A free email client still has engineers to pay, servers to run, and investors to satisfy. Nothing about reading and organizing mail is free to build or operate. So when the price to you is $0, the money is coming from somewhere else. This post walks through where, concretely, without the fearmongering.
The three common models
Ads inside mail. The oldest model. A free webmail or client inserts ads into your inbox view, sometimes contextual to the content of the emails you are reading. To target those ads well, the provider needs to know what is in your mail: purchase confirmations, travel plans, health topics, financial details. This was Gmail's original model until 2017, when Google stopped scanning consumer Gmail for ad targeting after years of public pressure. Many smaller free clients still do a version of this.
Data aggregation and resale. Less visible, more common than people assume. A free client with access to your inbox can extract structured signals (receipts, subscriptions, travel bookings, shopping habits) and sell an aggregated, “anonymized” version of that data to market research firms. This is a documented business model: Slice and Edison Trends, both owned by Rakuten's Slice Technologies, built exactly this on top of free email and shopping apps, parsing purchase receipts out of millions of inboxes to sell trend data to retailers and hedge funds. The user agreement disclosed it. Almost no user read that far.
Training AI on your inbox. The newest model, and the fastest growing. A free client with an “AI assistant” feature needs training data to make that assistant better. Some products use aggregate, permissioned, opt-in data. Others have opt-out defaults buried in a settings menu, meaning your email content trains a model unless you find and disable it yourself. The distinction between opt-in and opt-out defaults is the whole ballgame here, and it is rarely obvious which one a product uses without reading the actual policy text.
Why this is a rational business decision, not a scandal
It is worth being fair here: none of this is secret villainy. Building and maintaining an email client is expensive. If a product is free, one of two things is true: it is subsidized by a larger paid product (like Gmail is subsidized by Google's ad business broadly), or it is monetizing your data directly. Both are legitimate business choices that a company can make. The problem is not that these models exist. The problem is that the tradeoff is rarely stated plainly at the point of signup.
What to actually look for
Skip the vague privacy-policy language and check for these specific things:
- Does the privacy policy mention advertising partners, data brokers, or a “data-sharing” section? If yes, read exactly what category of data is shared and whether it is opt-in or opt-out.
- Is there a paid tier with the same features minus the data use? If a company offers an identical product for money without the data practices, that is a strong signal about where the free tier's revenue actually comes from.
- Does the AI feature have a toggle for training? And if so, is it on by default? Check
Settings > Privacyor equivalent before you assume it is off. - How old is the company, and has it been acquired? Data practices frequently change after an acquisition, when a new owner looks for ways to monetize an existing free user base. A free tool that stayed stable and free for years, without an acquisition, is a smaller risk than a startup that just got acquired by an ad-tech company.
Where STAMP is different, and why
STAMP has no free tier. Pricing is Monthly $9, Yearly $79, or Lifetime $199. There is no ad-supported path and no data-resale business hiding behind a lower price. Classification runs on-device, meaning your email content is not sent to our servers for processing in the first place, so there is no inbox data sitting on a server to monetize even if we wanted to. We do not train models on user email, and we never will without an explicit, paid, opt-in agreement we have no plans to build.
This is not altruism. It is math. A subscription business has one customer to satisfy: the person paying. A free, ad-or-data-funded business has two customers: the user and the advertiser or data buyer, and the second one usually wins ties. We would rather have one customer whose interests are fully aligned with ours.
The honest tradeoff
Paying $9 a month for an email client is a real cost, and free alternatives are genuinely fine for a lot of people, especially casual, low-stakes email use. The tradeoff is not “free is evil, paid is virtuous.” It is: figure out what a specific free product's business model actually is, decide if that tradeoff is acceptable for what you use email for, and choose accordingly. For a founder's confidential fundraising thread or a lawyer's client correspondence, that tradeoff usually looks different than it does for a personal inbox you check twice a week.
Where to go from here
For a deeper look at what “on-device” commits a company to, what “on-device” actually means. For our own reasoning on data as a liability rather than an asset, why we don't want your data.
No ads. No data resale. Ever. hello@stamp.email