What Mandrill users can do before the Mailchimp sale closes
On 19 November 2026 Intuit said it had agreed to sell Mailchimp to Bending Spoons at an enterprise value of $3.68 billion. Intuit paid about $12 billion for it in 2021. The deal needs regulatory approval and is expected
On 19 November 2026 Intuit said it had agreed to sell Mailchimp to Bending Spoons at an enterprise value of $3.68 billion. Intuit paid about $12 billion for it in 2021. The deal needs regulatory approval and is expected to close in early 2027.
Mandrill, which Mailchimp now calls Transactional Email, goes with it. It is an add-on to a paid Mailchimp plan, so whatever happens to Mailchimp's plans and teams happens to Mandrill too.
A company I used to work for sent everything through Mandrill for years, and I wrote earlier about what it cost to get the templates out. This post is narrower. It covers what has been announced, what the buyer did with its earlier purchases, and a few things that are cheap to do while nothing has changed yet.
What has been announced
Bending Spoons is buying 100% of The Rocket Science Group, the company behind Mailchimp. Its press release says that after closing it will "invest substantially in the business". Intuit says it wants a smooth transition for customers, partners and employees.
Neither statement mentions Mandrill or transactional email. No pricing or plan change has been announced, and until the deal closes Mailchimp is still run by Intuit. Anyone who tells you today what Mandrill will cost in 2027 is guessing.
One number from Intuit's earnings report the same day: Mailchimp's revenue fell 1% year over year, to $313 million for the quarter.
For reference, Mandrill's current pricing is blocks of 25,000 emails at $20 each (up to 500,000 a month), on top of a Mailchimp Standard plan that starts at $20 a month. Sending 50,000 emails a month costs from $60. I checked this on 13 November 2026.
What Bending Spoons did with earlier purchases
This part is history, and I don't want to turn it into a forecast.
Evernote is the best documented case. According to TechCrunch, 129 Evernote employees were laid off in February 2023, and in November 2023 the free plan was limited to 50 notes. The Pragmatic Engineer, citing ZDNet, writes that a plan that cost $37 a year before 2023 was $250 a year by 2026. WeTransfer was bought in 2024, and Bending Spoons said it would cut about 75% of the staff. Vimeo's acquisition closed in late 2025 and was followed by layoffs that reached most of the company.
The same Pragmatic Engineer piece describes the other side. Bending Spoons moved Evernote's data off 750 manually provisioned VMs and split the old monolith into services in about six months, without disruption for users, and it kept shipping features afterwards.
Most of these products are consumer subscriptions. Mandrill is a B2B add-on billed in blocks, and I don't know how much of that pattern carries over. What the history does show is a buyer that changes pricing and teams after it takes over.
Mandrill also has its own history with short notice. In February 2016 Mailchimp told Mandrill users to attach a paid Mailchimp account within 60 days or find another service (HN thread from that week). Sixty days is not much if your templates live there.
What you would have to move
The sending call is the small part. If your code calls messages/send with ready HTML, any provider accepts the same thing with different field names. The rest is state that lives on Mandrill's side:
- templates, each with a draft and a published copy, written in Mailchimp merge tags (
*|FNAME|*,mc:edit) or in Mandrill's Handlebars - the rejection denylist: hard bounces, spam complaints and unsubscribes
- webhooks and their signing keys
- inbound routes, if you receive mail through Mandrill
- sending domains, where DKIM and Return-Path records in your DNS point at
mandrillapp.com - subaccounts, tags, and messages scheduled with
send_at
Code that calls messages/send-template is the hard part, because Mandrill renders those emails and your application has never seen the final HTML.
Things worth doing now
None of these commit you to leaving.
Export templates and the denylist. The script below uses two endpoints from Mailchimp's Transactional API reference, templates/list and rejects/list. It needs Node 18 or newer and nothing else. Create an API key in the Transactional settings (SMTP & API Info), run the script on your own machine, and it writes everything into a mandrill-backup folder next to it.
// export-mandrill.mjs
// Run: MANDRILL_KEY=your_api_key node export-mandrill.mjs
import { mkdir, writeFile } from "node:fs/promises";
const API = "https://mandrillapp.com/api/1.0";
const key = process.env.MANDRILL_KEY;
if (!key) throw new Error("Set MANDRILL_KEY to a Transactional API key");
async function call(path, body = {}) {
const res = await fetch(`${API}/${path}`, {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({ key, ...body }),
});
if (!res.ok) throw new Error(`${path}: ${res.status} ${await res.text()}`);
return res.json();
}
const out = "mandrill-backup";
await mkdir(`${out}/templates`, { recursive: true });
const templates = await call("templates/list");
for (const t of templates) {
const dir = `${out}/templates/${t.slug}`;
await mkdir(dir, { recursive: true });
await writeFile(`${dir}/published.html`, t.publish_code ?? "");
await writeFile(`${dir}/draft.html`, t.code ?? "");
await writeFile(`${dir}/published.txt`, t.publish_text ?? "");
const { code, text, publish_code, publish_text, ...meta } = t;
await writeFile(`${dir}/meta.json`, JSON.stringify(meta, null, 2));
}
const rejects = await call("rejects/list", { include_expired: true });
await writeFile(`${out}/rejects.json`, JSON.stringify(rejects, null, 2));
console.log(`${templates.length} templates, ${rejects.length} denylist entries saved to ./${out}`);
Two limits. Templates built in Mailchimp's visual builder and published to Transactional are listed by a separate endpoint, mctemplates/list. And if the denylist is large, exports/rejects builds a downloadable file instead of returning everything in one response.
The denylist matters more than it looks. If you start on a new provider without it, the first weeks of sending go to addresses that already hard bounced or complained, and that is a bad way to introduce a new sending domain setup to Gmail.
Count the template sends. Search your code for send-template (or sendTemplate in the client libraries) and compare with plain send calls. That ratio is the size of the migration. A codebase with three template sends is an afternoon. One with eighty is a project.
Render one template outside Mandrill. Take the most complicated one, with conditionals and a loop, convert its merge tags to whatever you would use next, and render it with a real payload. Mandrill's templates/render endpoint returns its own output for the same payload, so you have something to diff against. At my old company this was the step that took longest, because we had no test payloads and no expected output for any email.
Write down the DNS records. Note which DKIM selectors and which Return-Path CNAME belong to Mandrill. A second provider can be authenticated on the same domain with its own selectors while Mandrill keeps sending, so the switch itself can be a config change.
Where the templates go
The export gives you HTML in Mailchimp's dialect, so wherever you go, the templates need a new home that renders without Mandrill. I see three options: files in your repo (React Email or MJML, rendered at send time), your own database with an editor in front of it, or a separate template layer that sends through whatever provider you choose.
I'm the author of Senlo, an open-source tool of the third kind: an email editor, versions and automations that send through your own Resend, Postmark, SES or SMTP account. It does not import Mandrill templates yet, so today moving to it means rebuilding them in the editor. What you get is that the next provider change is new credentials in settings.
On price, any of the common providers is cheaper than Mandrill at small volume. For 50,000 emails a month, read from pricing pages on 9 November 2026: Amazon SES $5 to $8, Resend $20, Mailgun $35, Postmark $55, against Mandrill from $60.
When this can wait
If Mandrill is only a pipe for HTML you render yourself, switching takes a day and you can wait for an actual announcement. The same goes if you use Mailchimp for marketing, the bundle works for you, and you have a handful of templates.
Teams with dozens of templates stored in Mandrill are in a different position. That work does not get smaller by waiting, and the last time the terms changed the notice was 60 days.
Sources, checked on 20 November 2026: Bending Spoons press release (SEC filing), CNBC on the deal, Yahoo Finance on Intuit's quarter, TechCrunch on the WeTransfer acquisition and Evernote, TechCrunch on Bending Spoons and Vimeo, Euronews on Bending Spoons, The Pragmatic Engineer on its acquisition strategy, Mailchimp Transactional pricing, Transactional API reference.
Senlo: github.com/IgorFilippov3/senlo Β· senlo.io
Originally published by Dev.to WebDev. Aggregated on AIWithGhost for educational purposes β full credit and traffic to the original publisher.