Mailchimp gave away the growth engine that built it

Most companies price a free tier the way they price a discount, as revenue they have agreed to forgo and hope to claw back later. Mailchimp treated it as the acquisition channel, and it was cheaper than anything the company could have bought.

The numbers from the year after launch are the sort you assume have been rounded up by someone in marketing. Users went from 85,000 to 450,000 in twelve months. That is a five-fold jump with no paid media story attached to it, driven by the simple fact that a small business owner with no budget could start sending email that day.

Free did not cannibalise paid. Paying customers rose by more than 150% and profit rose 650%, because the free users were the top of a funnel that had never really existed before, and every one of them became a candidate for an upgrade the moment their list outgrew the limit.

By 2016 the compounding had done its work. Over $400 million in revenue, 16 million users, 246 billion emails sent that year, and around 550 employees carrying the whole thing. Work out the revenue per head and you get a business generating more than $700,000 per employee without a sales team knocking on doors.

What Mailchimp had built by that point was rare. A self-funded, self-serve machine where usage created word of mouth, word of mouth created signups, and signups created upgrades on a delay. Nobody had to defend a CAC number in a quarterly review because the funnel filled itself while everyone slept.

Sixty days. That was the window Mandrill users got on 24th February 2016 to move their lists, rewire their transactional email setup and attach it all to a paid Mailchimp account starting at $10 a month. For a developer running a production system, sixty days is an eviction notice with a payment plan attached.

What makes it strange is the year it happened in. Mailchimp in 2016 was doing over $400 million in revenue with roughly 550 employees, 16 million users and 246 billion emails sent. The ten dollars was never the point. Mailchimp had spotted a cohort using its infrastructure without paying and chose to convert them by force rather than by making the paid product more obviously worth buying.

Developers are the cheapest acquisition channel any email company will ever have, because they embed you and then never think about you again. They also have the longest memories in software. I have watched teams carry a grudge about a pricing change for five years and quietly veto the vendor in a meeting nobody documented. Mandrill taught a generation of engineers that Mailchimp's free tier came with an expiry date.

The same shape returned in 2023, when the free plan fell from 2,000 contacts to 500 and from 10,000 monthly sends to 1,000. Each time, the revenue was pulled forward out of the exact pool of small senders that produced the 16 million user number in the first place.

Fourteen percent. That is what the largest email marketing company on earth managed during the nine months when US ecommerce sales went from roughly $570 billion to $815 billion. A quarter of a trillion dollars of new online spending appeared, most of it flowing through merchants who needed exactly one thing Mailchimp sold, and Mailchimp grew slower than a decent B2B SaaS company in a flat market.

Look at the revenue line and the problem is visible well before 2020. $525 million in 2017, $600 million in 2018, $700 million in 2019, $800 million in 2020. Stairs rather than a curve, roughly $75 to $100 million added each year regardless of what the market was doing, which tells you the business was adding customers at a steady mechanical rate rather than riding anything. When the biggest tailwind in the history of online retail arrived, the staircase carried on climbing at the same pace.

The reason, I think, is that Mailchimp kept selling a newsletter tool to everybody while the money moved to lifecycle marketing built specifically for stores. An abandoned cart flow, a post-purchase sequence, product feed sync, revenue attributed per email. A merchant doing $2 million a year does not want a general-purpose broadcast tool with an ecommerce add-on bolted to the side. They want software that already knows what a variant is.

Shopify settled the question in August 2022 by putting $100 million into Klaviyo. When the platform your customers live on funds your competitor, the positioning argument is over. Mailchimp had spent years being the email tool for everyone and ended up as the default for the segment with the least money and the least urgency.

You cannot buy a self-serve funnel with borrowed money and expect it to keep behaving like a self-serve funnel. Intuit paid roughly $12 billion in 2021 for a business doing $800 million in revenue and growing about 20% a year, and part of that was funded with a $4.7 billion term loan. Debt has a schedule. Free plans do not.

Intuit broke Mailchimp out exactly once, in Q4 FY22, at $265 million, or about $1.06 billion annualised. After that the number only shows up as a subtraction. Global Business Solutions grew 15% to $3.3 billion, or 17% without Mailchimp. Online Ecosystem grew 19% to $2.5 billion, or 22% without Mailchimp. When your asset is being quoted as the thing you exclude to make the chart look better, the internal conversation has already moved on from growth.

At that point there are two ways out. Spend to fix acquisition, which is slow and expensive and takes a couple of years to show up. Or squeeze the base you already have, which shows up next quarter. Intuit went for the squeeze. In April 2026, legacy accounts created before May 2019 that had never migrated took an 11% to 13% price increase. The daily send cap was cut from 500 to 250 in February 2026. Both are revenue actions dressed as product decisions.

Then came the 17% workforce reduction, more than 3,000 people, with Mailchimp specifically named as an area of reduced investment, and around $300 million of restructuring charges landing mostly in the quarter ending July 2026.

A roadmap written by a term loan does not build growth loops. It builds invoices, and it calls them features.

The 500 contact, 1,000 send free tier is a line item somebody trimmed to make a quarter look tidier. I would put it back to 2,000 contacts and 10,000 sends tomorrow, and I would book the cost in the acquisition budget rather than in gross margin, because that is what it actually is. Mailchimp already ran this experiment once and got 85,000 users to 450,000 in a year out of it. Cutting the daily send cap from 500 to 250 last February is the opposite bet, made against a funnel that has already lost five million users.

Then the harder half. Klaviyo is running 110% net revenue retention and 4,175 accounts paying more than $50,000 a year, up 38%. Those numbers come from customers spending more because the product does more for their store as it grows, rather than from squeezing the ones who cannot leave. Klaviyo manages it at 16% non-GAAP operating margin with revenue per employee above $600,000, which kills the argument that generosity at the bottom and discipline in the P&L are opposites.

Four moves, in the order I would run them.

Geography is where I would push hardest. Klaviyo's revenue outside the Americas grew 39% last quarter and EMEA excluding the UK grew 51%, so the land grab is still open in markets where neither brand has been settled as the default. Mailchimp still has name recognition with small senders almost everywhere, and that is the one asset the squeeze has not spent yet. I would spend it now, while it still buys something.