We started Mailapp because we kept building the same thing inside every company we joined. A spreadsheet that segments users. A Zap that pipes an event to a tool that drops it on the floor. A Slack channel for the "email people" to ask why a campaign didn't send. A quarterly migration to whichever ESP looked least bad.
We think the email tool of the next decade has to do four things really well.
1. Treat the audience as a stream, not a list.
Your customers are not rows in a CSV. They are streams of events with traits that change minute to minute. An email tool that treats your audience as a static list will be wrong every Friday afternoon. A tool that treats it as a stream — that re-evaluates a segment the moment a user opens the app, upgrades, refunds, cancels, logs in from a new device — will be right all the time.
This is why we built a real customer data platform inside Mailapp. Not a "lists" tab with extra fields. Events, traits, computed properties, profiles, the whole thing.
2. Treat automations like code.
Lifecycle automations are some of the most consequential code in your business. They onboard your users, save your churn, recover your carts, deliver your receipts. And yet most are built in a UI that doesn't support diffing, doesn't support code review, doesn't support version control, and certainly doesn't support letting an LLM author them.
Mailapp's automations are a graph. Every change is a commit. You can diff, revert, protect branches, require approvers, and re-run a contact through the entire graph for inspection. We did this on purpose. Lifecycle code deserves the same engineering treatment as production code.
3. AI should operate, not narrate.
The first wave of "AI in your ESP" was a chatbox that summarised dashboards you could already read. That's not interesting. The interesting thing is an assistant that, when you say "something is wrong with our open rate this week," doesn't reply with a paragraph — it opens the segment, sorts by delta, finds the broken campaign, drafts the fix, stages the change, and says here's what I'd do, approve or edit.
That's the bar. Every AI feature we ship is held to it. If the AI can't take the action itself, we'd rather ship a better UI than a worse chatbox.
4. The boring middle is the product.
Demos win the first call. The boring middle — deliverability, residency, sub-processors, render fidelity in Outlook for Windows, SCIM provisioning, audit logs, the way DMARC fails open vs closed — wins the deal. We work hard on the boring middle. It's where most ESPs cut corners. It's where we don't.
A few things we promise.
- Pricing will always be on the website.
- Free tier will always exist. You will never lose access to your data.
- You will never be paused for an engagement dip without warning.
- Changelog will always be dated and signed.
- Post-mortems for serious incidents will always be public.
- Documentation lands with the feature, not a quarter later.
- We will say no to features that look great in a demo and bad in a year.
If any of this resonates, give Mailapp a try. We think you'll like it. If it doesn't, we'll help you migrate back. The tools you choose are too important to be settled by inertia.
— Mateo, Aria, Pat
Co-founders