements compliant with applicable regulations, termination assistance including data extraction support and transition timeline, and lock-in prevention through clear data extraction and migration support provisions.
Related Glossary Terms
Frequently Asked Questions
Standard weightings are: deliverability (25%), automation capabilities (20%), analytics and reporting (15%), API quality and integrations (15%), scalability and reliability (10%), compliance and security (10%), and support and customer service (5%). These should be adjusted to reflect your organisation's specific strategic priorities and technical requirements.
A POC should test scenarios that mirror your actual programme requirements: complex data import and field mapping, automation workflow creation with your logic and triggers, template creation matching your design system, API integration with your tech stack, segmentation using your criteria, and reporting that covers your KPIs. Scoring each scenario consistently across vendors enables objective comparison.
Calculate TCO over a three-year period including: platform fees (licensing model based on contacts, sends, or accounts), implementation costs (platform setup, template development, integration development), migration expenses (data transfer, validation, parallel run), training costs (platform access, certification, team training), and ongoing operational costs (dedicated headcount, overage charges, API usage fees, premium support).
Essential protections include: explicit data ownership and portability clauses, service level agreements with measurable targets and penalty provisions, data processing agreements compliant with GDPR and other applicable regulations, termination assistance with defined data extraction support, access to platform improvements and updates, and liability limitations that are clearly defined and commercially reasonable.
The full process from initial research to contract signature typically takes three to six months. Phase one (needs assessment, criteria definition, market discovery) takes three to four weeks. Phase two (RFI, initial evaluation, shortlisting) takes four to six weeks. Phase three (structured POCs, detailed evaluation, references) takes six to eight weeks. Phase four (negotiation, contracting, procurement) takes four to six weeks.