Skip to content

Research what Resend offers, what users love, and what they miss #527

Closed

Description

@Silthus

Question

What does Resend offer as a product, which features do its users love most, and what do they miss or dislike?

Michael (2026-10-20) plans a full attack on Resend to move many of its organizations to PostHog. This survey finds where Resend is strong, so PostHog matches it, and where Resend users hurt, so PostHog wins them.

Answer with links for every claim:

  1. The product. Everything Resend offers today: sending API and SDKs (languages), React Email and its editor, hosted templates, domains and DNS setup, API keys and permissions, webhooks, logs, broadcasts and audiences or contacts, topics and subscriptions, inbound email, scheduled and batch sends, idempotency, test addresses and test mode, analytics, dedicated IPs, MCP server or agent tooling, regions, compliance. Pricing tiers and their limits (free tier, daily and monthly caps, retention). Note launch dates where they show the pace.
  2. Most loved. What developers praise, with sources: Hacker News, Reddit, X, Product Hunt, G2 and similar reviews, blog posts, YouTube, Resend's own customer stories, and GitHub stars and issues on resend/resend-node and resend/react-email. Rank the top features by how often they come up.
  3. Missing and disliked. Complaints and gaps with sources: deliverability, rate limits and caps, pricing jumps, support, account suspensions, outages, missing analytics or person context, marketing features, multi-environment setup, template management, webhooks reliability, no workflow or automation. Rank by frequency and severity. Include open GitHub issues and feature requests with many reactions.
  4. Who switches and why. Reasons people leave Resend or pick it over SendGrid, Postmark, Mailgun, Loops, Customer.io and AWS SES. Which of those reasons does PostHog already answer (analytics on the person, replay, flags, workflows, one bill), and which would it need to build?
  5. Agent and developer experience. How Resend reaches coding agents: docs for LLMs (llms.txt), MCP server, agent skills, onboarding (the "send your first email" flow), templates gallery. What makes an agent pick Resend when asked to "send an email"?

Write the report on branch research/resend-product-and-sentiment. Resolve this ticket with a ranked summary (top loved, top missing, top switching reasons, PostHog's openings) and the branch pointer.

Activity

  1. Silthus commented on Oct 20, 2026

    @Silthus
    OwnerAuthor

    Runtime: T3 Code
    Task: node:delegated-task:command%3Amcp%3Acbc48805-d1ea-4076-9165-d9cb3e20ac16%3Adelegate-task%3Amap461-r527-v1
    Child thread: thread:delegated-task:command%3Amcp%3Acbc48805-d1ea-4076-9165-d9cb3e20ac16%3Adelegate-task%3Amap461-r527-v1
    Cast: Sonnet 5.5 high. Wall-time cap: 60 minutes.

  2. self-assigned this
    on Oct 20, 2026
  3. Silthus commented on Oct 20, 2026

    @Silthus
    OwnerAuthor

    Claimed by the research sub-agent (Sonnet 5.5, high). Report goes to branch research/resend-product-and-sentiment. Wall-time cap: 60 minutes.

  4. Silthus commented on Oct 20, 2026

    @Silthus
    OwnerAuthor

    Resolution

    Report: research-resend-product-and-sentiment.md on branch research/resend-product-and-sentiment (commit 250c338f900). No PR, per the map's research convention. Builds on #522; not repeated.

    Top loved (by mentions across the sample)

    1. Developer experience and time to first email. Product Hunt tags: "developer-friendly API" 27, "easy setup" 25, "fast integration" 17.
    2. React Email: 9 Product Hunt mentions, 19.8k stars.
    3. No approval and no sandbox, so sending starts on day zero (HN).
    4. Dashboard, logs and event visibility. 5. Perceived reliability (polarized). 6. Pace: 54 changelog entries in 2026.

    Top missing or disliked (frequency and severity)

    1. Deliverability opacity on shared SES infrastructure. Worst case is silent loss: Anukari, Feb 2026. One detailed case plus drive-by "SES wrapper" remarks, so thin on frequency.
    2. Price shape: 100 per day free cap, $20 floor, per-domain cost, $35 to $90 jump at the same 100k volume (pricing, HN). Most frequent complaint.
    3. Automatic pause above 4% bounce or 0.08% complaint rate (Resend). First-person evidence is thin.
    4. No environments, message streams or sandbox. 5. Shallow analytics and no person context. 6. React bundling friction (closed GitHub issues). 7. 10 requests per second per team, same on every plan.
    • GitHub is a weak demand signal: open issues on resend-node and react-email carry 0 reactions; the top-reacted ones are all closed.

    Top switching reasons

    • Into Resend: simpler DX and instant start, from SendGrid and Postmark (Raycast, Depot).
    • Out of Resend: deliverability trust (to Postmark), cost at volume (to SES), pay-per-use (to Scaleway, Cloudflare Email Service launched 2026-04-16 at $0.35 per 1,000). Mostly stated intent, one first-person case.

    PostHog's openings

    • Lead with person context: templates read person.properties inside workflows; Resend's contacts are a separate list. Only uncontested differentiator.
    • Price is a real opening (SES $0.10, Cloudflare $0.35, Resend overage $0.90 per 1,000), but PostHog's email price was not checked here. Needs a decision.
    • Customer.io import already exists in the repo; Resend ships five migration guides and nobody ships an agent skill for leaving Resend.
    • Deliverability is not an opening yet: PostHog also sends through SES, so the "SES wrapper" critique applies. Needs per-team reputation isolation and visible suppression data.
    • Agent reach: Resend wins with llms.txt, .md page twins, OpenAPI, hosted OAuth MCP, CLI, five skills and plugins for Claude Code, Codex and Cursor. PostHog has template skills and MCP; the send-side surface is the gap (see Research what a switch from Resend needs from posthog-node and a skill #522).

    Not verified

    • Reddit and G2 were unreachable (Reddit blocked in every fetch path; G2 only via a search summary). X gave two posts, both announcements. The sample is HN, Product Hunt, Resend's own pages, GitHub and blogs.
    • Rankings beyond the Product Hunt tag counts are judgment over a small sample.
    • PostHog's domain features, inbound, webhooks, SDK breadth and email pricing were not checked in the repo.
    • No hands-on test of Resend's MCP, CLI or deliverability. "What makes an agent pick Resend" is inference from Resend's material and npm download counts.

    Two things for Michael:

    • Lead with person context, price, or agent-led migration? Evidence supports person context; price needs a number.
    • Is "no worse than Resend on deliverability" enough given both sit on SES? If not, isolation needs its own ticket.

    I did not edit the map's Decisions-so-far; the dispatcher owns that.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Projects

No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions