B2B software research · email
Email deliverability benchmark platform: sourced service data, two separate scores and free authentication tools
An independent reference site that compares email services on a sourced feature matrix, keeps deliverability and platform scores in two separate engines and offers free DNS and email-authentication tools.
- Client
- Client confidential
- Industry
- B2B software research · email
- Year
- 2026
Challenge
Every email platform says its deliverability is excellent, and most comparison sites repeat vendor feature lists. Buyers choosing a marketing or transactional email service need dated, sourced facts and a method that keeps paid features apart from what actually reaches the inbox. Senders also need practical tools to check the DNS records that decide whether their own mail authenticates.
Solution
We designed the platform around its data model first. Service facts live in git as YAML and CSV, with a source note on every cell of the feature matrix, and compile to JSON at build time. Two scoring engines run side by side: a deliverability score computed only from dated inbox-placement snapshots, and a platform score built from a weighted editorial rubric. On top of the data sit a service catalog, head-to-head comparisons, pricing hubs, an MDX content system and a set of free SPF, DKIM, DMARC and DNS tools.
How it connects
- 01 Service research
- 02 Git-based data model
- 03 Test snapshots
- 04 Two scoring engines
- 05 Site, tools and content
- 06 Build gates
What we delivered
- Information architecture: service profiles, rankings, category hubs, head-to-head comparisons, pricing hubs, tools, mailbox-provider requirement pages, methodology, blog, glossary and news
- Git-based data model with a source note on every feature-matrix cell, a pricing dataset across sending volumes and append-only test snapshots
- Two separate scoring engines: a versioned deliverability formula and a weighted platform rubric
- Free deliverability tools: DNS checkers that run on the server and record generators that run in the browser
- MDX content system with an editorial calendar and date-gated publishing
- Technical SEO: dynamic sitemap and robots rules, JSON-LD, generated Open Graph cards
- Build gates before every production build and an automated test suite that also covers SEO metadata
Functionality
- SPF, DKIM, DMARC and BIMI checkers backed by server-side DNS lookups with per-client rate limiting
- MX lookup, reverse-DNS (FCrDNS) check and DNSBL blocklist check
- SPF and DMARC record generators that run in the browser
- Email header analyzer that runs entirely in the browser
- Full email-DNS audit, authentication summary, deliverability setup score and troubleshooter
- DMARC aggregate-report analyzer
- Deep links that open any tool with the domain already filled in
- Services without a test snapshot are marked as untested instead of receiving an estimated score
- Scheduled pages go live on their publish date through hourly revalidation, without a redeploy
- Newsletter sign-up protected by a honeypot field and rate limiting
Integrations
- Server-side DNS resolution for the checkers
- Public DNSBL zones for the blocklist check
- Dated review-platform rating snapshots (Trustpilot, G2, Capterra)
- Vercel Blob storage for newsletter sign-ups
- Vercel Web Analytics
Technologies
Result
Built and operated, live and indexed. The platform publishes service profiles in several review tiers, the free tools, a blog, a glossary, news, mailbox-provider requirement pages and methodology pages, with more articles scheduled. Scores are recomputed from data at build time, so a new test snapshot or a price change is a data edit, not a development task.
How is the data kept?
The platform has no runtime database. Service data lives in git as YAML and CSV: a feature matrix of services against grouped features, a pricing dataset across sending volumes, append-only test snapshots, review-rating history, and an AI-readiness dataset that records official MCP servers, agent skills and developer docs for each service. The build compiles it all to JSON. Each change is reviewable in a diff, and each published value traces back to a dated line in a file.
Two scores that stay apart
Features and pricing say little about whether mail lands in the inbox, so the platform scores them separately. The deliverability score comes only from dated inbox-placement snapshots taken under a controlled seed-list protocol, with a versioned formula. The platform score is a weighted editorial rubric of features, pricing and reputation. User ratings from review platforms appear next to both as context.
Gates before every build
Before each production build, scripts validate frontmatter, audit internal links, recompute and verify every score, run a confidentiality scan and emit the rankings data. Tests also enforce title and description length and uniqueness for every page, so SEO metadata cannot drift as the content grows.
Discuss your project
Tell us about your project